Equity Gilt Study
Equity Gilt Study
Equity Gilt Study
Larry Kantor
Head of Research, Barclays
24 February 2015
Jim McCormick
Head of Asset Allocation, Barclays
1
CONTENTS
Chapter 1
Population dynamics and the (soon-to-be-disappearing) global savings glut
Chapter 2
Adjusting to a world of lower oil
26
The magnitude and speed of the collapse in oil have roiled markets; only the selloffs in 199798, 1986 and 2008 were larger than the recent one. In order to assess the sustainability of
lower oil prices and their effects on the global economy and markets, we construct a model to
explain real WTI oil prices based on the global demand-supply balance for crude, global IP,
OPEC market share and real US power prices. The medium-term drivers in our model suggest
that lower oil prices are likely to persist. Demand growth is slowing, driven by energy
efficiency and lower aggregate growth globally. Moreover, oil should remain a well supplied
market, with US tight oil keeping OPEC in check.
Chapter 3
EM is still an attractive asset class
60
The external backdrop for EM economies has grown tougher since 2011 and will likely
remain so over the next few years. On the domestic front, progress on structural reforms
has been disappointing. But EM economies have evolved since the start of the boom years
in the early 2000s, with many of their macroeconomic and financial vulnerabilities now
reduced. When we look at EM in the context of a global portfolio, the gap between EM and
DM risk premia is significant. Thus, we think allocations to EM assets make sense even if
asset returns are likely to be much lower than in the boom years.
Chapter 4
The great destruction
84
Severe recessions intertwined with financial crises have historically been associated with lost
output and slower potential growth. In applying a uniform framework across seven developed
economies that account for nearly half of world output, we estimate that potential growth in
these economies has fallen by 1.5pp since 1999 and, in turn, has reduced global potential
growth by 0.7pp. Our finding that slower growth in developed economies could slow global
growth by 0.7pp is of similar magnitude to the effect of a slowing China on global growth.
Slower potential growth in developed economies and a decelerating Chinese economy have
reduced global potential growth by 1.5pp a significant deceleration.
Chapter 5
The decline in financial market liquidity
107
Banking regulation has intensified since the financial and sovereign crises in a global effort
to improve the safety and stability of the financial system. New regulations have materially
improved the stability of the financial system. However, in an effort to reduce the risk of future
fire-sales financed by short-term debt, they have also reduced the supply of safe, short-term,
liquid assets such as repurchase agreements, causing them to trade at lower yields (and, by
extension, higher prices). The reduction in the supply of short-dated safe assets has caused
them to trade at lower yields and resulted in a transfer of fire-sale risk from traditional sources of
liquidity to less traditional ones, exposing end-investors to run risk.
24 February 2015
Chapter 6
India: A step change
123
India is enjoying multiple cyclical and structural tailwinds: the government under Prime
Minister Modi is pursuing an aggressive reform agenda to spur growth and employment;
Indias central bank is enjoying a fresh credibility boost; and the countrys twin deficits are
improving fast. Meanwhile, India remains among the biggest beneficiaries of lower
commodity prices, with inflation softening materially. We expect Indias economy to post
average real growth of 7-8% annually in the coming decade very strong for an economy
exceeding USD 2trn and with a 3% share of global GDP. Against a backdrop of generally
subdued global growth, including in China, we think India could emerge as the worlds
fastest-growing economy in the years ahead.
Chapter 7
FX risk in a multi-asset portfolio
143
After falling to historically low levels between mid-2012 and mid-2014, cross-asset volatility
has risen recently. We think a trend rise in volatility may be forthcoming in a highly
asynchronous global economic recovery, with elevated macroeconomic uncertainty related to
demographic and structural changes across major economies. An increase in foreign
exchange market volatility has the potential to erode returns and raise portfolio-level volatility
in international multi-asset portfolios. We construct a standard equities/bonds international
portfolio and find that higher risk-adjusted returns are achieved, both ex ante and ex post,
through FX hedging of the bond portfolio.
Chapter 8
UK asset returns since 1899
152
UK equities had a lacklustre year and underperformed other developed market indices in
2014. UK nominal total returns were just 1.2%, compared to 2.6% for the German DAX and
10.5% for US equities. The underperformance occurred despite a reasonable growth
backdrop. The UK was one of the few economies where the consensus growth forecast was
revised higher last year. Fixed income and credit had a very strong performance in 2014 as a
result of the deflationary fears fuelled by the oil price decline. Nominal and inflation-linked
gilts posted their best returns since the Euro sovereign debt crisis in 2011.
Chapter 9
US asset returns since 1925
157
US equity returns in 2014 outperformed both developed and emerging markets by a wide
margin as domestic growth remained robust. Despite periodic drags from global growth
concerns and deflationary fears, the upward momentum was maintained throughout the
year. Fixed income markets followed the trends in the UK: nominal bonds were the best
performing asset of 2014, producing a 23% real total return, in sharp contrast to the -13%
of the previous year, when investors first digested the prospect of monetary policy
normalisation by the Fed.
Chapter 10
Barclays Indices
161
We calculate three indices showing: 1) changes in the capital value of each asset class; 2)
changes to income from these investments; and 3) a combined measure of the overall
return, on the assumption that all income is reinvested.
Chapter 11
Total investment returns
186
Our final chapter presents a series of tables showing the performance of equity and fixedinterest investments over any period since December 1899.
24 February 2015
CHAPTER 1
Michael Gavin
the existing very low interest rate environment. But world interest rates have been
fluctuating around a strongly declining trend for more than 30 years. It is a
question of no minor significance whether asset markets will remain so wellsupported and real interest rates correspondingly depressed in the decades
ahead. Bond markets are pricing historically low real interest rates for the
foreseeable future. But we think that a key secular driver of world asset markets has
peaked and will be fading strongly in the years to come.
While other forces have also been at work, we believe (and present some evidence)
that demographic pressure on world saving has been an important secular driver of
upward pressure on asset prices and downward pressure on interest rates in
recent decades. Since the early 1980s, demographic trends have put upward
pressure on saving rates (and, by extension, asset prices) in every systemically
important region of the world except Japan, as rising old-age dependency ratios
have been more than offset by a growing share of mature, high-saving workers.
In the past 20 years, the country with the largest such shift was China, with Korea a
fairly close second. This suggests that demographic pressures have been a major
driver of the boom in emerging Asian, and specifically of Chinese savings.
FIGURE 2
Demographic support for saving has recently been growing
since the 1980s, but is on the cusp of a profound reversal
0.40
10%
0.35
5%
0.30
0%
0.25
-5%
0.20
-10%
0.15
0.10
-15%
0.05
-20%
1960 1966 1973 1980 1987 1993 2000 2007 2014
US
UK
DE
JP
Note: Short term interest rate deflated by forward 12-mo rate of CPI inflation.
Source: Barclays Research
24 February 2015
0.00
1950 1960 1970 1980 1990 2000 2010 2020 2030
Projected
Mature
Elderly
Difference
Were it not for the historic collapse in the price of oil that began in mid-2014, the year would
likely be remembered as the one in which financial markets began to price something like
secular stagnation into global financial markets. In light of the cyclical headwinds that
became apparent in Europe, Japan, and China during the first half of 2014, it is not surprising
that the short end of many yield curves priced lower interest rates. Certainly, the 2014 rally in
global bond markets has been validated by the dovish monetary policy actions of recent
months, including the launch of full-blown quantitative easing by the ECB, the establishment
of negative deposit rates in the euro area, Switzerland, Sweden and Denmark, and more
conventional forms of easing in China, India, and many smaller economies.
What we find more striking about the 2014 bond market rally is the degree to which it
extended to the long end of real interest rate curves. In the US, the 5y5y forward TIPS rate has
fallen from nearly 2% at end-2013 (itself low by historical standards) to less than 0.4% in
January 2015 (Figure 3). The collapse in the UK 5y5y real rate is even more extreme, leaving it
at an unprecedented negative 0.7% from around 1% a year earlier and much higher in
previous years.
Moreover, the collapse in forward rates has not been limited to the 5-year point. Inflationlinked swap markets are now pricing strongly negative real rates beyond 10 years in the
euro area and the UK. 10-year forward real rates are positive for the US and Japan, but at
historically abnormally low levels (Figure 4).
These abnormally low forward rates likely reflect, in part, negative term-risk premia, as bond
duration in the US, UK, core Europe and Japan has established itself as a negative-beta,
safe-haven asset class. But it seems unlikely that the term premium is sufficiently negative
to generate an implied rate forecast anywhere near historically normal real interest rates.
One interpretation of current bond market pricing is that participants are expressing the
view that real interest rates are likely to be abnormally low for a very long time much
longer than it takes for transitory or, loosely-speaking, cyclical developments to play out. In
this article, we take issue with this view.
FIGURE 3
Forward real interest rates plummeted in 2014
5%
FIGURE 4
10-year forward rates suggest low real rates forever
1.0%
4%
0.5%
3%
0.0%
2%
1%
-0.5%
0%
-1%
-1.0%
-2%
Dec-96 Dec-99 Dec-02 Dec-05 Dec-08 Dec-11 Dec-14
-1.5%
US5x5
UK5x5
Note: 5y5y forward real interest rates computed from inflation-linked bond
market. Click here to view an interactive Barclays Live Chart Source:
Bloomberg, Barclays Research
24 February 2015
US
Euro area
UK
Japan
Note: 10-year forward real interest rates from inflationlinked swap market.
Source: Barclays Research
economic response to Aprils tax hike in Japan, a fading of the weak recovery that had seemed
to be in place in Europe, and growing evidence that Chinese demand was decelerating faster
than expected. Weak inflationary pressures likely contributed to the rally, but should probably
be viewed more as a reflection of the weak cyclical context than as an independent driver.
Disappointment in 2014 global
growth may have been a
catalyst for the bond market
rally
However, although growth disappointed, the bottom did not fall out of the world economy in
2014. From end-2013 to the present, for example, Barclays forecasts of 2014 and 2015 world
GDP growth have fallen by 0.3pp; it would surprise us if consensus forecasts fell much further.
This raises the question: Why would so modest a deceleration, which likely reflects cyclical
developments, at least in part, have affected investors assessment of the long-run outlook
enough to generate such a strong bond-market response? A reasonable answer, in our view,
is that the coincidence of sluggish output growth with robust labor market recoveries (in the
US and UK) or stable labor markets (as in Japan and China) led investors to buy into the view
that sluggish output growth in the recent economic recovery reflected a weak secular
outlook, attributable to some combination of demographic and productivity-related factors.
For what its worth, we have a lot of sympathy with the view that trend growth has slowed
significantly in most systemically important economies, a view that is laid out in convincing
detail in Gapen (2015).
If a downgrade of market participants assessment of the secular outlook for growth was
indeed a market theme in 2014, the relevant question would seem to be what sort of
downgrade seems plausible and how large the impact on asset prices might be in the long
run. We think this is not quite the right question, and are not going to address it here.
It is true that some basic economic theory provides reason to believe that the economic
growth rate and the real interest rate are positively related, although the strength of the
theoretical relationship is sensitive, in particular, to assumptions about how savings are
determined. But the same theories suggest that, even if we abstract from cyclical and focus
on secular drivers of interest rates, as we wish to do here, interest rates and asset prices are
also influenced by many other factors.
Experience suggests that these other factors loom large in practice and, as a result, that the
empirical relationship between economic growth and real interest rates is not strong. In two
recent analyses of US economic history, for example, Bosworth (2014) found a weak link
between economic growth and real interest rates, while Hansen and Seshadri (2013) found a
negative long-run relationship.1 We suspect that the intuitive presumption of a strong link
between trend growth and the real interest rate is at least partly due to a failure to distinguish
completely between cyclical, mainly demand-related, fluctuations in the rate of growth and
secular variations, which are longer-lasting and driven predominantly by the economys
capacity to supply output. A boom in demand will naturally elicit a rise in the rate of interest; it
is far less clear that rapid trend growth in supply capacity has the same implication.
International comparisons also provide weak evidence, at best, that variations in trend
growth are a powerful driver of the equilibrium real interest rate. Interest rates have (for
example) been consistently high in Brazil, and low in China and Korea, even though trend
growth has been substantially higher in China and Korea. The key difference, in our view, is
that in Brazil, domestic saving is very low compared with underlying investment demand; in
China and Korea, the opposite is true.
Thus, we think a more promising approach to understanding the outlook for real interest rates
(and asset prices more broadly) is to focus on the drivers of world saving and investment. (Of
course, the trend rate of growth can be introduced into this framework as one driver of saving,
investment, and asset prices.) This is conventional; for example, the IMF recently adopted a
broadly similar framework.2 Our analysis differs from the IMFs in its more concentrated focus
1
Barry Bosworth Interest Rates and Economic Growth: Are They Related?, 2014, Brookings Institution, and Bruce
Hansen and Ananth Seshadri Uncovering the Relationship Between Real Interest Rates and Economic Growth,
2013, University of Michigan.
2
Perspectives on Global Real Interest Rates, in IMF World Economic Outlook, April 2014.
24 February 2015
on the systemically significant economies of the world and on what we think is a particularly
powerful driver of the recent and prospective savings/investment balance: the evolution of
population structures globally. We think this more sharply focused discussion is warranted
by the powerful and, we suspect, still under-appreciated, influence that population
dynamics have had and may have on financial markets in the decades ahead.
For 30 years, an increasingly
supportive demographic
context has combined with
other drivers to deliver everlower real interest rates.
This is not, of course, to suggest that demographic trends are the only relevant drivers of
interest rates and asset prices. In the immediate future, the weak cyclical backdrop
associated with the rebalancing in China, de-leveraging and reflation in the euro area, and
still incomplete recovery from the 2008-09 financial collapse, will continue to exert a
powerful influence over monetary policy and real interest rates around the world. Some
longer lasting, more secular drivers also point toward low interest rates, at least on highquality, liquid, safe haven assets. Indeed, we have addressed some of these drivers in
recent editions of the Equity Gilt Study. (On the so-called safe asset shortage, see for
example Gavin, Ghezzi, Brown and Gregory (2012) and Gapen (2013).)
But during the past 30 years, powerful demographic trends have combined with these other
drivers, providing steadily increasing support for asset markets and downward pressure on
interest rates. The reversal of this demographic support for saving and, by extension, asset
markets is at hand. Although it will be slow-acting, we believe that the ebbing demographic
tide will transform the investment landscape as powerfully as the global savings glut
shaped the landscape of recent decades.
2000
China
Other emerging
2005
2010
US
Note: Real investment as share of world real GDP. Source: Barclays Research
3
We begin with real investment as published by the national statistical office, for example, bn 2005 JPY for Japan. We
then transform this series into 2014 local-currency prices using the deflator for investment spending (except in China,
where this is not published and we use the GDP deflator). This is a level adjustment only, and leaves growth rates
unchanged. We then transform 2014 local currency data into USD using the 2014 USD exchange rate.
24 February 2015
just over 24% in 2014. During the same period, real investment in China has increased from
roughly 2.5% of world real GDP to almost 9%. In absolute terms (ie, constant 2014 USD),
the two-decade expansion of real investment in China accounts for more than 60% of the
growth in world investment during the same period.
It might once have been reasonable to confine an analysis of global savings and investment
trends to the advanced economies, but whether one is seeking an explanation for recent
developments or to make assessments of the future, this is no longer at all possible.
As investment in China has boomed, investment in the rest of the world has declined as a
share of world GDP (and, in Japan, in absolute terms). This decline has been concentrated in
Europe and Japan, where investment rates have fallen and economies now account for a
smaller share of the world economy. Indian investment has grown rapidly, but from such a
small base that it still comprises only about 1% of world GDP, substantially smaller (for
example) than Chinas in 1995.
China has played an even more
central role in the global
saving glut
FIGURE 6
Real investment has risen strongly in the past two decades, led by China
2014 USD
1995
2000
2005
2010
2014
US
1,628
2,494
2,806
2,226
2,802
China
755
1,000
1,875
3,691
4,929
Euro area
2,221
2,717
2,825
2,743
2,579
Japan
1,172
1,121
1,069
918
1,007
UK
561
542
505
477
525
India
na
174
295
494
577
Korea
255
277
336
387
408
Brazil
212
225
231
371
371
Mexico
106
185
211
234
263
Russia
272
201
293
352
390
TOTAL
7,182
8,935
10,445
11,894
13,851
1995
2000
2005
2010
2014
US
5.0%
6.3%
6.1%
4.3%
4.9%
China
2.3%
2.5%
4.1%
7.1%
8.6%
Euro area
6.8%
6.9%
6.2%
5.3%
4.5%
Japan
3.6%
2.8%
2.3%
1.8%
1.8%
UK
1.7%
1.4%
1.1%
0.9%
0.9%
India
na
0.4%
0.6%
1.0%
1.0%
Korea
0.8%
0.7%
0.7%
0.7%
0.7%
Brazil
0.6%
0.6%
0.5%
0.7%
0.6%
Mexico
0.3%
0.5%
0.5%
0.5%
0.5%
Russia
0.8%
0.5%
0.6%
0.7%
0.7%
TOTAL
22.0%
22.6%
22.8%
23.0%
24.2%
24 February 2015
FIGURE 7
Chinas contribution to global saving has been even more significant
0.60
0.50
0.40
0.30
0.20
0.10
0.00
1995
2000
2005
World
China
2010
Ex-China
Note: Ratio of regional saving to regional GDP. In all cases the rate of national saving is measured as the rate of
investment plus the current account surplus. Source: Barclays Research
There is much that conventional models of national saving do not fully explain. But one
prediction of basic life-cycle theories of consumption is generally borne out by the data. This
is that individual saving rates tend to be hump-shaped, rising from a fairly low level in young
adulthood to a peak in late adulthood, then declining after retirement. This pattern creates a
presumption that national savings rates may be correlated with the national age structure.
To explore this idea, we need to quantify the demographic structure of a population. The
age composition of a population is a distribution, not a number, and for purposes of
discussion or statistical analysis the distribution needs to be condensed into a reasonably
compact set of numbers. This involves compromises and arguably a degree of oversimplification. Here is how we approached the problem:
The interplay between the high saving of mature workers and the lower saving of the
elderly population is the central theme of this article, so we start with measures of the
share of mature (presumably high-saving) workers and elderly (presumably lowersaving) people. It is conventional to identify the elderly with the share of the population
aged 65 and older. We adopt this conventional measure, which seems sensible and
allows us to compare our results with the many other studies that include the ratio in
their analysis. But we should bear in mind that the measure may be culture-bound (since
not every society has a conventional retirement age of about 65) and obscures
potentially important variations of saving rates within the population that we categorize
as elderly.4
We identify mature workers as those aged 40-64, roughly corresponding to the second half
of a typical individuals working life. This division is equally arbitrary and also obscures likely
differences in saving behaviour within the categories young worker and mature worker.
Thus, these ratios should be viewed as indicators of broad trends in population dynamics, not
precise measures that capture every relevant detail of the demographic structure.
Having simplified a complex reality by focusing on these two ratios, it will be useful in much
of the ensuing discussion (eg, in most of the graphical analysis that follows) to simplify it
even further with a single number that summarizes the overall effect of the two interacting
demographic terms. We do this as follows:
4
For example, Poterba, Venti and Wise (2011) find that retired households tend not to draw down their housing or
other assets to support consumption in the early years of retirement, and seem to do so in response to health-related
shocks later in retirement.
24 February 2015
The underlying idea is that the saving rate of economy i in period t is something like:
= + +
Here is the share of mature workers and is the share of the elderly in the
population.
This bivariate approach is fine for statistical analysis of the sort that we report in the
appendix. One obvious way to estimate the overall effect of the population structure on
saving propensities would be to estimate the parameters b and c (or use estimates from
existing studies, if available) and compute a synthetic measure: = . The
problem with this is that the proposed measure of demographic pressure is too dependent
on a necessarily incomplete and fallible empirical study.
For graphical analysis, we use
the difference between the
shares of mature workers and
of the elderly in the total
population.
If the coefficient b is not much different from c, then the difference between the share of
mature workers and of the elderly is a natural measure of overall demographic pressure on
saving: = ( ). There is no theoretical reason to believe that this should be the
case, but in our statistical analysis we find that it is a very reasonable approximation. So, in
what follows, we take a two-track approach. In the scatter charts with which we illustrate
relationships and in order to structure the subsequent discussion of historical and
prospective trends, we use the difference between the shares of mature workers and of the
elderly as a measure of overall demographic pressure on saving. In the more formal
statistical work that is reported in the appendix, we analyze the separate influences of the
two ratios.
There are many studies of the determinants of national saving and some of these introduce
demographic factors as a potential driver. (A number of these studies are discussed in the
appendix to this chapter. The old-age dependency ratio is included far more often than a
measure of high-saving, mature workers.) But there is no consensus on the strength of the
link between age structure and saving. Thus, we begin with a look at 10 systemically
important economies that comprise the lions share of global output, saving, and investment.
The economies that we included in our analysis are the US, the euro area, China, Japan, the UK,
India, Korea, Brazil, Mexico, and Russia. We generally split our 20-year sample (1995-2014)
into four five-year subsamples to smooth out some high-frequency fluctuations and, perhaps,
measurement errors, that may contaminate annual data. Not all of the countries provide
complete estimates of the income side of the national income accounts, so we have estimated
national saving as the sum of domestic investment and the current account balance.
In Figure 8, we plot for each of these four periods and 10 economies the rate of saving
against our measure of demographic pressure on saving. There is clearly much about
national saving that is not explained by this measure of demographic pressure, but there is
also a strong, positive relationship, as the simple theory would suggest.
In fact, the size of the simple co-movement between demographics and saving illustrated
by Figure 7 is suspiciously large.5 This turns out to be caused by a very strong correlation
between demography and saving across our 10 economies. (Figure 9) We view the
magnitude of this cross-economy co-movement with some skepticism. The long-run crosscountry correlation is driven largely by China and Korea (the highest-saving observations in
the upper right portion of Figure 9), where savings have likely been high during the past 20
years for other reasons in addition to demographics. With such a small number of
economies in the analysis, the possibility of at least partly spurious, or accidental,
correlation is strong.
In theory, the slope of the trend line should be something like the difference between the saving rates of mature
workers and of the elderly.
24 February 2015
10
FIGURE 8
Demographic pressures are correlated with national saving
rates
FIGURE 9
20-year averages show an exaggerated (and likely spurious)
co-movement
S/GDP
S/GDP
0.60
0.50
0.45
0.50
0.40
0.40
0.35
0.30
0.30
0.25
0.20
0.20
0.10
y = 2.5663x - 0.2199
R = 0.3983
0.15
0.10
0.00
0.12
0.14
0.16
0.18
0.2
0.22
0.24
0.26
0.14
0.16
0.18
0.20
0.22
0.24
One way to reduce this problem is to subtract the long-run average from all data from each
of the five-year sub-periods. This removes persistent cross-country differences from the
analysis, which is then driven by variation over time within each economy.6
We find a strong co-movement
between age structure and
saving, although other drivers
are also important
Figure 10 illustrates that the co-movement between demographics and saving remains
strong after removing these economy-specific fixed effects from the data. (For readers
interested in a more formal quantitative analysis, the statistical results that correspond to
this illustrative chart are given in columns 2 and 3 of Figure 28 in the appendix.) The
numerical analysis suggests that a one percentage point increase in our measure of
demographic pressure is associated with an increase in national saving of about 0.7pp. This
is plausible, in light of the theory, and broadly consistent with our reading of the evidence
from other studies. The multipliers are larger if we estimate the effects of mature workers
and elderly people separately, and the negative impact of a rising share of elderly appears
somewhat larger than the positive impact of mature workers. But the difference between
the constrained and the unconstrained estimates is not statistically significant, nor does it
alter the qualitative discussion of the outlook, below.
If demographic pressures affect national saving, they should also affect the current account.
Figure 11 shows that the current account has in fact been positively correlated with our
measure of demographic pressure on saving (see also columns 4 and 5 of Figure 28 in the
appendix). The evidence is a little more tenuous here, which is not surprising given the
importance of shocks to the terms of trade, economic policy, and other sources of highfrequency fluctuations, as well as the fact that the current account depends not only on
developments at home, but also in key trading partners. Still, the positive association
between demographic pressure and the current account is congruent with the view that
demographic pressures exert upward pressure on saving.
We have so far focused on the level of the saving rate, either in absolute terms or relative to
the economys long run (1995-2014) average. As a test of robustness, we can also explore
whether changes in demographic factors explain changes in the national or regional saving
rate. This is relevant for the forward-looking discussion that follows because we are
ultimately interested in understanding potential changes in saving behaviour that may be
implied by forecast changes in demographic structure in the coming 10 or 20 years. In
6
In the statistical work described in the appendix, we do this with a fixed-effects estimator. In the scatter plot in
Figure 10 we subtract the country-specific 1995-2014 averages for saving and demographic pressure from each
observation.
24 February 2015
11
FIGURE 10
Co-movement of saving and demographics is also strong
after removing economy-specific fixed effects
CAS/GDP
0.06
0.10
0.04
0.08
0.02
0.06
0.00
0.04
0.02
-0.02
0.00
-0.02
-0.04
-0.04
-0.06
-0.06
0.12
-0.08
-0.06
-0.04
FIGURE 11
Demographic pressure on saving is also associated with a
current account surplus
-0.02
0.02
0.04
0.14
0.06
0.16
0.18
0.2
0.22
0.24
0.26
Figure 12, we show the relationship between the change from 1995 to 2014 in our measure
of demographic pressure and the change in national saving over the same period.
Demographic forces help
explain the large decline in
Japanese saving and the boom
in Chinese saving of the past
20 years.
One drawback of viewing the data this way is that we are left with only nine observations,7
which makes the correlation potentially sensitive to outliers. But for what it is worth, the
resulting correlation is consistent in magnitude with the previous, arguably more robust
analysis.8 It also highlights that demographic pressures go a long way toward explaining
the change in saving in the country with the largest decline (Japan, in the lower left of Figure
12) and the one with the largest increase (China, in the upper right).9
FIGURE 12
Long-term changes in demographic structure have been correlated with changes in
national saving rates
Change in
national saving 1995 - 2014
0.10
0.05
0.00
-0.12
-0.08
-0.04
0.00
0.04
0.08
0.12
-0.05
-0.10
-0.15
Change in demographic pressure - 1995-2014
Source: Barclays Research
It probably comes as no surprise to most readers that Japan has been the most
demographically challenged of the major economies during the past two decades. We have
the impression that the exceptionally strong demographic backdrop in China has figured
7
We have no measure of saving for Russia in the early years of our sample, so Russia is not included in Figure 12.
The co-movement in Figure 12 suggests that a 1pp increase in our measure of demographic pressure would be
associated with a 0.73pp increase in savings demand, similar to the results of the preceding analysis.
9
We note that the observed relationship is not driven by these two observations, which remains broadly similar if the
China and Japan are excluded from the sample. However, as observations are dropped from an already small sample,
the uncertainty surrounded the estimated co-movement naturally rises. In any event, while it is reassuring that the comovement is not overly sensitive to their inclusion, we no strong reason to exclude China and Japan from the sample.
8
24 February 2015
12
less prominently in academic and market analyses of Chinas recent history. In our view,
prospective demographic developments in China are likely to have a profound effect on
global savings and world asset markets, as we discuss below.
Our data seem to be consistent with the idea that savings tend to be positively correlated with
a demographic structure concentrated in mature workers, and negatively correlated with a
high proportion of the elderly. We take a closer look at recent historical and prospective
developments in the world as a whole and within the systemically most important countries
and regions. The discussion highlights the role of population trends in the development of a
global savings glut in recent decades, and suggests that demographic pressures on saving
(and, by extension, on asset markets) have peaked and are in the process of a potentially
momentous reversal.
Both China and the rest of the world experienced strong upward demographic pressure on
saving after the early 1980s. But the Chinese experience stands out for the magnitude of the
swing, which, at more than 10pp, is more than twice the change experienced by the rest of
the world. The demographic swing in China is also more recent than in the rest of the world,
dating from the early 1990s, compared with the early 1980s elsewhere.
To put this swing in a rough quantitative context, we can attach the multiplier of about 0.7
suggested by the statistical work summarized above to the roughly 10pp increase in the
measure of demographic pressure from 1995 to 2014, which suggests that roughly 7pp of
the 10pp swing in the measured rate of Chinese saving may plausibly be attributed to the
marked shift in the demographic context.
This suggests a perspective on the post-2000 investment boom in China that emphasizes a
demographically induced surge in saving as the primary driver of the episode, rather than
policy distortions or other drivers of investment demand. A saving-driven interpretation is
supported by the contemporaneous surge in the current account surplus.
FIGURE 13
Demographic pressures help explain the Chinese saving
boom, but are poised to fade in the decades to come
FIGURE 14
The rest of the world also experienced upward demographic
pressure on saving since 1980, though smaller than Chinas
0.40
0.40
0.35
0.35
0.30
0.30
0.25
0.25
0.20
0.20
0.15
0.15
0.10
0.10
0.05
0.05
0.00
0.00
1950 1960 1970 1980 1990 2000 2010 2020 2030
Projected
24 February 2015
Mature
Elderly
Difference
Mature
Elderly
Difference
Note: Average for US, Euro area, Japan, UK, India, Korea, Brazil, Mexico, and
Russia, weighted by 2014 USD GDP. Source: Barclays Research
13
Looking ahead, Chinas pro-saving demographic evolution is set to reverse, starting about
now, because of a projected acceleration in the share of the elderly population, combined with
a sharp deceleration of growth in the share of the mature workforce.
The rest of the world is in the midst of a similar, but more complete, reversal of the
demographic trend of the past two decades. Whereas Chinas demographics in the next two
decades are projected to remain more saving-supportive than in the 1980s and early 1990s,
in the rest of the world the demographic structure is projected to shift to a substantially
more saving-unfriendly composition than at any time in its post-war history. What this
suggests to us is that China is likely to remain a net supplier of world savings in the years to
come, but in a world where savings overall are becoming ever scarcer.
Although much smaller and systemically less significant than China, the other countries of
emerging Asia that we consider here provide an interesting contrast. Korea is one of the
fastest-aging societies in the world, and its age structure is projected to turn very rapidly
from one of the most pro-saving in the world, to one of the least. Like China, upward
demographic pressure on saving has been building rapidly in recent decades, although the
surge has been less abrupt, dating from the 1970s rather than the 1990s (Figure 15). By
2035, however, the share of the elderly population is projected to rise from a moderate level
by international standards to a level even higher than Japans, and only marginally below that
of the other demographic pioneer, Germany. At the same time, the share of the population in
the high-saving middle-age years is projected to decline sharply. Our summary measure of
demographic pressure on saving is thus projected to drop from one of the highest in our
sample of 10 economies (essentially tied with China in 2015) to one of the lowest (essentially
tied with the UK and US, and only marginally more savings-friendly than Japan, by 2035).
This provides a useful context for Koreas very large recent current account surpluses,
suggesting that they may be associated with demographic pressures created by a rapidly
aging population that is saving to provide for its imminent retirement.10 If so, upward
pressure on saving and the current account balance is likely to plummet in the years ahead
as the generation now saving for its retirement enters its lower-saving elderly stage of life.
India provides a stark contrast and, of the countries that we discuss here, is the only
exception to the global demographic norm. As in the rest of the world, the share of Indias
elderly population is rising (from a very low level Figure 16). But a relatively high birth rate
ensures that a large cohort of children and young people will be entering their high-saving
mature working years for decades to come. The net result is that Indias demographic
structure uniquely among the 10 economies that we discuss will become gradually more
savings-friendly. In fact, by 2035, Indias demographic structure is projected to be more
saving-friendly than any other region in our sample, including China. This suggests that
India may be transformed into a net supplier of global savings in the decades to come, and
that the current account deficits of the past may be replaced by structural surpluses.
However, from the perspective of the global savings glut and world asset markets, Indias
savings-friendly demographic backdrop is unlikely to provide a major offset to the
unfavourable demographic trends in most of the world. In 2014, Indian investment
comprised roughly 1% of world GDP (and Indian saving even less), compared with Chinas
2.5% in 1995 and nearly 9% in 2014. Therefore, it is likely to be a long time before Indias
saving is large enough to provide a meaningful offset to developments in China and the
advanced economies.
10
Koreas trading partners have been aging rapidly as well. Our point is that the associated demographic pressure on
saving has been larger in Korea than in any other large economy, save China.
24 February 2015
14
FIGURE 15
Korean demographics are poised for a very sharp reversal
FIGURE 16
India is a rare exception to the global demographic norm
0.40
0.40
0.35
0.35
0.30
0.30
0.25
0.25
0.20
0.20
0.15
0.15
0.10
0.10
0.05
0.05
0.00
0.00
Mature
Elderly
Projected
Difference
Mature
Elderly
Difference
Emerging Asia looms very large in the 21st century saving/investment balance. But the US,
UK, and the euro area still comprise the majority of the worlds output, saving, and investment.
What happens in these economies still matters, and will continue to matter for decades.
Their demographic patterns are broadly similar. In all three advanced-economy regions, the
share of the elderly is projected to rise more rapidly, while that of the mature workforce has
either started to fall (US, UK) or is projected to do so soon (euro area).
FIGURE 17
US demographics have only recently turned less savings-friendly
0.40
0.35
0.30
0.25
0.20
0.15
0.10
0.05
0.00
1950
1960
1970
Projected
1980
1990
Mature
2000
2010
Elderly
2020
2030
Difference
As a result, after a 25-year surge in demographic support for saving, population structures
are becoming strongly less saving-friendly in all three areas. Within two decades, the age
structure is projected to become markedly less savings-friendly than it was in 1980.
With saving rates as low as they are in the US and UK, it is not easy to imagine a large,
sustained decline from current levels. Then again, we have limited experience with
demographic structures like the ones that face the advanced economies, and it may be that
we will simply have to adjust our idea of whats normal.
24 February 2015
15
FIGURE 18
Euro area population dynamics are similar to the US
FIGURE 19
UK demographics are also projected to turn sharply less
saving-friendly in the next two decades
0.40
0.40
0.35
0.35
0.30
0.30
0.25
0.25
0.20
0.20
0.15
0.15
0.10
0.10
0.05
0.05
0.00
0.00
Mature
Elderly
Projected
Difference
Mature
Elderly
Difference
Japan is sometimes cited as evidence for the view that demographics affect market
valuations, because the post-war financial market boom that reached its climax at the end
of the 1980s coincided with a boom in demographic support for saving (Figure 20). Both
market valuations and demographic support for saving declined sharply after the early
1990s. One need not buy the idea that demographic forces fully explain the bubble
economy of the 1980s and Japanese markets subsequent, protracted, correction, to accept
the idea that pro-saving demographic factors may have set the stage for the high market
valuations of the late 1980s, and provided an impetus for normalization thereafter.
FIGURE 20
Japanese demographics supported a saving boom into the
early 1990s, and reversed sharply thereafter
FIGURE 21
In Germany, the demographic structure is only now
becoming less supportive of saving
0.40
0.40
0.35
0.35
0.30
0.30
0.25
0.25
0.20
0.20
0.15
0.15
0.10
0.10
0.05
0.05
0.00
0.00
1950 1960 1970 1980 1990 2000 2010 2020 2030
Projected
Source: Barclays Research
24 February 2015
Mature
Elderly
Difference
Mature
Elderly
Difference
16
Germany fits less neatly into the paradigm that we have been discussing in this note. The
demographic factors that we have highlighted explain neither the increase in the German rate
of national saving since the early 2000s, nor the (even larger) decline in the rate of investment.
Our measure of demographic pressure on saving has become marginally less saving-friendly
during this period (Figure 21). Germany thus illustrates that factors other than demographics
do drive saving, and that, when changes in the demographic driver are relatively small, other
factors will predominate. Whether German savings can remain resilient to the very sharp
decline in demographic support projected for the coming 20 years is an interesting question
with non-negligible implications for the European and world economy.
In the coming 20 years, our proposed summary indicator of global demographic support for
saving is projected to decline by about 8 pp. Our statistical results suggest that this could be
associated with a decline in desired saving (at any given interest rate, which is to say a
leftward shift of the saving supply schedule) of nearly 6 pp of world GDP, or about 25% of
world saving. Of course, we are unlikely to see world saving and investment fall by the full
25%. The effect of this leftward shift of the saving supply schedule on actual
saving/investment and the real interest rate will depend upon the slopes of the investment
demand and the saving supply curves, among other things. With no strong view on the
magnitude of these slopes, we are not in a position to provide an estimate of the impact on
asset prices and interest rates. Suffice it to say that this would be a very large shock to the
balance between saving and investment if it were half the size. It compares, for example,
with an increase of about 5 pp in demographic pressure during the 1980-2015 period of
strong secular support for asset markets, and downward pressure on interest rates, which
reflected an increase of nearly 12 pp in China and about 3.5 pp in the rest of the world.
We have suggested that demographic factors have been a key driver of the global savings glut
of the past 20 years. Intuition and economic theory suggest that a demographically induced
bulge in saving should be associated with an increase in asset prices (and a corresponding
decline in the real interest rate).11 When societys need to save is high, the price of saving
vehicles will be bid up and the expected returns to saving will be depressed. It is tempting to
explore whether the demographic factors that have been shown to be associated with high
world saving are also associated with high asset prices. In this section, we succumb to this
temptation and consider a measure of the real interest rate and equity valuations.
With highly integrated capital markets, and over the extended periods that concern us here,
real interest rates and other asset prices should be equalized to a very substantial degree.
We are therefore led to focus on a measure of the world interest rate as the appropriate
object of analysis. Figure 22 shows a measure of the real short-term interest rate in the US,
UK, Germany and Japan. There are occasionally very sharp divergences among them, but
longer-term trends appear highly correlated. (There is also almost certainly a large element
of measurement error related to the high and volatile inflation rates of the 1970s and early
1980s.) In what follows, our measure of the world interest rate is the simple average of the
real interest rates depicted in Figure 22.
11
One fully articulated theoretical model of a demographic cycle is in Geanakoplos, Magill, and Quinzii (2004), which
also provides evidence that US stock valuations have been positively correlated with the ratio of high-saving mature
workers. Poterba (2004) also documents that real US interest rates tend to be depressed when the share of the
population aged 40-64 is high, while a high share of the elderly is associated with a rise in real interest rates, although
he characterizes the correlation as weak. Bond (2009) also suggested that demographic trends would put upward
pressure on real interest rates and downward pressure on equity valuations in the US and the UK.
24 February 2015
17
FIGURE 22
A measure of the short-term real interest rate
FIGURE 23
Long-term trends in world interest rates and demography
10%
5%
0%
8%
0.14
6%
0.15
4%
0.16
2%
0.17
0%
-5%
0.18
-2%
-10%
0.19
-4%
-15%
-20%
1960 1966 1973 1980 1987 1993 2000 2007 2014
US
UK
DE
-6%
0.20
-8%
0.21
-10%
1960
JP
0.22
1970
1980
1990
2000
2010
Demographic pressure
By the same token, asset prices in the financially integrated regions of the world should be
influenced by global demographic developments, rather than national. This means that
there is little to be learned from cross-country variations in real interest rates and
demographics, or other national economic drivers. We have little choice but to evaluate the
historical co-movement of the world real interest rate and global demographic trends. With
only a few decades of postwar experience available for study, the available history provides
us with quite limited variation in the slow-moving demographic driver; we must therefore
interpret historical co-movements with some caution.
There is a strong historical comovement between the global
age structure and the real
interest rate.
Figure 23 shows how our measure of the real interest rate and global demographic pressure
on saving have evolved since 1950, and highlights the fact that history provides us with only
one long, slow, instance of deterioration in demographic pressure on saving (1950-early
1980s) followed by a long, slow increase in demographic pressure on saving, which is only
very recently beginning to reverse. This reinforces the case for a cautious interpretation of
the statistical correlations.
Caveats aside, the fact is that the historical co-movement has been strong. This is more
clearly seen in Figure 24, which plots annual versions of the demographic data and real
interest rates shown in Figure 23 against one another. There is a lot of noise around the
FIGURE 24
The real interest rate has been negatively correlated with
demographic pressure on saving (annual data)
Real rate
FIGURE 25
Real interest rate and saving (smoothed data)
Real rate
8%
5%
6%
4%
3%
4%
2%
2%
1%
0%
0%
-2%
-1%
-4%
-2%
-6%
14%
16%
18%
20%
24 February 2015
22%
-3%
0.140
0.160
0.180
0.200
0.220
18
trend line, reflecting other short-term drivers of real interest rates and almost certainly a lot
of measurement error in the high-inflation era, but the relationship is statistically and
economically significant; the demographic pressures that we have found to promote saving
have also been associated with lower real interest rates.
For what it is worth, the statistical evidence presented in the appendix (Figure 29) suggests
that a 1pp point increase in the share of mature workers has been associated with a 0.75pp
decline in the world real interest rate. A 1pp increase in the share of the elderly has been
associated with a 1.15pp increase in the real interest rate. (The estimated impact of shift in
the age structure and the real interest rate is considerably smaller if the separate effects are
constrained to be equal and opposite in sign, as is implicitly done in Figure 24.)
Annual data are contaminated by short-term fluctuations that have nothing to do with secular
trends and by potentially large errors in the measurement of expected inflation. In Figure 25
we have tried to reduce these problems by sorting the data on demography, then averaging
over groups of five annual observations apiece. Smoothed in this way, history suggests a
negative relationship between the real interest rate and age structure that is quantitatively
similar to but less noisy than the one in Figure 24. The relationship does not seem to be
driven by a single outlier or cluster of outliers; if anything, the outlier seems to be attributable
to the monetary disorder of the late 1970s, when demographics were unfavourable yet our
measure of the world real interest rate was very low. We think it makes more sense to view
this outlier as the result of monetary policy mistakes and measurement error than as a
counter-example to the generally negative co-movement between of demographic pressure
and the real interest rate. Eliminating this observation would considerably strengthen the
observed historical co-movement between real interest rates and age structure.
The same theory that suggests that demographic pressure on savings should reduce the real
interest rate also suggests that it should support equity valuations. This is because increased
demand for saving vehicles should push all asset prices up (and expected returns down), with
the additional possibility that an aging population may shift its asset allocation in the direction
of less volatile, safe haven fixed-income assets, resulting in a higher equity risk premium.12
Here, we focus on the US cyclically adjusted PE (CAPE) ratio as one plausible and easily
computed valuation metric. Partly to minimize (although, as we shall see, not eliminate)
problems created by the exaggerated level of equity valuations in the late 1990s bubble, we
analyze the cyclically-adjusted earnings ratio, which is simply the inverse of the CAPE.
Figures 26 and 27 illustrate that, as a purely statistical matter, equity valuations have been
rather strongly correlated with our measure of global demographic pressure on saving. (The
corresponding statistical analysis is given in Figure 29 of the appendix.) A literal reading of
Figure 27 suggests that this relationship could be non-linear; indeed, a nonlinear relationship
does a much better job of explaining the data. However, we suspect that conclusions like this
would be pushing the analysis beyond what the data can support. In this analysis we rely
entirely on a relatively brief (in demographic time) time series of information, during which
two events dominate the valuation experience: the monetary disorder of the late 1970s and its
subsequent correction (when equity valuations were exceptionally low and demographic
support for saving happened to be rather weak), and the equity market bubble of the late
1990s and early 2000s (when demographic support for saving happened to be strong).
This does not mean that demographic pressures have not contributed to equity valuations in
recent decades, but they were clearly not the only influences at work. Although it fits neatly with
our view that demography has exerted a powerfully supportive influence on the investment
climate in recent decades, we would not take the observed historical co-movement between age
composition and equity valuations as a strong guide to the future until the impacts of these
other factors been more convincingly controlled for than we have been able to do here.
12
24 February 2015
19
FIGURE 26
US equity valuations have been correlated with global
demographic pressures on saving
FIGURE 27
Strong co-movement between equity valuations and
demographic pressure on saving
14%
0.14
1/US CAPE
12%
0.15
14%
0.16
12%
0.17
10%
0.18
8%
0.19
6%
0.2
4%
2%
0.21
2%
0%
1960
0.22
0%
14%
10%
8%
6%
4%
1970
1/US CAPE
1980
1990
2000
2010
16%
Demographic pressure
18%
20%
22%
Demographic factor
Source: Barclays Research
Despite these limitations of the analysis, it remains noteworthy that, in the post-war period,
asset prices seem generally to have had the association with demographic trends that would
be expected if demographic pressure on saving were a key secular driver of asset markets.
Conclusions
Demographics are not the only drivers of world savings, investment, and asset prices, but
they seem to us to be among the most powerful. Moreover, we are living through a
demographic inflection point with potentially profound implications for asset markets
implications that have been overshadowed by the existing demographic structure and the
weak cyclical context, both of which have contributed to abnormally low real interest rates.
We think it is a mistake for market participants to extrapolate current circumstances into
the distant future to the extent that they seem to have done.
Demographic fundamentals have become highly supportive of world savings, and by extension
asset prices, since 1980, particularly in the past 20 years. This is because the impact of a steady
rise in the share of the elderly has been more than offset by a rise in the share of mature
workers who save a lot. This has been a global phenomenon, with Japan the only significant
exception, and has been particularly powerful in China, bedrock of the global savings glut.
But demographic support for global saving is peaking, and it will be getting steadily and
substantially less supportive in the decades ahead. When this happened to Japan in the early
1990s, the Japanese saving rate fell, as expected, although asset prices in Japan and globally
continued to be supported by a surge in saving in the rest of the world. It seems likely to us
that, as demographic support for savings recedes in the US, Europe, UK, China, and Korea, the
global savings glut will similarly be reversed. Although the demographic tide will ebb gradually,
the impact on financial markets could be very large. Our statistical analysis suggests that the
decline in demographic support for saving could shift the global saving supply schedule back
by almost 3% of global GDP (more than 15% of world saving) in 10 years, and nearly 6% of
world GDP (roughly 25% of world saving) in 20 years. This would be a substantial dislocation
of the balance between world saving and investment if it were half the size.
The fading of the global savings glut seems very likely to put upward pressure on interest
rates and downward pressure on asset prices around the world. Although we think that they
should not be taken as strong guides to the impact of future, our statistical analysis of the
historical co-movement between demographic pressures and asset prices corroborates this
view, which seems quite inconsistent with market pricing of very low or negative 5-year and
even 10-year forward interest rates.
24 February 2015
20
Appendix
This appendix discusses the underlying conceptual framework, and numerical results of the
statistical analysis that is illustrated graphically in charts, above. We also provide some
background on related studies. The literature is enormous, and we cannot possibly provide
a comprehensive survey in the space available. We have attempted to put our discussion
within the context of existing studies that we consider representative.
13
The most recent and comprehensive analysis of global saving of which we are aware (Grigoli et al, 2014) finds a
modest positive relationship between the rate of saving and the real interest rate. Many empirical studies have found no
relationship between real interest rates and saving. For our purposes, the sensitivity of saving to the interest rate is not
important, unless saving is so negatively related to the interest rate that the saving function is negatively sloped and flatter
than the investment function. This leads to paradoxical results that seem implausible to us, so we leave it aside.
24 February 2015
21
The theory also offers some guidance on the question whether countries like Brazil, India,
and China should be considered part of the world for purposes of this analysis. Those
countries maintain sizeable barriers to and regulation of portfolio capital flows, which
prevent full equalization of expected returns on otherwise similar financial securities, as we
would expect among the advanced financial markets. However, all of these countries
experience large fluctuations in their current account balances. If a shock to saving or
investment in one of them affects their own current account, the current account of the
rest of the world must of necessity be affected equally and in the opposite direction. This
will require an adjustment of the world interest rate (and, in the background, real exchange
rates), even if there is no direct arbitrage between national financial markets.
For recent examples of this type of analysis, see Grigoli et al (2014) and Furceri et al (2014). Grigoli et al provide a
summary of 16 previous panel studies of national saving behavior.
24 February 2015
22
Statistical results
All statistical results reported here are based upon a sample of data that includes the US,
China, the euro area, Japan, the UK, India, Korea, Brazil, Mexico and Russia. We compiled
data for 1995-2014, and computed five-year averages of the data to reduce the influence of
high-frequency cyclical influences and measurement error. Data for 2014 were estimated
using the four quarters through 2014:Q3 in every case except China, where we used
Barclays estimates. Savings were estimated as the sum of the broadest available definition
of investment and the current account surplus. Saving, investment, and the current account
were all normalized by GDP. Missing data for Russia in 1995-2005 reduced our sample from
40 to 38 data points.
FIGURE 28
Co-movement of national saving and current account balances with national
demographic indicators
Dependent variable
Middle (40-64)
Old (65+)
(1)
(2)
(3)
(4)
(5)
Saving
Saving
Saving
CAS
CAS
1.90
0.684
0.41
(4.40)
(3.03)
(2.75)
-2.57
-0.962
-0.33
(-4.65)
(-2.17)
Middle - Old
Number of observations
(-1.73)
0.6681
0.41375
(2.99)
(2.92)
38
38
38
38
38
Adjusted R2
0.356
0.951
0.931
0.153
0.169
Fixed effects
No
Yes
Yes
No
No
Column 1 of Figure 28 shows the simple relationship between the demographic variables
that we emphasize and the rate of saving. (This is numerical analogue of the scatter plot in
Figure 8, except that the impact of mature worker and the elderly are not constrained to be
equal and opposite in sign, as they implicitly are in the scatter plot.) The estimated
coefficients are statistically significant at standard significance levels, and the signs are in
line with the theory. As we discuss in the text, the coefficients are larger than seem
plausible, in light of the theory. We think that this arises from unobserved country-specific
influences on saving that happen to be correlated with the demographic variables, biasing
the estimated coefficients.
To mitigate this problem, we introduce country fixed effects in column 2 and 3. These fixed
effects explain much of the variation in national saving rates, but the coefficients on the
demographic variables remain statistically significant and of the correct sign. In these
regressions, the values of the coefficients seem more plausible in light of the theory.
Column 3 is the numerical analogue of the scatter plot in Figure 10.
In columns 4 and 5 we explore the relationship between demographic variables and the
current account. The coefficients are of the expected sign and statistically significant,
although the demographic variables explain less of the variation in the current account than
of the domestic saving rate.
In our exploration of the observed historical co-movement between demography and asset
prices, we had first to compute a world real interest rate. For each of the US, UK, Germany
and Japan, we used the money market or T bill rate as reported by the IMF International
Financial Statistics. As our measure of expected inflation, we used the actual 12-month
forward rate of inflation. This almost certainly led to important measurement error,
particularly in the late 1970s when inflation accelerated rapidly and likely included a strong
unexpected component. The world real interest rate was defined as the simple average of
24 February 2015
23
the four national rates. Data are annual, and the statistical analysis includes the years 19602014. We note that the world demographic variables are constructed using 2014 GDP
weights. This grossly over-emphasizes the role of China for most of the period in question,
and it may be possible to improve the results by constructing an index whose weights vary
over time in line with the shifting roles of regional economies. This would, however, not
address the more basic limitations of the limited information in the time series that we
discuss in the text.
FIGURE 29
Co-movement of the real interest rate and equity valuations with world demographic
indicators
(1)
Dependent variable
Middle (40-64)
Old (65+)
(2)
-1.167
(-3.38)
1.147
1.423
(2.28)
Number of observations
(4)
(-2.53)
Middle - Old
Adj R2
(3)
(2.52)
-0.491
-1.000
(-2.07)
(-3.79)
55
55
55
55
0.1342
0.118
0.402
0.401
Note: t-statistics are based upon Whitney-West heteroskedasticity and autocorrelation consistent standard errors. The
cyclically adjusted earnings yield is the inverse of the Shiller US (SPX) CAPE. Source: Barclays Research
The historical co-movement between demographic variables and the real interest rate
points to very large effects. To illustrate, if we apply the coefficients in column 1 to the
changes in age structure projected for the coming 10 and 20 years, the equation would
imply an increase in the real interest rate (relative to 2014) of nearly 4.5% in 10 years and
9% in 20 years.
We think there are good reasons not to take these estimates as reliable guides to the impact
of future demographic change on the real interest rate, as we discuss in the text. The
calculation serves to illustrate, however, that the historical co-movement between real
interest rates and age structure is not only statistically, but economically very significant.
The results do not, in our view, provide a convincing forecasting framework. But they do
corroborate the implication of the results on saving propensities that demographic forces
comprise a slow-moving but potentially very powerful influence on asset markets.
The numerical analysis of equity valuations points in the same direction, and comes with
the same limitations.
24 February 2015
24
References
Abel, Andrew B. (2004) The effects of a baby boom on stock prices and capital
accumulation in the presence of social security, NBER working paper no 9210.
Bond, Tim (2009) The lost decade, in Barclays Research Equity Gilt Study, 2009.
Bosworth, Barry (2014) Interest rates and economic growth: Are they related?, Brookings
Institution.
Callen, Tim, Nicoletta Batini and Nicola Spatafora (2004) How will demographic change
affect the global economy?, in IMF World Economic Outlook, September 2004.
Furceri, Davide and Andrea Pecatori (2014) Perspectives on Global Real Interest Rates, in
IMF World Economic Outlook, April 2014.
Gapen, Michael (2013) Demand for safe havens to remain robust, in Barclays Research
Equity Gilt Study, 2013.
Gapen, Michael (2015) The Great Destruction, in Barclays Research Equity Gilt Study,
2015.
Gavin, Michael, Piero Ghezzi, Sebastian Brown and Alanna Gregory (2012) The equity risk
premium: Cheap equities or expensive bonds? in Barclays Research Equity Gilt Study,
2012.
Geanakoplos, John, Michael Magill and Martine Quinzii (2004) Demography and the longrun predictability of the stock market, Cowles Foundation working paper no. 1099.
Grigoli, Francesco, Alexander Herman and Klaus Schmidt-Hebbel (2014) World Saving,
IMF working paper no WP/14/204.
Hansen, Bruce E. and Ananth Seshadri (2013) Uncovering the relationship between real
interest rates and economic growth, University of Michigan.
Modigliani, Franco and Shi Larry Cao (2004) The Chinese saving puzzle and the life-cycle
hypothesis, Journal of Economic Literature.
Poterba, James (2004) The impact of population aging on financial markets, NBER
working paper no. 10851.
Poterba, James, Steven Venti and David Wise (2011) The composition and draw-down of
wealth in retirement, NBER working paper no. 17536.
Terrones, Marco and Roberto Cardarelli (2005) Global imbalances: A saving and
investment perspective, in IMF World Economic Outlook, September 2005.
24 February 2015
25
CHAPTER 2
The magnitude and speed of the collapse in oil has roiled markets; only the selloffs
in 1997-98 (61%), 1986 (67%) and 2008 (73%) were larger than the recent one
(60%). In order to assess the sustainability of lower oil prices and their effects on
the global economy and markets, we construct a model to explain real WTI oil prices
based on the global demand-supply balance for oil, global IP, OPEC market share
and real US power prices. The model explains 89% of oil price moves since 1991,
including the boom, and, importantly, the collapse since June; oil prices fell more
than the drivers implied but recovered some losses recently. The US dollar and
speculative positioning provide additional explanatory power.
The medium-term drivers in our model suggest that lower oil prices are likely to
persist. Demand growth is slowing, driven by energy efficiency and lower aggregate
growth globally. Moreover, oil should remain a well supplied market, with US tight oil
keeping OPEC in check.
Inflation expectations and, thus, bond yields have reset lower in response to the
collapse in oil. Our findings suggest that emerging market inflation will be affected
more than developed market inflation. Headline inflation volatility should be lower,
all else equal, with lower energy weights in CPI on sustained lower oil. Importantly,
the fall in bond yields (US 10y) that is typical in oil selloffs tends to be fairly sticky,
with yields settling 15% below the higher levels prior to the oil selloff.
Global growth should get a 0.1pp boost for every 10% drop in oil prices based on
our model, or 0.4-0.5pp in 2015 if oil prices stay in the current range. As the growth
benefits tend to manifest with a 2-3 quarter lag, the market also prices the benefits
with a lag; the S&P 500 rallies 12% on average the year after oil troughs.
The fall in Brent is a $1.5-2.0trn annualized redistribution from oil producers to oil
consumers and is equivalent to more than 2% of global GDP. Sector beneficiaries of
lower oil, such as consumer discretionary, should continue to outperform,
particularly since both earnings forecasts and relative valuations do not reflect the
upside. Asia disproportionately benefits as it is the largest net importer of oil and
equity valuations are still in line with the rest of EM.
Current accounts and terms of trade will be profoundly affected by sustained lower
oil. A narrower US petroleum deficit should ease the impact of oil price moves on the
dollar. The currencies of oil exporters have largely adjusted, but we think
Asian currencies that benefit from the terms of trade shock have further to run (eg
INR, KRW).
Financial stress often follows periods of oil weakness. However, key differences in
country fundamentals and market dynamics suggest the risks to financial stability
are lower this time. That said, risk premiums for energy-related assets should
remain elevated in an environment of sustained lower oil.
26
FIGURE 1
The recent oil selloff was near historical extremes
FIGURE 2
A few oil selloffs have been longer
0%
800
-10%
700
-20%
600
-30%
500
-40%
400
-50%
300
-60%
Apr-00
Jul-86
Jun-12
Feb-91
Mar-86
Oct-11
Jan-07
Dec-08
Jun-90
Recent
Nov-01
Oct-88
Dec-98
Jun-12
Apr-00
Oct-11
Jan-07
Jun-90
Jul-86
Dec-93
Oct-88
Feb-91
Nov-01
Recent
Dec-98
-90%
Mar-86
100
Dec-08
-80%
Dec-93
200
-70%
First and foremost, crude stock changes drive oil prices: when demand is greater than
supply, oil prices rise (and vice versa).
Second, the global business cycle explains much of the movement in oil prices not
explained by stock changes. When global industrial production (IP) is growing above
trend, oil prices rise even faster as supply tends to lag.
Third, OPEC is an important driver of oil prices. A higher OPEC share of crude supply has
coincided with higher oil prices.
Fourth, power prices help to explain oil spikes when oil is burned to meet unanticipated,
peak power demand. Moreover, we believe power prices also capture the potential
long-term threat to oil from energy substitutes such that investment responds when oil
prices go well above the alternative cost of energy like we saw this past cycle.
Finally, the US dollar and speculative oil positions affect oil prices on a shorter horizon.
Coming out of the EM crises of the late 1990s and the US recession of 2001, the global
economy embarked on a synchronized and unparalleled growth cycle fueled by credit.
Against a backdrop of surging aggregate and energy demand, oil supply growth lagged
notably after years of underinvestment amid weak oil prices. In particular, OPEC supply
failed to react to much higher prices, sowing the seeds of the structural trends that are now
in place, including energy efficiency, US tight oil supply growth, and oil substitution.
Ye, Zyren, Shore (2002). Forecasting Crude Oil Spot Price Using OECD Petroleum Inventory Levels. International
Advances in Economic Research.
24 February 2015
27
FIGURE 3
Demand-supply imbalances explain the majority of oil price moves
2.00
4.00
2.05
2.10
3.50
2.15
2.20
3.00
2.25
2.30
2.50
2.35
Jun-13
Sep-14
Dec-10
Mar-12
Jun-08
Sep-09
Sep-04
Dec-05
Jun-03
Dec-00
Mar-07
Mar-02
Jun-98
Sep-99
Mar-97
Dec-95
Jun-93
Sep-94
Dec-90
Mar-92
Jun-88
Sep-89
Dec-85
2.00
Mar-87
2.40
From the 1999 OPEC decision to adhere to a quota until 2011, demand considerably
outstripped supply and real oil prices more than doubled. Even though OPEC producers
have the vast majority of the worlds recoverable reserves, OPEC supply rose from 34mb/d
in 2004 to 36mb/d in 2013, a CAGR of just 0.5%.
The recent fall in oil prices has
been larger than supplydemand imbalances would
have suggested
US crude production began to surge in 2012, rising by an extraordinary 3mb/d in about two
years to more than 9mb/d in 2014. It took about two years for US supply to get back to
1985 levels, the point at which Saudi Arabia abandoned the quota. Middle East disruptions
offset much of the supply growth until 2014, when Libya came on line, demand weakened,
and OPEC decided not to cut its production. At 2mb/d y-o-y, non-OPEC supply grew at a
record level and at three times the rate of demand growth in 2014. Accordingly, the fall in
oil prices has been notable since mid-2014, but larger than recent supply-demand
imbalances would have suggested.
A Fed paper in 20112 highlighted the significance of global industrial production (detrended) in explaining crude prices. Global IP that is growing faster than trend means that
oil demand is likely growing faster than trend, which has a big impact on prices because
supply tends to adjust more slowly. In our analysis, we also find that global IP in log terms
and de-trended explains much of the residual from the demand-supply balance variable
mentioned above (Figure 4); the trend growth rate from 1991 to 2013 is 2.8%. The
coefficient for global IP is highly significant and sizeable in a joint regression. From 2002 to
2008, global IP grew at an above-trend rate for more than six straight years in a
synchronized global recovery and, accordingly, oil prices surged. Since the world has
recovered from the financial crisis, global IP has grown at essentially trend rates and oil
prices have remained relatively flat. Structural factors suggest that potential growth may be
lower, as we highlight in Chapter 4, The great destruction.
Alquist, Killian, Vigfusson (2011). Forecasting the Price of Oil. Board of Governors of the Federal Reserve System,
International Finance Discussion Papers, Number 1022.
24 February 2015
28
FIGURE 4
The global business cycle is a key driver of oil prices
0.10
4.50
0.06
4.00
3.50
0.02
3.00
-0.02
Lower
growth
potential?
-0.06
2.00
IP vs Trend (lhs)
Mar-16
Mar-15
Mar-14
Mar-13
Mar-12
Mar-11
Mar-10
Mar-09
Mar-08
Mar-07
Mar-06
Mar-05
Mar-04
Mar-03
Mar-02
Mar-01
Mar-00
Mar-99
Mar-98
Mar-96
Mar-97
Mar-95
Mar-94
Mar-93
Mar-92
1.50
Mar-91
-0.10
2.50
Source: Netherlands Bureau for Economic Policy and Analysis, Haver, Barclays Research
OPEC share of supply is a key variable; US tight oil was a game changer
We view OPECs current desire
to maintain and regain share
as reflective of the competitive
balance provided by US tight
oil supply
The role of OPEC is discussed at length in the literature. We use OPECs share of supply to
capture its power. The higher OPECs market share, the higher the price of crude (Figure 5).
OPEC market share grew from about 36% in 2002 to about 42% in 2008 and was again
near that level at the end of 2012, before US production began to surge. This rise in OPEC
pricing power magnified the boom in oil prices. With the boom in US tight oil, OPECs share
has now fallen to about 39.5%, its lowest level since 2004, which, all else equal, points to a
lower price for crude. That said, a concerted rise in Saudi and OPEC production, as occurred
after 1986, significantly pushed out the supply curve, with the new supply added at the
lowest cost point on the curve. We view OPECs current desire to maintain and regain share
as reflective of the competitive balance provided by US tight oil supply. This is different from
1986, when Saudi production needed to return toward normal levels after years of cutting
supply unsuccessfully.
FIGURE 5
Oil prices have moved with OPEC market share
4.50
0.44
4.00
0.40
3.50
3.00
0.36
2.50
0.32
2.00
Sep-14
Jun-13
Mar-12
Dec-10
Sep-09
Jun-08
Mar-07
Dec-05
Sep-04
Jun-03
Mar-02
Dec-00
Sep-99
Jun-98
Mar-97
Dec-95
Jun-93
Sep-94
Mar-92
Dec-90
Sep-89
Jun-88
Dec-85
1.50
Mar-87
0.28
Power prices capture peak power demand for crude, and oil substitutes
Oil is often burned during peak power demand (cold winters, hot summers, etc). As a result, it
is important to capture marginal and often unanticipated demand for oil, although the two
tend to move contemporaneously. We use real power prices from US CPI data (ie, energy
services divided by CPI). Additionally, the longer-term path for power is seemingly converging
24 February 2015
29
around a grid that draws from multiple energy sources. Crude as a source for power is likely to
continue to decline (even in Saudi Arabia), as other sources become cheaper. The ultimate
threat to crude and gasoline is the electrification of cars (or fuel cells), which is far off but still
worth including in a macro framework for oil prices, in our view.
The ultimate threat to oil is the
electrification of cars, which is
far off but still worth including
in a macro framework for oil
prices
The oil price rise that began in 1999 coincided with the rise in real power prices as demand
for energy in aggregate soared (Figure 6). Big spikes in oil (eg, 2000 and 2005-07)
coincided with surges in power prices driven by large demand-supply imbalances. Overall,
since natural gas production began to take off in 2007, real power prices have continued to
decline as natural gas prices have fallen.
FIGURE 6
Power prices capture peak oil demand as well as potential threats to oil
4.60
4.50
4.55
4.00
4.50
?
4.45
3.50
4.40
3.00
4.35
2.50
4.30
Dec-15
Jun-13
Sep-14
Dec-10
Mar-12
Jun-08
Sep-09
Dec-05
Mar-07
Jun-03
Sep-04
Dec-00
Mar-02
Jun-98
Sep-99
Mar-97
Dec-95
Jun-93
Sep-94
Dec-90
Mar-92
Jun-88
Sep-89
Dec-85
2.00
Mar-87
4.25
The dollar and speculative positioning are likely exacerbating oil price moves
A number of researchers find that speculative CFTC positioning in oil is a significant factor in
explaining oil price movements 3 as it often captures the price premia associated with
geopolitical risk and price shocks from supply disruptions. We find similar results using net noncommercial CFTC futures for WTI, scaled by open interest; oil prices and speculative positioning
often move together (Figure 7). Speculative positions in WTI futures reached a record 21% of
open interest at the June 2014 peak. The covering of oil longs exacerbated the selloff.
FIGURE 7
CFTC speculative oil positioning exacerbates price moves
25%
4.50
20%
4.00
15%
10%
3.50
5%
3.00
0%
2.50
-5%
Sep-15
Nov-14
Jan-14
Mar-13
Jul-11
May-12
Sep-10
Nov-09
Jan-09
Mar-08
May-07
Jul-06
Sep-05
Nov-04
Jan-04
Mar-03
Jul-01
May-02
Sep-00
Nov-99
Jan-99
Mar-98
Jul-96
May-97
2.00
Sep-95
-10%
24 February 2015
Antonio Merino and Alvaro Ortiz, Explaining the so-called price premium in oil markets, 2005, OPEC Review.
30
The literature on the relationship between the trade-weighted US dollar and oil prices
essentially shows that the dollar has less long-term power in terms of predicting and
explaining oil price moves45. However, over short-term horizons, there is a more robust
relationship between the dollar and oil6. As the goal of our model is to have a medium-term
framework for oil prices, the coefficients on the level of the US dollar become less
significant when the fundamental drivers are included. However, including q-o-q log
changes in the trade-weighted dollar into our fundamental model captures the essence of
the short-term impact of dollar moves on the price of oil.
A macro model of oil prices: The recent collapse in the context of the drivers
The r-square of the model is
89% since 1991; the fit
improves to 94% after
including the dollar and spec
positions
Based on our findings we construct a model to explain real WTI oil prices (in log terms) over
longer horizons based on the global demand-supply balance for crude, global IP (de-trended),
real US power prices and OPEC market share (Figure 8). The r-square of the model is 89%
using quarterly data available since 1991. Including CFTC speculative positioning with data
available since 1995, as well as the dollar, improves the fit slightly. The model does a good job
explaining all of the major turns in real oil prices in the last 20 years.
FIGURE 8
Table of select variables
Univariate
Coeff
T-stat
Coeff
T-stat
Coeff
T-stat
Stock change
-6.0
-13.7
-4.7
-14.0
-4.3
-15.0
Global IP
2.8
1.5
4.5
6.3
3.2
5.1
Real power
5.4
10.3
2.3
6.8
2.3
7.1
OPEC share
22.1
7.1
6.4
4.0
7.5
3.6
USD qoq
-1.0
-1.7
0.9
2.3
0.89
R-square
0.94
FIGURE 9
Fundamental variables do a good job of explaining oil price
moves
FIGURE 10
The recent plunge looks to have been exacerbated by the
dollar and spec positioning
Nov-13
Jan-11
Jun-12
Aug-09
Oct-06
Mar-08
May-05
Jul-02
Dec-03
Feb-01
Apr-98
Sep-99
Nov-96
Jan-94
Mar-91
Nov-13
Jan-11
Fitted
Jun-12
Aug-09
Oct-06
Mar-08
May-05
Jul-02
Dec-03
Feb-01
Apr-98
Sep-99
Nov-96
Jan-94
2.0
Jun-95
2.0
Mar-91
2.5
Aug-92
2.5
Jun-95
3.0
Aug-92
Fitted
Galbraith, Dufour, Zhang (2014). Exchange rates and commodity prices: measuring causality at multiple horizons.
CIRANO working paper 2013-s39. Submitted to Journal of Empirical Finance.
5
Christian Grisse (2010). What Drives the Oil-Dollar Correlation? Federal Reserve Bank of New York Working Paper.
6
Ferraro, Rogoff, Rossi (2011). Can Oil Prices Forecast Exchange Rates? Federal Reserve Bank of Philadelphia.
Working Paper No. 11-34.
24 February 2015
31
Fundamentals explain roughly 35pp of the selloff; the dollar and speculative positioning
another 10pp (Figures 9-10). Based on Barclays commodities analysts estimates in the Blue
Drum, the stock build since Q2 2014 has been nearly 5% of demand, which would imply a
23% decline in oil prices since June. Global IP has been volatile but at fairly flat levels through
October. An extrapolation of the 1pt decline in the global manufacturing PMI would point to a
1pp slowing in global IP in Q4, which in turn would point to a 4% decline in oil prices. Real US
power prices ticked higher in December after a cold November, but the collapse in natural gas
since late December suggests power and heating demand had fallen as US winter
temperatures have been warmer. The decline in natural gas explains part of the last leg down
in oil based on our model, though gas production may have played a bigger part. The fall in
OPEC share due to rising US supply in 2014 contributed just 2% to the oil price fall, though the
signalling impact was much greater. Including the surge in the dollar and the covering of net
speculative positions would account for a further 10pp of the decline. Our oil strategists see
fundamental drivers leading oil even lower before a recovery to $60 in 2016 (see Blue
Drum, 28 January 2015). They assume that OPEC will maintain its position, that non-OPEC
supply growth will stay firmly positive, and that oil consumption will be slow to respond to
lower prices. Brent prices are forecast to average $44/b in 2015. During 2015, lower prices
should have only a muted effect on the demand response and not be sufficient to balance
the market on their own given various obstacles preventing the full feed-through in enduser pricing. Even non-OPEC supplies ex-US will likely grow, riding the momentum from
several years of sustained $100/b oil before the crash. After averaging very low levels
during 2015, we expect prices to rebound to $60 in 2016. Low prices will likely curb nonOPEC supply growth to just 0.3 mb/d as demand grows 1.3 mb/d in 2016.
FIGURE 12
Headwind from vehicle fuel efficiency will be considerable
85
75
EU 2025: 72.5-82.3
MPG,
normalised
to US CAFE
test cycle
65
US 2025: 56.2
45
China 2020:
50.1
24 February 2015
Dec-14
Jun-11
Mar-13
Sep-09
Dec-07
Jun-04
Mar-06
Sep-02
Dec-00
Jun-97
Mar-99
Sep-95
Dec-93
Mar-92
Jun-90
Sep-88
Dec-86
35
OECD (ln)
EU 2021: 60.6
55
25
2000
2004
2008
2012
2016
2020
2024
EU enacted
EU proposed
US enacted
China enacted
China proposed
Japan enacted
Source: ICCT
32
Global growth potential is lower. Slower potential growth in developed economies and a
decelerating Chinese economy have reduced global potential growth by 1.5pp a
significant deceleration (see Chapter 4, The great destruction). Therefore, a return to
above-trend growth is unlikely and a lower trend rate is probable. From a cyclical
perspective, global IP is near trend levels (using the trend in place since the early 1990s).
This reflects developed market production, which is still below potential, and emerging
market output, which is slowing. As a result, the likelihood of a strong cyclical rebound like
2009 and the early 2000s is low. Given that potential growth rates have fallen and growth is
so asynchronous, we see weaker global IP growth pointing to flat to slightly lower real oil
prices in the medium term.
Rising fuel efficiency will be a 4% headwind for DM and China. Our auto analysts believe
that globally tightening fuel economy and emissions standards have forced automakers to
focus on Powertrain engineering to design more fuel efficient vehicles and that stringent
European standards mean that electrification is here to stay (see Future Powertrain: Premium
pain?, 3 July 2014). In the US, CAFE7 standards for passenger cars were basically the same
from 1985 to 2010, at 27.5 mpg. Now, standards in the US as well as in Europe and China are
forecast to rise at a 4+% CAGR until 2025 (Figure 12). The average age of the US vehicle fleet
is 11 years. CAFE standards are set to go from 35 to 56.2 to 2025; this is a 4.4% annual
headwind to oil gasoline demand on the simple assumption that the entire fleet turns over in
11 years. The adoption of more efficient standards by EM countries as they become more
affordable could also weigh on the trajectory of demand growth, though, as older cars find
their way into EM markets anyway, the efficiency standards should have a lagged impact.
Fossil fuel subsidies of $550bn8 globally are being reduced. As oil prices have fallen,
governments in China, India and Indonesia have used the opportunity to reduce fuel
subsidies. As a result, lower oil prices are not fully flowing through to the consumer, which
should in turn lower the demand response. In the Middle East, nearly 2 mb/d of oil are used
to generate electricity when renewable energy would be competitive absent subsidies,
according to the IEA. In Saudi Arabia, fuel subsidies discourage shifts to more fuel efficient
cars and usage patterns. We also find that Middle East demand is well above the trend that
was in place prior to 2004, likely reflecting positive spillovers from the oil boom.
7
8
24 February 2015
33
US tight oil is quicker to adjust, reducing the likelihood of persistent undersupply. The
marginal source of supply is essentially US tight oil, which has much lower lead times.
Therefore, the likelihood of persistent periods of undersupply (eg, 1999-2011) should be
greatly reduced. In the near term, supply growth has once again been slow to adjust to
shifting demand growth, but the market has moved from broadly undersupplied with little
excess capacity to one with greater excess capacity.
US tight oil supply costs have plummeted resilient supply ahead (see Re-examining the
cost of oil production, 9 January 2015). Our E&P analysts believe that US supply costs have
fallen sharply and that lower supply costs will enable US producers to continue growing
volumes despite a sharp pullback in CAPEX. Reduced capex will slow the growth rate, but
not materially, as producers can earn 6-9% at $60/bbl and perhaps lower. After the 1986
collapse in oil, non-OPEC supply stayed relatively flat at around 44mb/d up until 1991 and
was fairly resilient, while Saudi Arabia was raising production by 5mb/d. On the other hand,
planned Middle East investment growth, despite falling prices means that oil supply is
unlikely to collapse in response to lower oil prices in the medium term, as some worry. The
IEA expects supply growth to come from the US, Canada and Brazil prior to 2020, with
Middle East production growing about 10mb/d from 2020 to 2040.
Geopolitical risks remain a concern. Only last summer, investors feared that ISIS could
disrupt Iraqi oil, driving oil futures higher. Disruptions are still a big risk, but the build-up of
inventories should provide a buffer. Additionally, sanctions against Iran have removed a
significant amount of crude from the overall market that could eventually return.
FIGURE 14
US crude production rose 3mb/d in 2 years
9,000
52
8,000
OPEC decides
not to cut
production
34
48
7,000
19
44
6,000
14
40
5,000
24 February 2015
Jan-15
Jan-13
Jan-11
Jan-09
Jan-07
Jan-03
Jan-01
Jan-99
Jan-97
Jan-95
Jan-93
Jan-91
Jan-89
Jun-12
Mar-14
Sep-10
Dec-08
Jun-05
Sep-03
Dec-01
Mar-07
Non-OPEC (rhs)
4,000
Jan-87
OPEC (lhs)
Mar-00
Jun-98
Sep-96
Dec-94
Jun-91
Mar-93
Sep-89
Dec-87
Mar-86
Jan-83
24
OPEC 1999
quota marks
shift
OPEC
abandoned
quota
Jan-85
29
Jan-05
39
34
Real power prices in the US rose by nearly 30% between 2000 and 2008, driven by the
cyclical boom in energy demand and lagging supply. Additionally, emission abatement
costs have been and will continue to be a headwind, with cheaper coal plants being retired,
but significant utility investment and the growing supply of natural gas should more than
compensate. Importantly, the cost of alternatives continue to decline, with solar
photovoltaic costs (PV) falling at a 10% annual rate and wind costs fairly inexpensive
(Figure 15). From a longer-term perspective, our strategists believe that a confluence of
declining PV cost trends and residential-scale power storage is likely to disrupt the status
quo (see The Solar Vortex, 23 May 2014). Falling alternative energy prices should in theory
put a medium-term cap on oil prices; if oil gets too expensive and the payback period for an
electric car is short enough, consumers will convert. Overall, the IEA sees consumer energy
costs rising only modestly in Europe and Japan to 2040, slightly more in the US, and the
most in China and India.
The real price of oil more than doubled from 2000 to 2014, while prices overall remained
much more benign, particularly for core goods. From a high level, there is little apparent
economic justification for the relative rise outside of scarcity value (real or perceived). The
marginal cost of supply was incrementally more expensive, but the capex boom and its
relative inefficiencies followed nearly two decades of lackluster investment. On the demand
side, fuel efficiency standards were flat for nearly 25 years. Typical productivity
improvements did not naturally flow through to oil prices. The fall in oil prices likely reflects
in part the efficiency gains of continued investment.
The average real price of oil from 1986 to 1999 would point to a current price of $25/b to
$45/b (Figure 16). Similarly, the price of oil relative to the price of natural gas remained in a 515x band until 2009, when it broke out as oil prices rebounded and natural gas prices
remained weak. Although the dynamics are different, the gap to natural gas also points to the
potential for costs to come down.
Marginal cost curve should continue to show efficiency gains. From an economic
perspective, the removal of cartel pricing power in an oversupplied/balanced market means
prices should settle around the marginal cost. Our US E&P and oil analysts note that US
tight oil costs are likely to continue to fall, driven by technology, lower service costs (~20%
declines) and improved efficiency from high-grading (drilling the best wells). In the
current environment, there is an incentive for producers to lower costs and maintain/grow
production. The magnitude of the boom likely exacerbated input cost increases for capital
FIGURE 15
Solar PV costs continue to decline and become competitive
13.00
60
12.00
50
11.00
40
10.00
-10% CAGR:
2007-2013
9.00
30
2000
2002
24 February 2015
2004
2006
2008
2010
2012
WTI/CPI
Feb-12
May-14
Nov-09
Aug-07
Feb-03
May-05
Nov-00
Aug-98
Feb-94
May-96
Nov-91
4.00
1998
Aug-89
5.00
Feb-85
May-87
6.00
Nov-82
10
Feb-76
7.00
Aug-80
20
May-78
8.00
FIGURE 16
On a relative basis, oil is falling toward pre-boom levels
WTI/Natural Gas
35
(rig rates, etc) and labor (energy wages costs in the US have risen much faster than the
national average). We see scope for oil supply costs to decline as the industry restructures
to persistently lower prices.
Oil prices should be less volatile after settling near a new equilibrium
Inventories are back at very high levels, providing a cushion. Global inventories of crude
and product have risen notably in recent months. Saudi inventories are near their peaks and
US petrol inventory is back near peak levels of 95-100 days of average demand. We find
that oil price volatility is inversely related to the level of total crude and product inventory.
High crude and product inventories should mean lower oil price volatility, all else equal,
because unanticipated demand can more easily be met by inventory (Figure 17).
After bouncing off the trough, oil prices have stayed fairly flat in past episodes.
Performance of oil is highly varied in the 13 episodes in which oil has sold off by more than
30%. But in 10 of the episodes, oil prices two years after the trough were basically the same
as prices one year after the trough (Figure 18). The flat trajectory likely reflects a more
balanced demand-supply backdrop after the corrections.
The fall in oil prices has an immediate impact on inflation. It is logical then that the collapse
in oil prices has been quickly priced into inflation expectations and thus bond yields. We
estimate the inflation effects from lower oil and find that emerging markets should feel a
greater impact than developed markets. Sustained lower oil prices mean that energy
weights in CPI will be lower and thus headline inflation vol should be lower, all else equal.
Importantly, the fall in bond yields that is typical in oil selloffs tends to be fairly sticky, with
yields settling 15% below higher levels prior to the oil selloff.
The effect of a sharp decline in oil prices on domestic inflation varies greatly across
countries and derives predominantly from the direct effect of oil prices on headline CPI
(weight of energy in CPI), and the indirect effect, which is estimated by the long-term passthrough of a change in oil price to the CPI via other prices. The direct weight of energy in
CPI averaged 10.4% in EM versus 8.9% in advanced economies. Since overall inflation in EM
is considerably higher than in advanced economies, the contribution of energy to EM
inflation is proportionately higher. In terms of the indirect effect, we found that the
aggregate EM long-term pass-through from oil onto inflation is 0.051 versus 0.045 in
FIGURE 18
After bouncing off a trough, oil prices have had a flat trajectory
65
24 February 2015
Feb-91
Dec-08
Mar-86
Nov-01
Feb-15
Feb-13
Feb-11
Feb-09
Feb-07
Feb-05
Feb-03
Feb-01
Feb-99
Feb-97
Feb-95
Feb-93
0.1
Feb-91
110
Oct-11
0.3
105
Dec-93
100
Apr-00
0.5
95
Median = 83.45
Jun-90
0.7
90
Jun-12
85
Oct-88
0.9
80
Jan-07
75
180
160
140
120
100
80
60
40
20
0
Jul-86
1.1
70
Dec-98
FIGURE 17
Oil prices should be less volatile if inventories stay high
trough + 24 month
36
FIGURE 19
Energy pass-through effects vary notably
20
18
16
FIGURE 20
EM currency depreciation has been much lower this time
0.2
0.18
0.16
50
40
12m %
change
12m $
change
80
60
0.14
30
40
12
0.12
10
0.1
20
20
0.08
10
0.06
0.04
0.02
0
Hungary
Slovakia
Poland
India
Czech
Turkey
Malaysia
Thailand
Russia
S. Africa
Korea
China
Brazil
Mexico
Chile
Philippines
Israel
Colombia
Singapore
Indonesia
Peru
Euro zone
USA
UK
Japan
14
-20
-10
-40
-20
-60
-80
-30
75 77 79 81 84 86 88 90 93 95 97 99 02 04 06 08 11 13
vs. DM
vs. EM
WTI (rhs)
advanced economies (see: The crude reality, 8 April 2011). Unsurprisingly, in places where
energy prices are heavily regulated, eg, India and the euro area, the long-term pass-through
coefficients are lower (Figure 19).
It should be noted that the pass-through coefficients presented here might be slightly
overstated because the importance of oil for the global economy has in fact been declining9.
In addition, the loosening of real wage rigidities implies a more muted response from
inflation and output, which means that the long-term pass-throughs are lower today
relative to earlier in the sample. Nevertheless, the overstatement would be mostly
consistent. We therefore expect a higher impact from lower energy prices on EM inflation
relative to DM inflation, on an aggregate basis.
There is less of an offset from
EM currency depreciation in
this oil selloff
There is less of an offset from EM currency depreciation in this oil selloff. In the 1980s
and 1990s, oil price falls coincided with large EM currency declines vs the dollar (Figure 20).
EM currency depreciation through the recent fall in oil was notable, but considerably less
than in past episodes and also relative to advanced currencies. Thus, the overall inflation
response in EM countries should be greater relative to history as well as versus many other
developed markets.
The fall in rates that is typical in oil selloffs tends to be fairly sticky
The US 10y Treasury yield declined in 11 of the 13 episodes in which oil has sold off by
more than 30%. We find that the recent decrease in bond yields aligns with this pattern
(Figure 21). The larger-than-implied decline in 10y yields in 2011 was likely affected by the
Feds calendar date guidance. On the other hand, the slight rise in yields in 1988 and 1990
was likely due to core inflation running above 4%.
The decrease in nominal bond
yields was essentially in line
with the fall in inflation
expectations until early 2015
Falling inflation expectations have fueled most of the decline in nominal yields. As our
inflation strategist has stated (see Oil slick, December 11, 2014), the link between oil and
front-end inflation breakevens is clear, given the pass-through of gasoline to CPI; however, the
flattening of the inflation curve and the fall in longer duration breakevens have been marked.
Indeed, the correlation between daily oil price changes to daily changes in US 10y inflation
breakevens has risen to exceed 60% in the past three months (30% for German breakevens).
Since the June peak in oil, 10y inflation breakevens for the US and Germany fell by about 75bp
from peak to trough (Figure 22). Until early this year, the decrease in nominal bond yields was
essentially in line with the fall in inflation expectations as real yields in the 10y sector stayed
See: Olivier J. Blanchard and Jordi Gali, The macroeconomic effects of oil shocks: Why are the 2000s so different
from the 1970s?, 2007, NBER working paper No. 13368.
24 February 2015
37
FIGURE 22
Breakevens have fallen globally since June as oil has fallen
Brent
160
0%
140
US 10y infl BE
Dec-14
Nov-14
-50%
-20%
Oct-14
-30%
Sep-14
-40%
20
Aug-14
-50%
40
Jul-14
-60%
60
Jun-14
-70%
80
May-14
-80%
100
Apr-14
-40%
2011
120
Mar-14
-30%
2.5
2.3
2.1
1.9
1.7
1.5
1.3
1.1
0.9
0.7
0.5
Feb-14
-20%
Infl BE
Jan-14
-10%
10%
Dec-13
Jan-15
FIGURE 21
The fall in bond yields is in line with past oil selloffs
relatively flat; ECB QE likely helped drive the recent decline in real yields as oil and inflation
expectations stopped falling. Clearly, central bank policy is also a key driver.
A lower energy weight in CPI should mean less inflation volatility, all else equal. The fall in
energy prices has led to a fall in the weight of energy in the CPI from over 10% in 2011-13 to
probably less than 8% now; the weight of energy commodities has also fallen, by about 2pp.
Since 1986, the volatility in monthly energy prices has been more than 9x that of the headline
CPI index. Therefore, a lower weight of energy should logically lead to lower headline inflation
volatility (Figure 23). Indeed, the correlation between the energy weight in CPI and the twoyear volatility of the headline index is 57%. Additionally, as noted above, higher oil and oil
product inventories could eventually lead to lower volatility in oil prices, all else equal. So the
combination of a lower energy weight and potentially lower oil volatility points to lower
energy inflation vol and thus lower overall inflation vol in the medium term.
US 10y has had a flattish trajectory after past oil troughs, but is Fed-dependent. Although
oil has tended to bounce after hitting a trough, the US 10y yield one and two years after an
oil trough has actually been flat, using the median move of the episodes. On average, the
US 10y yield settles about 15% below the higher levels prior to the oil selloff and remains in
a lower range for two years (Figure 24); this implies a range around 2.2% for the US 10y.
24 February 2015
80
75
Average
23
21
70
19
Jun-14
Oct-10
Aug-12
Feb-07
Dec-08
Jun-03
Apr-05
Oct-99
Aug-01
Feb-96
Dec-97
Jun-92
Apr-94
Oct-88
Aug-90
0.005
Dec-86
85
17
0.01
15
90
13
0.015
11
10
95
0.02
11
100
0.025
12
0.03
-1
13
105
-3
Stdev
-5
Wt
FIGURE 24
The fall in the US 10y yield that is typical in oil selloffs tends
to be fairly sticky
-7
FIGURE 23
Headline inflation volatility has historically tracked the
energy weight in CPI
Current
38
The rise in the 10y after 1986, 1993, and 1998 coincided with subsequent Fed hikes as
growth and inflation recovered, while the increase after 2012 stemmed from the taper
communication. The bounce in bond yields after 2008 was from extreme levels. During the
other eight episodes, US 10y yields were flat to lower a year after oil troughed.
The positive impact from lower oil on consumption and growth tends to manifest with a 23 quarter lag. Similarly, the market tends to price the benefits into growth assets with a lag;
the S&P 500 rallies 12% on average the year after oil troughs. The sector beneficiaries of
lower oil, such as consumer stocks, should continue to outperform amid sustained lower
oil, particularly since relative valuations do not reflect the upside. Across regions, Asia
disproportionately benefits since it is the largest net importer of oil; thus, an equity
valuation that is in line with the rest of EM suggests the upside is not fully priced.
To assess the impact of oil price moves on global growth under different price scenarios, we
use the framework developed in Easy money is not easy for all EM, (23 November 2010) and
The Crude Reality (8 April 2011). The main channel via which oil can affect global growth is
consumption, which generally accounts for 60% of GDP. In particular, we model annual real
consumption growth as a function of annual real GDP growth, the level of the policy rate,
and the annual change in oil prices all variables in the coincident period and with one lag
(using quarterly data). This allows us to isolate the effect of oil prices on consumption,
controlling for overall activity. We ran this exercise for both oil exporters and oil importers.
The resulting long-term coefficient has the expected negative sign for oil importers (ie, an
increase in oil price has a negative effect on consumption, and vice versa). It is also negative
for oil exporters, but much less so. This makes sense given the contrasting wealth effects
arising from such price shocks in each group. At the global level, every 10% decline in oil
prices is a 0.1pp boost to GDP (within one year of the price shock).
We use the coefficients from this exercise10 to simulate the global growth effect under
different price scenarios. In a scenario where prices reach a floor of $40/b and remain there
for the next two years, as Figure 26 shows, we estimate an incremental positive effect on
global GDP of up to 0.5pp in 2015 that gradually weakens to 0.2pp by end of 2016. If prices
stabilize at a level of ~$60/b by Q3 2015, the effect on global growth would reach 0.4pp in
2015 and -0.03pp in 2016 (since by then the bounce back up in price would translate to a
negative real income effect on consumers). In the event that prices rise back up to $80/b,
the effect on global GDP is expected to reach 0.3pp in 2015 and -0.2pp by end 2016. Our
findings, based on broad sensitivities, are consistent with IMF and World Bank estimates as
well as Barclays economists forecasts of a modest boost to global growth from lower oil.
10
24 February 2015
That would still be valid today, given that it included 26 years of data.
39
FIGURE 25
Energy exports are highly concentrated
45%
35%
25%
15%
5%
-5%
Singapore
Ukraine
Thailand
Korea
Taiwan
S.Africa
Hungary
Turkey
India
Chile
Japan
Israel
Germany
Poland
China
USA
Indonesia
Brazil
Mexico
Malaysia
Russia
Nigeria
Venezuela
Iraq
Saudi Arabia
-15%
FIGURE 26
Global GDP growth impact (YoY) based on oil price scenarios
Effect on global growth of the continuation of the oil price shock
under different assumptions for the oil USD/bbl price
0.6
pp
0.5
0.4
0.3
0.2
0.1
0.0
-0.1
-0.2
-0.3
-0.4
Mar-14 Sep-14 Mar-15 Sep-15 Mar-16 Sep-16
80
60
40
The major role of supply in the current oil decline would point to a potentially larger boost
to growth, but a few factors point to a smaller impact. Low consumer confidence, as in
the euro area, lessens the positive impact. The freeing up of disposable income may be
saved instead of spent, and the impact on inflation may exacerbate deflation risks more
than it boosts real income (see Euro Themes: Assessing the impact of lower oil prices
and the euro on growth, inflation and ECB policy, 17 October 2014). Additionally, fuel
subsidies are being normalized. Governments have used the opportunity of the recent price
decline to gradually normalize prices, which means consumers are not feeling the full
benefits directly (Figure 27). However, as an offset, consumers will probably still enjoy the
benefits of improved fiscal balances. Finally, the zero bound limits the policy response,
which is typical after oil selloffs. Interest rates that are near or below zero narrow the ability
of monetary policy to respond to a disinflationary shock11. However, BoJ and ECB QE have
showed signs of affecting inflation expectations.
FIGURE 27
Fuel subsidies and taxes limit the pass-through effect
140%
FIGURE 28
The zero bound limits the central bank policy response
140
% y/y
120
120%
80%
60%
40%
10.0
100
7.5
80
100%
5.0
60
Brent
EU
Korea
China
Brazil
Poland
South Africa
Indonesia
20%
Jan-14 Mar-14 May-14 Jul-14
US
Japan
Australia
India
Mexico
Philippines
Turkey
Sep-14 Nov-14 Jan-15
12.5
2.5
40
0.0
20
0
-2.5
85 87 89 91 93 95 97 99 01 03 05 07 09 11 13 15
Oil prices (lhs)
US Central Bank policy rate (EOP, %, rhs)
CPI (% y/y, rhs)
See: Olivier Blanchard and Rabah Arezki, Seven questions about the recent oil price slump, December 2014, IMF
Direct blog.
24 February 2015
40
Asia-Pacific countries consume more than a third of global oil output, led by China, Japan
and India (Figure 29). Asias largest exporting economies are big importers of energy inputs
and, as such, their external balances are very sensitive to oil prices. Among the heaviest
importers of oil are India, Thailand, Taiwan and Korea (see: Asia Themes: Its oil good for
Asia, 2 December 2014).
We expect Chinas current account surplus to rise to $356bn this year as result of lower
oil. However, low energy prices and additional factors are driving inflation substantially
lower, creating the conditions for PBOC intervention, which we expected in H1 (see
Lower oil ignites fresh structural reforms, 5 December, 2014).
In Japan, sensitivity to oil prices has risen since the 2011 earthquake as the economy
has become more dependent on oil. As a result, we now estimate that a 10% decline in
oil prices would lead to an initial 0.1pp positive GDP shock in Japan, with greater impact
in the second and third years after the decline (see: Sizing up recent shocks to the global
economy, 24 October 2014).
In India, the most direct beneficiary of lower oil is the current account, which is now
expected to post a small surplus of 0.1% GDP in FY 15-16 (from a previous forecast of a
1.5% deficit). Growth is also expected to improve on increased fiscal revenues from
reduced subsidies and indirect taxes; however, the pass-through will likely take longer
(see: Lower oil- a boon for India, 29 January 2015).
In the US, we estimate 0.2pp additional GDP growth for every 10% fall in oil prices, more
than offsetting the expected drag from a stronger dollar. This is because of the high
propensity to consume, which means a rapid transmission mechanism from lower energy
prices to stronger real consumer spending (see Consumer windfall ahead, 17 October 2014).
As the worlds biggest oil consumer, the US should benefit notably from the fall in oil prices.
In the euro area, the impact of lower oil should be negligible, though still positive. The
transmission mechanism works much less effectively, largely because of high energy taxes,
which account for up to 80% of the retail price of oil in a number of member states (see
Euro Themes: Assessing the impact of lower oil prices and the euro on growth, inflation and
ECB policy, 17 October 2014).
FIGURE 29
Asia-Pacific consumes more than one-third of global oil output
Middle East
9%
S. & Cent.
America
8%
Mn barrels
per day
Africa
4%
Asia Pacific
33%
Europe &
Eurasia
20%
North
America
26%
Source: BP Statistical Review of World Energy 2014, Barclays Research
24 February 2015
FIGURE 30
Asia-Pacific is by far the largest net importer of oil
2.5
2.0
1.5
1.0
0.5
0.0
-0.5
-1.0
-1.5
-2.0
-2.5
Asia
Pacific
Africa
Middle
East
41
Global trade could rise notably on the back of lower fuel costs
Fuel is still a dominant factor in
trade costs; thus lower prices
should boost volumes
Previous research12 found that the elasticity of trade with respect to the freight cost factor
was -313, suggesting that the pass-through of lower oil could substantially boost global
trade. UNCTAD calculated the elasticity of transport costs with respect to fuel prices and
came up with a rule of thumb of 0.4. Assuming that the effect of a change in oil prices on
trade is only via the freight channel, the elasticity of trade with respect to fuel costs rises to
0.4*(-3)= -1.2. Although overall trade costs have fallen in recent years, the drop has not
been significant enough to dismiss them. One main reason is the dominance of the fuel
component in the overall costs. As such, a large enough drop in oil prices, if sustained,
could mean a significant reduction in transportation costs, which in turn should boost trade
volumes.
Equities have rallied after oil troughs, reflecting lagged impacts on growth
From the June peak in oil to the recent trough, the S&P 500 rose 2%. The flattish
performance is in line with the mixed performance of equities during past oil selloffs. Of the
14 episodes in which oil fell by more than 30% (including the current one), the S&P 500 was
down by more than 20% four times, was flat four times, and up by more than 10% six times.
The disparate performance likely reflects different demand and supply shocks in each case.
The S&P 500 has rallied by an average of 12% in the year after an oil trough (median 21%).
The effect of lower oil on economic growth has been lagged by 2-3 quarters following the
initial price shock (see Oil and growth mind the lag and keep the faith, January 22, 2015).
The positive effect on growth also tends to get priced into equities with a lag. After oil troughs,
the S&P 500 has rallied by an average 12% over the next year (Figure 32). The only time
equities were lower was when the tech bubble burst (2000, 2001). Adjusting for negative and
positive outliers, the median S&P return in the year after a trough is a strong 21%.
Performance in the second year is still positive, on average (7%), but more mixed, suggesting
that the positive effects are priced in relatively quickly before other factors take over.
FIGURE 31
Past oil price declines came with slower global growth and a
lagged growth rebound
1.2
1.0
FIGURE 32
Equities have rallied notably after oil troughs
120
115
0.8
110
0.6
105
0.4
0.2
100
0.0
95
-0.2
-0.4
90
-8 -6 -4 -2 0
Average
8 10 12 14 16 18 20 22 24
Current
12
See: http://economics.ouls.ox.ac.uk/14816/1/paper488.pdf
The rationale behind such a large number is the very wide definition of costs, which includes infrastructure of
transit, trade facilitations, policies and logistics, technology, distance and geography (eg, landlocked countries were
found to trade 30-60% less than coastal countries with otherwise similar characteristics).
13
24 February 2015
42
Equity sector and industry relative performance has been 73% correlated to oil betas for the
US and Europe since the June peak in oil (see Investor Intel: Cross-asset effects of lower oil, 4
February 2015). Beneficiaries of lower oil are significantly outperforming and we find that
the outperformance tends to persist. Our equity strategist points out that the worlds 50
largest consumer discretionary and transportation stocks have seen no improvement in
earnings or share prices since July, yet they benefit significantly from lower oil (see Crude
Calculations, January 29, 2015). Importantly, valuations do not reflect the potential upside of
higher consumption, either, particularly the consumer discretionary sector.
Healthcare, consumer and industrials have outperformed after oil troughs. Using FamaFrench data, we analyze how sectors have performed during and after selloffs. Consumer
non-durables, retail and healthcare have outperformed by more than 5% during oil selloffs;
consumer non-durables flattish performance has lagged the typical pattern (Figure 34).
Consumer durables and manufacturing have underperformed the market since June, when
relative performance was historically in line. After oil bottoms, healthcare, chemicals,
manufacturing and consumer non-durables and durables have historically outperformed the
market (Figure 35). Retail, banks and tech have typically performed in line with the market
after oil troughs, while utilities and telecoms have underperformed.
Valuations of sector beneficiaries are not pricing the upside, particularly consumer
discretionary. Historically, the relative valuation of the consumer discretionary sector has
inversely tracked oil prices, with yoy declines in oil typically coinciding with a rise in the relative
forward P/E multiple. However, current consumer discretionary relative multiples are near
historical lows, while the fall in oil points to a relative valuation that is about 20% higher
(Figure 36). Similarly, we find that staples, healthcare and materials multiples inversely track
oil prices; their relative valuations are in line with or below historical averages, suggesting the
upside potential from lower oil is also not fully priced.
FIGURE 33
S&P 500 sector and industry performance have been closely correlated to respective oil betas
Beneficiaries outperforming
0.25
0.20
0.15
0.10
0.05
0.00
-0.05
-0.10
-0.15
-0.20
-0.25
25
20
15
10
5
0
-5
-10
-15
-20
-25
Retailing
Banks
HC Equip
Pharma
Insurance
Real Estate
Cons Durable
Staples Retail
Transportation
Household
Divers Financials
Software
Capital Goods
Food Bev,Tobacco
Media
Semi's
Consumer Services
Telecom
Utilities
Tech Hardware
Materials
Energy
Financials
Health Care
Staples
Discretionary
Industrials
Telecom
Info Tech
Utilities
Energy
Materials
24 February 2015
43
FIGURE 35
Healthcare, consumer and industrials have outperformed after
oil troughs
%
Median
Utils
Telcm
Hlth
Enrgy
BusEq
Utils
Chems
Durbl
Manuf
Money
Telcm
Hlth
Shops
NoDur
-25
BusEq
-20
Enrgy
-15
Money
-5
-10
Shops
Manuf
Durbl
10
8
6
4
2
0
-2
-4
-6
-8
-10
10
NoDur
% rel perf
Chems
FIGURE 34
The decline in energy stocks has been greater than average
Average
FIGURE 36
The decline in oil prices points to a rise in consumer discretionary valuations
WTI yoy
Rel P/E
Feb-14
Dec-14
Jun-12
Apr-13
Aug-11
Oct-10
Dec-09
Feb-09
Jun-07
Apr-08
Aug-06
Oct-05
Dec-04
Feb-04
Jun-02
Apr-03
Oct-00
Aug-01
Feb-99
Dec-99
Apr-98
Jun-97
Aug-96
35%
30%
25%
20%
15%
10%
5%
0%
-5%
-10%
-15%
Oct-95
-60%
-40%
-20%
0%
20%
40%
60%
80%
100%
120%
140%
Cons Disc
24 February 2015
44
Asian equities have further to run amid lower oil. Of the main global regions, Asia ex Japan
has the largest (negative) sensitivity to oil prices, such that lower oil should lead to
outperformance. Asian equities have held up relatively well in the selloff, but in-line
performance since June versus outperformance by many other beneficiaries of lower oil
suggests Asian equities have further to run. Indeed, the relative performance of Asia ex Japan
versus EM has inversely tracked oil prices and the recent collapse in oil points to further
outperformance (Figure 39). From an economic perspective, as noted above, Asia should
disproportionately benefit from sustained lower oil prices. Additionally, lower fuel prices could
lead to a pickup in demand for Asian goods if world trade rises. P/E valuations for Asia that
are essentially in line with EM overall suggest that the relative benefits are not fully priced,
particularly given that Asia has traded at a premium to the rest of DM historically. Finally, as
discussed later, many Asian currencies have not priced in the positive terms of trade shock
from lower oil, so currency could also be a tailwind.
FIGURE 37
Country performance has tracked sensitivities to oil
30
20
10
0
-10
-20
Russia
Norway
Canada
Brazil
Colombia
South Africa
Poland
Thailand
United Kingdom
Hong Kong
AC Europe
India
Germany
France
USA
Singapore
Indonesia
Taiwan
Europe ex UK
Turkey
Australia
EMU
Malaysia
Japan
Switzerland
Sweden
Netherlands
China
Italy
Chile
Mexico
Korea
All China
Spain
EM EMEA
EM Latin America
EM (Emerging
AC World
The World Index
EAFE
AC Asia ex JP
0.25
0.20
0.15
0.10
0.05
0.00
-0.05
-0.10
-0.15
-0.20
FIGURE 38
On average, EM and oil exporting countries have
outperformed after oil troughs
%
FIGURE 39
Asian equities should outperform EM amid lower oil
30
0.75
25
0.70
20
0.65
15
0.60
10
0.80
0
10
20
30
0.55
40
0.50
0.45
-10
0.40
24 February 2015
50
60
0.35
Apr-15
Mar-14
Jan-12
Feb-13
Dec-10
Oct-08
Nov-09
Sep-07
Jul-05
Aug-06
May-
Apr-02
Mar-01
Jun-04
70
0.30
Jan-99
Russia
Korea
Brazil
South Africa
Mexico
Norway
Hong Kong
Singapore
Taiwan
Asia ex JP
India
Malaysia
Australia
Sweden
France
Canada
US
Switzerland
United Kingdom
Spain
Germany
Indonesia
Italy
Japan
-5
Feb-00
45
The magnitude of the oil and commodity boom from the early 2000s to the peak in 2011 was
extraordinary and led to a massive build-up in trade imbalances. Oil was essentially the last
commodity standing, and the collapse since June shocked markets. Sustained lower oil prices
are likely to have a profound effect on current accounts and currencies for some time. Thus,
we assess the sensitivity of countries external balances to a fall in oil prices. The decline in oil
is largely attributable to US tight oil supply; the resultant narrowing of the US petroleum deficit
should reduce the impact of oil price moves on the dollar. Finally, while currencies of oil
exporters have largely adjusted, the currencies that benefit from the oil terms of trade shock
have further to run, in our view.
Oil exporters depend much more on energy exports than importers are affected by them, so
the shock to exporters trade balances will be felt more acutely. Oil accounts for 64% of Gulf
region exports, 70% of Russian exports, and 94% of Venezuelan exports. In contrast,
petroleum products account for only 10% of total US imports and 16% in China. To quantify
the effect of oil price declines of different magnitudes on countries external balances, we
estimate the change in the value of the 2013 trade balance that would arise from a decline in
the average price of oil as it stood in 2013 ($108/b) to $40/b, $60/b and $80/b. This simple
analysis assumes perfect inelasticity of trade, which slightly overstates the response, and
should therefore be used to gauge mainly short-term effects.
The results are shown in Figure 40. As the largest energy importer in the region (in % GDP
terms), Singapore stands out with a net 2% increase in its trade balance as a result of a $20 oil
price decline. However, given its already large surplus, the increment is relatively marginal.
Another notable beneficiary is India, which enjoys a nearly 20% improvement in its trade
deficit (1.5% increase from a deficit of 8.1% in 2013) from each $20 price drop. On the other
side of the chart, the most significant negative impact would be incurred by Saudi Arabia,
which should take a hit of roughly 8%, a 30% reduction in its trade surplus. Russia would
experience a shock similar in magnitude (a 2.6% decline from a 9% surplus), but is left with a
much smaller cushion from external balances. Oil exporters Canada and Mexico would
experience a relatively muted reduction in trade balance of only 0.7% and 0.8%, respectively,
but those come off a deficit of 0.9% in 2013 for both, so not marginal.
FIGURE 40
Current account impacts of changes in oil prices
2013
80
60
40
$20 change in oil price
35%
25%
15%
5%
10%
8%
6%
4%
2%
0%
-2%
-4%
-6%
-8%
-10%
-5%
-15%
-25%
UAE
Saudi Arabia
Singapore
Nigeria
Iran
Russia
China
Brazil
Mexico
Canada
Japan
United States
United Kingdom
India
-35%
24 February 2015
46
The beta of the trade-weighted US dollar to oil prices has historically tracked the US petroleum
trade balance (Figure 41). In the early 2000s, as the US began importing more petroleum
products on net and oil prices began to rise, the petroleum deficit burgeoned. The beta of the
US dollar to oil prices became much more negative, with higher oil coinciding with a lower
dollar. Since US crude production began to surge in 2012, the petroleum deficit has continued
to narrow. The collapse in crude oil prices will further narrow the deficit. Accordingly, the beta
of the dollar to oil price moves has become much less negative, even turning briefly positive in
2014. As noted earlier, however, causality between the dollar and oil runs both ways.
FIGURE 41
The beta of the dollar to oil has tracked US net petroleum imports
Beta
$bn
0.05
5
0
-5
-10
-15
-20
-25
-30
-35
-40
-45
0.00
-0.05
-0.10
-0.15
USD beta to oil fell
as oil deficit grew
-0.20
Feb-15
Feb-13
Feb-11
Feb-09
Feb-07
Feb-05
Feb-03
Feb-01
Feb-99
Feb-97
Feb-95
Feb-93
Feb-91
Feb-89
-0.25
Click here to view an interactive Chart of the beta of Oil vs USD REER
Source: Bloomberg, Haver, Barclays Research
24 February 2015
Oil has been a primary driver of terms of trade shifts and thus currency movements. Figure
42 shows the moves in our measure of terms of trade, which is calculated as the ratio of
each countrys commodities export and import price indices. A number of oil producing
countries enjoyed extremely large increases in their terms of trade (RUB and NOK) as did
those producing base metals (AUD, IDR and CLP) and precious metals (ZAR). Terms of
trade movements explain about 20% of the cross-sectional variation in real exchange rates
since 2002-11, when oil/commodity prices increased. This makes sense given that terms of
trade is a key driver of a currencys medium-term value. Indeed, our behavioral equilibrium
of exchange rates (BEER) model includes terms of trade as one of four drivers of a
currencys value alongside relative productivity, a countrys net foreign asset position and
the relative level of government spending (for details, see Currency valuation from a macro
perspective, 14 June 2011). Only about a third of the terms of trade shifts of 2002-11 have
been unwound, implying that real exchange rates are unlikely to revert to 2002 levels given
that fair values have changed. This raises the question: how have currencies evolved relative
to fair value?
47
FIGURE 42
Only about 1/3 of the 2002-11 commodity boom terms-of-trade impact has been unwound
60
50
% terms of trade
change
40
30
20
10
0
-10
-20
RUB
NOK
ZAR
CLP
AUD
IDR
BRL
CAD
NZD
MYR
MXN
DKK
ILS
SEK
HKD
GBP
PLN
CZK
SGD
EUR
HUF
CHF
CNY
RON
THB
PHP
USD
TRY
KRW
JPY
INR
-30
2002-2011
2011-Now
Countries where terms of trade improved as a result of the commodity price run-up largely
went from being cheap in 2002 to expensive in 2011 and cheapened recently (and vice
versa). Figures 43 and 44 show how the gap between the currency level and its BEER fair
value has changed for the countries that were the big winners and losers from the terms of
trade shifts (we look at the top and bottom seven countries). By and large, the terms of
trade winners went from cheap to expensive in 2011 and have erased some of this richness
more recently. The NOK is a notable exception that can be ascribed to the offshore
petroleum fund that prevents oil revenues from coming onshore unless the government is
running a non-oil fiscal deficit. For the terms of trade losers, the opposite pattern holds.
Effective exchange rates for a basket of terms of trade winners and losers do not suggest
significant misvaluations at present, despite large currency movements. The REER and BEER
of a GDP-weighted basket of commodity exporting countries have cheapened recently, but
it is not particularly cheap. Nor are commodity importers at particularly rich levels.
Oil exporting currencies have adjusted notably, oil importing currencies have not. We
believe currencies are sensitive to oil prices via three major channels: 1) the effect on
external balances; 2) the effect on economic growth; and 3) the effect on monetary policy.
Figure 45 summarizes the exposures of different currencies through these channels (see
Global FX Quarterly: Oil matters, November 20 2014 for details). Baskets of the top and
bottom seven currencies based on this oil sensitivity score are shown in Figures 46 and 47.
FIGURE 43
Terms of trade winners - % misvaluation vs BEER fair value
FIGURE 44
Terms of trade losers - % misvaluation vs BEER fair value
40
30
30
20
20
10
10
-10
-10
-20
-20
-30
-40
-30
RUB
NOK
2002
24 February 2015
ZAR
CLP
mid 2011
AUD
IDR
BRL
today
TRY
JPY
2002
INR
USD
mid 2011
KRW
ILS
THB
today
48
The results are striking: whereas currencies of oil exporting countries have adjusted lower
and their overvaluations have declined, currencies of oil importing currencies have not yet
done so.
Among oil-exporting countries, currency movements since the end of Q3 2014 have been
largely consistent with the deterioration in their terms of trade fundamentals stemming
from oil/commodity price declines (Figure 48). The depreciation has been especially sharp
for the RUB but also for CAD, MXN, MYR and DKK on a relative basis. We find it interesting
that the terms of trade boost among oil-importing countries has so far been largely ignored
(most notably Asia FX, including INR, KRW, etc). The market seems to have settled the
debate over whether lower oil/commodities intensify disinflationary forces or support
growth through consumption in favor of the former. In our view, the terms of trade boosts
are likely to yield faster growth, rising inflation, tighter monetary policy and stronger
currencies on a relative basis in the medium-term.
FIGURE 45
Currency rankings based on an oil sensitivity score
15
10
Oil sensitivity
score
5
0
-5
-10
NOK
CAD
MYR
RUB
MXN
RON
DKK
ARS
COP
AUD
CHF
CLP
USD
IDR
BRL
CNY
GBP
SEK
NZD
EUR
HUF
CZK
ZAR
ILS
PLN
HKD
INR
TWD
TRY
PHP
JPY
THB
KRW
SGD
-15
Score vs.USD
Source: Barclays Research
FIGURE 46
GDP-weighted misalignment between REER and Barclays
BEER model
GDP-weighted misalignment between real effective FX
rate and Barclays BEER model
20
15
FIGURE 47
GDP-weighted misalignment between REER and its 10y
average
25
20
15
10
10
-5
-5
-10
-10
-15
-15
01
03
05
07
Oil winners
24 February 2015
09
11
Oil losers
13
01
03
05
Oil winners
07
09
11
13
Oil losers
49
FIGURE 48
REER move versus the terms of trade shift since Q3 2014
REER move (%)
20%
ARS
10%
0%
-10%
NOK
-20%
y = 0.2781x + 0.0131
R = 0.0172
DKK
MYR
MXN
CAD
COP
y = 1.3714x + 0.0255
R = 0.6203
-30%
-40%
-50%
-25%
RUB
-20%
-15%
-10%
-5%
0%
5%
10%
FIGURE 50
Volatility tends to subside once oil bottoms
-3
27
Oil
24 February 2015
30yr
Dollar
Dec-14
Dec-12
Dec-10
Dec-08
-3
Dec-06
Dec-04
-2
Dec-02
Dec-00
-1
Dec-98
23
Dec-96
Dec-94
Dec-92
25
Dec-90
-1
Dec-88
Dec-86
-2
Dec-84
29
21
19
17
-12
-9
-6
-3
12
15
18
21
24
Average
Click here to view an interactive Chart of the VIX. Source: Haver, Barclays Research
50
35%
FIGURE 52
The weight of energy in equity indices
30%
25%
20%
15%
10%
5%
Euro HY Corp
Russell 2000*
US HY Loans
Euro IG Corp
S&P 500
US IG Corp
US HY Corp
EM Local Sov
EM Corp
EM Sov
0%
50%
45%
40%
35%
30%
25%
20%
15%
10%
5%
0%
Russia
Italy
India
UK
EM
Brazil
US
China
DM
World ex US
France
SA
Europe
Spain
Taiwan
S Korea
Japan
Germany
HK
FIGURE 51
The weight of energy is highest in EM hard currency debt
US HY, posing notable default risk, and thus spillover risk, to investors (Figure 51). The
energy weight in global equities is much smaller, 7%, with Russia having the largest
exposure by far (Figure 52).
Volatility tends to subside after oil finds a bottom, however. Using the average of the VIX
around past oil selloffs, we find that volatility tends to peak in the weeks right around the
trough in oil, but returns to normal levels once oil bottoms (Figure 50).
Oil-exporting countries are the clear losers from a decline in oil prices
Russia stands out as having
taken the largest hit from the
oil shock so far
Apart from the real activity implications, a large shock to these energy-dependent countries
can pose financial stability risks. However, a closer look suggests the risks are lower than in
the past.
Russia is the biggest sufferer, with the ruble taking the largest hit. The RUB selloff is
causing large inflationary pressures, though Russia has the ability to at least partially
offset the negative growth effect via imports, which tend to contract more than
consumption. Nevertheless, there is a high risk of financial instability and overdepreciation because of the spiral effect on business confidence, productivity growth
and investment growth (see Russia quarterly: Problems accumulate, 5 December 2014).
Russias Big Four companies, Gazprom, Rosneft, Lukoil and Novatek essentially have
cash/short term debt ratios in excess of 100%; this, coupled with their low-cost
position and the flexible Russian oil taxation system, suggests that a default on their
hard currency debt in coming months is unlikely (see Russia Big 4 Credit & Equity
Cashflows: Surviving the Time of Troubles, 15 December 2014).
Venezuela, with 94% of exports from oil, is highly exposed, but default risk is priced.
The hard currency regime means that the oil price decline is hardly offset by
devaluation, and ongoing wild fiscal spending is expected to continue, maintaining a
deep deficit of roughly 16% of GDP. (The Emerging Markets Quarterly: Content with
carry, concerned with commodities, 5 December 2014).
Middle Eastern oil producers have accumulated significant FX reserves. The IMF
estimates a $98 oil price is needed to balance Saudi Arabias budget, and above $65 to
maintain a CA surplus. However, with FX reserves covering more than 900% of Saudis
external debt, a net asset position of over 100% of GDP and among the worlds lowest
levels of public debt (less than 3% of GDP), the kingdom has ample fiscal space and
plenty of policy tools to accommodate the lower oil price (see The end of OPECs golden
age? 24 November 2014).
24 February 2015
51
Mexico has also been hurt by the oil price drop, but concerns look to be overstated.
The opening of the economy to international trade has intensified the links to the US,
and in turn reduced reliance on oil and gas, which now account for only 7.3% of GDP. In
the past 10-15 years, oil and gas production has shrunk by nearly 25% in real terms and
the exposure of government revenues to oil prices has declined by 60% (see Oil price
collapse; for Mexicos economy, concern looks overstated, 21 January 2015).
The primary loser in Asia is oil-exporting Malaysia, which is expected to take a 0.5%
CA hit for every USD10 drop in oil prices. However, lower fuel subsidy spending should
clear some fiscal policy space.
Market dynamics are also different, providing buffers to the oil shock
Financial markets are more flexible, with fewer currency pegs. Many past crises
stemmed in part from currency pegs that did not allow adjustment to occur through
exchange rates. The 50% decline in the Russian ruble since June has meant that the
resultant increase in import prices will lead the adjustment, while FX reserves were not
exhausted to defend an overvalued currency. Importantly, Russian equities held up
reasonably well in local currency terms. The currencies of other oil exporters, including
Colombia, Mexico, Canada and Norway, have also depreciated considerably.
FX reserves provide a sizeable cushion for the terms of trade shock from lower oil.
The reserves of major oil exporters mushroomed when oil was rising, to almost $1.5trn
(Saudi Arabia, $718bn; Russia, $316bn; UAE, $315bn; and Qatar, $106bn).
Maturity profile in HY and much of EM provides some breathing room, for now. A
number of US E&P companies will find it difficult to keep operating even at the current
level of oil prices, raising the potential for defaults. However, most HY energy companies
have some breathing room until 2018-19, when debt maturities will start to be
significant (Figure 53). Within EM, LatAM and EMEA debt maturities should provide
some time to ride out lower oil prospects and the start of Fed hikes, but less so for Asia.
The Fed is not hiking (yet) and other central banks are easing. The recent oil selloff
has led to considerable central bank easing, while the Fed has stayed the course. This
contrasts with the Mexican peso devaluation in 1994 and the crises in Brazil and
Argentina in 1999-2001, when the Fed was hiking. Additionally, the Asian and Russian
crises were set against a backdrop of high US real yields, which is certainly not the case
currently. However, the prospect of sustained lower oil and Fed hikes starting this year
could put further external pressure on at-risk countries and companies.
Investor liquidity remains a buffer, but market liquidity remains a concern. Liquidity
ratios dropped below 4% prior to the euro crisis in 2010, the financial crisis in 2007 and
the bursting of the tech bubble in 2000. Current post-QE cash ratios have remained
above 5%, despite cash yielding essentially zero (Figure 54). Cash levels for bonds and
asset allocation funds are elevated, while equity fund cash holdings are average.
Importantly, US HY bond funds have raised their cash holdings to 5.6% from a low level
24 February 2015
52
FIGURE 53
Debt maturities are not significant for another few years
FIGURE 54
Investor liquidity is higher than other pre-crisis periods
9
%
45,000
40,000
35,000
Financial
crisis
30,000
Euro
crisis
25,000
20,000
15,000
Russia
default
10,000
Tech
bubble
Sep-14
Jan-12
May-13
Sep-10
Jan-08
May-09
Sep-06
May-05
Jan-04
Sep-02
Jan-00
May-01
Sep-98
Jan-96
May-97
Sep-94
May-93
2038
2037
2035
2024
2023
2022
2021
2020
2019
2018
2017
2016
Jan-92
5,000
in June (4.5%), a scenario that has historically pre-dated volatility in HY (eg, 2007, 2000
and 1998). That said, redemption cycles, such as the one during the taper tantrum,
show that markets are much less liquid after a financial crisis in a world of smaller
dealer balance sheets.
FIGURE 56
FX reserve growth and US dollar index changes
120
0.7
100
0.6
80
60
0.5
40
0.4
20
0
0.3
-20
0.2
-40
0.1
-60
-80
0
01
03
05
07
24 February 2015
09
11
13
%y/y (inverted)
-20
0.7
-15
0.6
-10
0.5
-5
0
0.4
0.3
10
0.2
15
0.1
20
25
0
01
03
05
07
US dollar index (lhs)
09
11
13
FX reserves in SDRs (rhs)
53
US energy sector will be hard hit, but it is smaller than TMT and financials
The oil price decline will be felt globally, but US energy companies stand to be hurt most.
However, to put this into perspective, US energy stocks had a $1.9trn market cap in June 2014
(11% of S&P) and energy high yield companies were about 15% of the index. In comparison,
in 2000, TMT comprised more than $5trn of US market value (44% of market cap) and about
44% of high yield. In 2006-08, US financials were $3trn in market cap (22%) with $15trn of
mortgages outstanding before the housing bust. In terms of the US economy, the financial
sector employed 8.4mn people in 2007 and lost 800k jobs, while TMT employed 3.7mn and
600k jobs were lost. Energy and related sectors in the US employ fewer than 1mn people.
Finally, oil-related assets and debt typically do not have the type of collateral function that can
cause larger systemic effects, as was the case with mortgage debt.
US high yield spreads reflect lower oil prices, but maybe not low enough
(This is excerpted from US HY Energy 2015 Outlook, 14 January 2015, and Energy
Effects, 9 January 2015)
If this oil cycle is lower/longer, as we expect, HY energy spreads could widen further,
based on our sensitivity analysis of high yield OAS to oil prices (Figure 58). Assuming
$50/bbl WTI, E&P debt/EBITDA could jump to 4.3x from 2.8x and about 1/3 of the credits
we model will have leverage above 5x. In 2016, even under higher assumed WTI prices of
$60/bbl, ~40% of the names under coverage will likely have debt/EBITDA over 5x as
weaker hedges more than offset a rise in assumed oil prices. Spreads could approach
1200bp in such a scenario. We model a 4.5% default rate in such a scenario, taking into
account hedges that should help insulate certain credits in 2015, but stopping short of the
full bottom-up analysis that will be necessary for a more comprehensive view on potential
default losses if oil prices are at these levels a year from now.
FIGURE 57
EM and US energy spreads have widened with the fall in oil
*OAS,rebased to
100 on Jan 14
1,400
Oil price
210
190
40
170
60
1,300
1,200
1,100
1,000
900
150
80
130
100
110
90
120
70
50
Mar-12
FIGURE 58
High yield OAS sensitivity to oil prices
140
Sep-12
Mar-13
Sep-13
24 February 2015
Mar-14
Sep-14
current
800
700
600
500
400
300
$45
$50
US HY
$55
$60 $65
HY Energy
$70
54
FIGURE 59
Lower energy demand growth should also be a headwind
GDP %
FIGURE 60
Energy P/B valuations near the lows
Rel price
3.0%
0.6
2.8%
0.5
2.6%
P/B
3.9
Relative
1.3
1.2
3.4
1.1
0.4
2.4%
2.2%
2.0%
1.8%
0.3
2.9
0.2
2.4
1.0
0.9
0.8
0.1
24 February 2015
0.6
Feb-14
Feb-12
Feb-10
Feb-08
Feb-06
Feb-04
Feb-02
Feb-00
Feb-98
0.5
Feb-96
1.4
Feb-94
Dec-13
Oct-10
May-12
Jan-06
Mar-09
0.7
1.9
Feb-92
Aug-07
Jun-04
Apr-01
Nov-02
Sep-99
Jul-96
Feb-98
Dec-94
Oct-91
May-93
Mar-90
1.4%
Feb-90
1.6%
55
24 February 2015
56
Bibliography
External Research
Forecasting Crude Oil Spot Price Using OECD Petroleum Inventory Levels. Ye, Zyren, Shore,
International Advances in Economic Research, 2002
Explaining the so-called price premium in oil markets, Merino, Ortiz, OPEC Review, 2005
Exchange rates and commodity prices: measuring causality at multiple horizons. Galbraith,
Dufour, Zhang. CIRANO working paper 2013-s39. Submitted to Journal of Empirical
Finance, 2014.
What Drives the Oil-Dollar Correlation?, Christian Grisse Federal Reserve Bank of New York
Working Paper, 2010
Can Oil Prices Forecast Exchange Rates?, Ferraro, Rogoff, Rossi, Federal Reserve Bank of
Philadelphia. Working Paper No. 11-34, 2011
Forecasting the Price of Oil, Alquist, Killian, Vigfusson, Board of Governors of the Federal
Reserve System, International Finance Discussion Papers, Number 1022, 2011
The macroeconomic effects of oil shocks: Why are the 2000s so different from the 1970s?,
Olivier J. Blanchard and Jordi Gali, NBER working paper No. 13368, 2007
Seven question about the recent oil price slump, Olivier Blanchard and Rabah Arezki, IMF
Direct blog, December 2014
Transportation costs and international trade, Alberto Behar and Anthony J. Venables,
Handbook of Transport Economics, 2010
Barclays Research
Economics
Oil and growth mind the lag and keep the faith, Christian Keller, 22 January 2015
Easy money is not easy for all EM, Piero Ghezzi, 23 November 2010
The Crude Reality, Tal Shapsa, 8 April 2011
Consumer windfall ahead, Dean Maki, 17 October 2014
Euro Themes: Assessing the impact of lower oil prices and the euro on growth, inflation and
ECB policy, Thomas Harjes, Aroop Chatterjee, 17 October 2014
Sizing up recent shocks to the global economy, Apolline Menut, 24 October 2014
Asia Themes: Its oil good for Asia, David Fernandez, 2 December 2014
Lower oil ignites fresh structural reforms, David Fernandez, 5 December, 2014
New Oil Forecasts, Macro Implications, David Fernandez, 28 January 2015
Lower oil- a boon for India, Siddhartha Sanyal, 29 January 2015
Commodities
The Blue Drum: Everybody Hurts, Michael Cohen, 28 January 2015
Saudi Arabia: From swing producer to flexible petroleum supplier, Miswin Mahesh, 19
February 2015
The end of OPECs golden age? Michael Cohen, Alia Moubayed, 24 November 2014
Middle East Geopolitics The ISIS threat: a new era of MENA risk, Alia Moubayed, 17
September 2014
Credit
US HY Energy 2015 Outlook, Gary Stromberg, 14 January 2015
HY E&P Cost Structures and Sensitivities to Lower Oil Prices, Gary Stromberg, 17 October 2014
OPECs Black Friday: Implications for High Yield Energy Companies, Gary Stromberg,
1 December 2014
24 February 2015
57
E&P: Some More Stress Testing Looking at $50/bbl, Harry Mateer, 15 December 2014
Asking the Right Questions: E&P Spreads Are Pricing in Lower Oil; Remain Overweight,
Harry Mateer, 18 December 2014
Credit Strategy
Energy Effects, Bradley Rogoff, 9 January 2015
The Solar Vortex, Ryan Preclaw, 23 May 2014
The Low Oil-Price Playbook, Ryan Preclaw, 16 January 2015
There Will Be Oil: A Look at Past Oil Moves and Impact on E&P Spreads, Ryan Preclaw,
17 October 2014
Asia Credit Strategy Oil price drop: Positive macro, mixed micro, Krishna Hegde, 2
December 2014
Asia Credit Alpha: Oil Bounce Based Ideas, Krishna Hegde, 6 February 2015
Cross Asset
Investor Intel: Cross-asset effects of lower oil, Keith Parker, 4 February 2015
Navigating the new EM landscape: Where to find the best returns, Equity Gilt Study, Michael
Gavin, 20 February 2011
Future Powertrain: Premium pain?, Kristina Church, Brian A. Johnson, Paul Hwang, 3 July 2014
Emerging Markets
The Emerging Markets Quarterly: Content with carry, concerned with commodities,
Christian Keller, 5 December 2014
Russia quarterly: Problems accumulate, Eldar Vakhitov, Durukal Gun, Andreas Kolbe, 5
December 2014
Russia Big 4 Credit & Equity Cashflows: Surviving the Time of Troubles, Matthew Thomas,
15 December 2014
Mexico: Oil Price Collapse for Mexicos economy, concern looks overstated, Marco Oviedo,
21 January 2015
Commodities and LatAm Corporates: Here is the Fall, Aziz Sunderji, 16 October 2014
GCC Corporates: Assessing the fundamental impact of lower oil prices, Bayina Bashtaeva,
28 January 2015
South African macroeconomic forecast: Adjusting for lower expected oil prices, Peter
Worthington, 2 February 2015
Equities
Re-examining the cost of oil production, Thomas R. Driscoll, 9 January 2015
Energy: Downgrading the Majors due to Unattractive Valuation, Paul Y. Cheng, 14 January 2015
Global Energy - Positioning for a Low Oil Price Environment, Paul Y. Cheng, 14 January 2015
Lower oil prices a net positive, Jonathan Glionna, 6 November 2014
Crude Calculations, Ian Scott, 29 January 2015
Foreign Exchange
FX Mid-Quarter Update: Oil tests monetary policy limits, Jose Wynne, 2 February 2015
Global FX Quarterly: Oil matters, Jose Wynne, 20 November 2014
Currency valuation from a macro perspective, Aroop Chatterjee, 14 June 2011
Oil Low For Longer Transforms Asia Outlook, David Fernandez, 4 February 2015
Rates
Oil slick, Michael Pond, 11 December, 2014
24 February 2015
58
External accounts
Inflation
C/A forecast
(% GDP)
2014
C/A forecast
(% GDP)
2015
% of energy in
CPI basket
g
revenues
(% of
government
revenues)
Fiscal accounts
Energy
subsidies
(% of GDP)
EM net oil exporters (darker colour = greater vulnerability from lower oil price )
Kuwait
53
57
-4.7
-9
33
38.9
21.1
92
4.9
54
Saudi Arabia
46
43
-3.9
-32
64
14.8
5.3
90
8.3
98
Oman
41
43
-3.5
-3
82
8.1
-1.3
86
99
Iraq
40
37
-3.3
-8
94
1.6
-8.2
98
6.0
111
Qatar
36
62
-3.0
-7
56
26.4
17.2
56
3.1
55
Venezuela
30
30
-2.7
-6
77
6.6
1.5
5.0
10.2
600
UAE
25
51
-2.1
-9
64
12.3
6.8
64
5.6
79
Bahrain
20
15
-1.8
-1
67
9.1
2.3
88
7.8
125
Nigeria
13
15
-1.2
-6
77
3.5
-1.6
8.5
66
1.0
113
Russia
14
17
-1.5
-26
72
3.6
6.7
9.0
6.4
51
2.2
95
Colombia
-0.6
-3
-4.3
-5.0
6.8
6.6
0.0
Malaysia
-0.5
-2
48
5.0
3.0
12.1
5.5
25
1.5
Ghana
-0.1
204
-9.6
-7.8
0.3
Mexico
-0.1
-1
-1.7
-1.9
8.9
32
0.8
Argentina
-1
0.0
0.9
0.9
9.9
4.1
Peru
-1
0.0
-5.0
-4.6
5.7
3.2
Kenya
0.6
-7.3
-7.5
3.7
Egypt
0.0
-1.2
-2.3
6.6
EM net oil importers (darker colour = greater benefit from lower oil price )
Brazil
-1
-1
0.1
-3.9
-3.2
8.9
6.1
Indonesia
-2
0.1
-2.8
-1.9
6.0
13.2
21
2.4
China
-2
-3
0.2
23
3.3
3.6
2.1
0.2
Hungary
-3
-6
0.4
3.7
3.8
17.6
0.3
Israel
-3
-4
0.3
2.9
2.9
7.2
3.1
Turkey
-3
-6
0.5
-5.4
-4.3
13.6
17.3
Poland
-4
-3
0.3
-1.2
-0.7
17.8
8.4
Philippines
-4
-5
0.3
4.4
5.4
8.9
12.7
Chile
-5
-5
0.5
-1.6
-2.1
10.5
7.1
S. Africa
-5
-4
0.6
-5.4
-4.1
4.2
5.5
India
-5
-6
0.4
-1.3
-1.2
9.5
1.0
Ukraine
-5
-10
0.7
-3.0
-2.4
9.4
5.8
Morocco
-8
-10
0.6
-4.8
-4.5
3.8
Singapore
-8
-8
1.4
19.4
19.2
2.4
2.0
Korea
-9
-10
0.8
10
5.7
6.2
10.0
2.3
Taiwan
-9
-11
0.7
12.0
12.6
6.7
Thailand
-10
-10
0.9
3.0
4.4
11.4
4.2
0.7
Zambia
0.4
-0.8
2.6
DM net oil importers (darker colour = greater benefit from lower oil price )
US
-2
-2
0.2
28
-2.3
-2.3
9.0
3.1
Euro Area
-3
0.3
33
2.2
2.0
10.8
1.4
UK
-1
-1
0.1
-5.0
-5.1
8.0
3.4
Japan
-3
-5
0.3
14
0.2
0.9
7.7
3.6
Switzerland
-2
-2
0.2
10.1
9.8
6.4
8.0
DM net oil exporters (darker colour = greater vulnerability from lower oil price )
Norway
13
19
-1.1
-6
7.8
2.0
Canada
-0.4
-8
-2.4
-2.3
8.6
7.4
24 February 2015
59
CHAPTER 3
The external backdrop for EM economies has grown tougher since 2011 and will likely
remain so over the next few years. On the domestic front, progress on structural
reforms has been disappointing. But EM economies have evolved since the start of the
boom years in the early 2000s, with many of their macroeconomic and financial
vulnerabilities now reduced.
EM risk premia are again moving toward levels more consistent with the early boom
stages, especially in local currency debt. Our analysis suggests that investors have
so far been compensated for the risks they have taken. However, because of the
negative backdrop, annualized returns in the down period have been far smaller
than annualized gains in the boom.
When we look at EM in the context of a global portfolio, the gap between EM and DM
risk premia is significant. Thus, we think allocations to EM assets make sense even if
asset returns are likely to be much lower than in the boom years.
A changing landscape
The EM investment landscape
has changed dramatically
over the past two decades
EM economies have gone through a remarkable economic transformation over the past two
decades as the crisis-plagued 1990s led to a boom in 2002-11. Three key factors drove that
transformation, in our view. First, China emerged as a global manufacturing power and a
major source of demand for commodities. Second, EM countries reduced their dependence on
external debt finance and the associated currency mismatches. This was reinforced by a
secular downtrend in term interest rates in the developed world, notably the US. Third, most
EM economies achieved macro stability through fiscal responsibility, lower inflation, more
flexible exchange rates and the build-up of external buffers.
These factors reinforced each other, producing a virtuous cycle that led to strong
macroeconomic outcomes and high asset returns. However, the boom years were followed
by a landing phase that began in 2011 with the reversal of the favourable external
backdrop: a slowdown in China, lower commodity prices and a stronger US dollar. As these
positive factors recede, the likelihood of much needed structural reforms to boost potential
growth will likely fall. In that context, the environment for EM asset returns over the next
few years is likely to be a lot more challenging, even if some of the macroeconomic
improvements (notably lower inflation, the shift toward exchange rate flexibility, and more
stable sovereign external debt) are proving to be long-lasting.
The EM investment landscape must be considered in the context of a new reality: a more
challenging external backdrop and the need for structural reform to boost growth. First, we
discuss the bust, boom and landing cycles of the past two decades. Second, we assess the key
challenges of the changing external environment. Third, we highlight the main domestic
challenges and remaining vulnerabilities. Finally, we review how EM risk premia have evolved
and where investment opportunities look most attractive given these new challenges.
24 February 2015
Our assessment is that EM is much more resilient than it was 15-20 years ago, though still
vulnerable to various external macro headwinds. Activity and earnings growth are likely to
remain challenging and so are equity returns. Local bonds have better prospects as real yields
remain high and growth and inflation prospects subdued. However, there remains plenty of
scope for differentiation. In equities, we favour countries that benefit from macro/structural
reforms, lower oil prices and a stronger US economy, notably EM Asia. High carry and
undervalued currencies also make some bond and FX markets attractive.
60
Emerging market economies have gone through three major phases over the past two
decades: the economic and financial fragility of the crisis-plagued 1990s; the economic and
financial boom of 2002-10; and the landing period that started in 2011.
The crises of the 1990s and the early 2000s are well documented in the academic literature.1
The macroeconomic outturns at the turn of the 1990s were dismal: growth was slow by EM
standards, inflation high and indicators of external vulnerability poor (Figure 1).
FIGURE 1
EM macro fundamentals in bust, boom and landing cycles
Domestic factors:
Real GDP growth (% YoY)
Inflation (% YoY)
Government gross debt (% GDP)
Fiscal balance (% GDP)
Current account balance (% GDP)
External debt (% exports)
External debt service (% exports)
1998-2001
2002-2010
2011-2014
3.9
10.6
52.3
-2.1
0.1
147.1
38.1
6.6
6.6
43.8
-1.1
2.7
92.0
29.4
5.1
6.2
39.1
-1.3
1.1
79.6
27.1
Economic performance
improved materially in the
boom cycle (2002-10)
The economic performance of EM changed markedly for the better in the early 2000s on a
combination of macro stability post-EM crises and a benign external environment. For a
detailed discussion of the changes in the macroeconomic and financial landscape and the
resulting resilience to negative shocks, see Navigating the new EM landscape: Where to find
the best returns, Equity Gilt Study 2011.
FIGURE 2
Asset returns* were very high during the boom years, but have been lackluster since 2011
30
Avg. annual
returns (%)
25
20
15
10
5
0
-5
Equity
Credit
2002-2010
FX
2011-2014
The boom years were supported by a virtuous cycle of benign external conditions and
improvements in domestic fundamentals. On the external front, the emergence of China as a
global economic power, the associated boom in commodity prices and lower external
financing costs (notably via falling US interest rates) fuelled the EM growth recoveries.
Meanwhile, macro stability and the build-up of external and fiscal buffers led to significant
improvements in macro fundamentals. These synergies led to strong economic performance
and a reduction in risk premia, which had soared following the EM crises. As a result, EM
experienced very high asset returns (Figure 2).
See for example, Calvo (2005), Emerging Capital Markets in Turmoil: Bad Luck or Bad Policy?; or Roubini, Nouriel
and Brad Setser (2004), Bailouts or Bail-Ins: Responding to Financial Crises in Emerging Markets.
24 February 2015
61
However, in 2011, the benign external environment began to reverse. At the same time,
little was achieved in terms of institutional progress and structural reforms beyond the post
crises adjustments, and EM asset returns have been uninspiring apart from EM sovereign
credit. As a result, economic growth and asset returns deteriorated markedly.
FIGURE 3
Slower China and euro area GDP growth since 2011
% y/y
Boom Years
China
%y/y
18
16
14
12
10
-2
-4
-6
-8
96
97
98
99
00
01
02
03
04
05
06
07
08
09
10
11
12
13
14
Indeed, Felices and Wieladek (2012) find that the link between external factors and fundamentals remains tight
even though macro fundamentals have improved materially.
24 February 2015
62
Boom Years
150
Brent Oil
USD/MT
Copper (RHS)
12000
130
10000
110
8000
90
70
6000
50
4000
30
2000
10
-10
0
96 97 98 99 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14
Stronger US dollar
A weak US dollar was also very favourable for emerging markets during the boom (200210). During those years, a large fraction of total returns in equity and local bond holdings
for US dollar-based investors was the appreciation of EM currencies relative to the US dollar.
Average annual returns for EM currencies, for example, were c.7% against the US dollar
during the boom years, but only -2.4% in 2011-14.
FIGURE 5
UST 5y yields set to edge up
9
Boom Years
US
Germany
8
7
6
5
4
3
2
1
0
96
97
98
99
00
01
02
03
04
05
06
07
08
09
10
11
12
13
14
Click here to view an interactive Barclays Live Chart. Source: Bloomberg, Barclays Research
24 February 2015
63
FIGURE 6
The US dollar is on the rise (USD vs broad basket of currencies including EM)
Index
150
140
130
120
110
100
90
80
96
97
98
99
00
01
02
03
04
05
06
07
08
09
10
11
12
13
14
Click here to view an interactive Barclays Live Chart. Source: Bloomberg, Barclays Research
We combine the five key external headwinds (slower China growth, euro area growth, weaker
commodity prices, higher US yields and a stronger USD) to form a measure of external
supportiveness. The measure is an average of the Z-scores (standard deviation away from the
mean) of each of the external factors. UST yields and the USD are included in the indicator
with a negative sign to reflect their negative effect on EM economies. In the Appendix, we
present the exposure of various EM economies to the five external measures discussed above
as well as an aggregate ranking of EM countries to those factors.
Figure 7 shows our measure of external supportiveness. Three observations stand out. First,
the measure shows very clearly the transitions from bust, boom and landing since the mid1990s. Second, the global financial crisis was a major negative for the asset class but the
external backdrop improved rapidly, largely because of Chinas huge counter-cyclical effort,
easier US monetary policy and a weaker USD. Third, these phases clearly mirror the pattern
of the asset class in the boom and landing phases (Figure 2).
FIGURE 7
A measure of external supportiveness (Higher = more supportive of EM)
Z Score
Boom Years
1.2
0.8
0.4
0.0
-0.4
-0.8
-1.2
96
97
98
99
00
01
02
03
04
05
06
07
08
09
10
11
12
13
14
Source: Haver Analytics, Bloomberg, Barclays Research. Note: We combine the five key external headwinds (slower
China growth, slower euro area growth, weaker commodity prices, higher US yields and a stronger USD) to form a
measure of external supportiveness. The measure is an average of the Z-scores (standard deviation away from the
mean) of each of the external factors.
24 February 2015
64
We expect the external environment (and our indicator of vulnerability) to become even
more challenging and to continue weighing on asset returns. China is undergoing structural
changes and the economy is expected to experience slower GDP growth in the coming
years. We forecast growth to slow to 7% in 2015 and to 6.6% in 2016, from an average of
about 10% in 1995-2007. In the euro area, we expect a very gradual recovery to 1.1% in
2015 and 1.5% in 2016, still below the average in 1996-2010. We see these slowdowns as
structural shifts that will have long-lasting effects, especially on the EM countries that
export goods and services to them. Indeed, potential growth has slowed in the euro area
and other developed economies and it is likely to remain challenged over the next few years
(see Chapter 4, The great destruction).
Crude oil prices in 2015 have recovered from recent lows, but our commodities analysts
expect the price to remain low and think that for consistent gains in commodity prices, global
growth must exceed 4% (see Cross commodity themes and strategy: Plumbing the depths, 10
February 2015). Lower commodity prices affect EM countries in various ways, with the EM
commodity exporters3 that benefited materially from the commodities boom most at risk.
The first three headwinds are structural in nature and, as such, likely to persist for several
years. The remaining two headwinds are more cyclical, but also likely to be long-lasting. First,
we expect US interest rates to rise as the Fed continues to normalize its monetary policy. This
process follows a prolonged period of monetary policy expansion that included quantitative
easing (QE), which led to a compression of market volatility and a search for yield that
benefitted EM assets. EM countries are at risk of renewed capital outflows as this process
reverses. The taper tantrum of May 2013, for instance, showed how destabilizing the change
in US monetary policy can be for EM. Second, the rise of the USD versus EM currencies,
although cyclical, is also likely to persist. USD cycles tend to be long, and this one should be no
exception, as it will likely be underpinned by higher Fed policy rates.
Domestic EM fundamentals improved markedly during the boom years. Figure 1 summarises
some of the positive changes in these economies in 2002-10. Inflation fell to single digits as
central bank independence and inflation-targeting began to play a bigger role. Governments
took action to ensure debt sustainability and enhance their external buffers. As a result,
government debt and fiscal deficits fell and external vulnerability indicators improved
markedly. Another crucial development was the reduction in foreign-currency-denominated
debt and the transition to debt issuance in local currency. This was important because localcurrency depreciation had increased debt burdens and hurt GDP growth. According to the
IMFs Global Financial Stability Report (2006), the share of local-currency-denominated
bonds in EMs marketable sovereign debt rose by 9pp between 1996 and 2004, reflecting
mainly increased local currency issuance. Naturally, the improved fundamentals vary by
country, but broad improvement was visible across EM (country-specific macro
fundamentals in the bust, boom and landing cycles can be found in the Appendix).
Another important area of improvement was the transition from fixed or heavily managed
exchange rate regimes to more flexible ones. Flexible exchange rates have been very
effective shock absorbers (eg, during the global financial crisis and more recently for
countries under stress, such as Russia). Using a classification of exchange rate
arrangements conducted by Reinhart and Rogoff (2004), we construct an indicator of FX
liberalisation for a group of 16 EM economies (Figure 8). This indicator takes a value of 1 if
24 February 2015
These include, for example, Russia, Venezuela, Brazil, South Africa, Mexico, Chile, Peru, Colombia and Malaysia.
65
the country has a hard peg and 15 if the FX market is fully flexible. The values in between
take into account other shades of FX regimes. The chart also shows that FX liberalisation
has improved markedly since the early 2000s.
FIGURE 8
FX liberalization has improved materially since the early 2000s (higher = more flexible)
11
Indicator of FX
liberalisation
10
8
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
16 country average
Source: Barclays Research, Reinhart and Rogoff (2004)
The adjustment toward more stable and resilient macro frameworks led to improved
confidence in the asset class among market participants and credit ratings agencies. The
latter can be seen in the evolution of credit ratings on sovereign and corporate external
debt. Figure 9 shows that in the early 2000s, only around 30% of the asset class was rated
investment grade; this had risen to more than 70% by end-2014. Note also, however, that
this trend is starting to turn. For example, Russias sovereign external debt has been
downgraded by several ratings agencies and it lost its investment grade (IG) status in
February 2015.
FIGURE 9
EM external debt is now mostly investment grade (weighted using Barclays EM sovereign
index weights)
100%
90%
70%
60%
50%
40%
Index weight
80%
30%
20%
10%
IG rated
Ba
2014
2013
2012
2011
2010
2009
2008
2007
2006
2005
2004
2003
2002
2001
2000
0%
NR
24 February 2015
Despite these major improvements, at least three areas of domestic vulnerability remain of
concern. First, structural reforms needed to boost potential output have been disappointing.
Second, activity growth has slowed markedly and has disappointed analysts expectations.
Finally, although sovereign external debt as a share of GDP has been reduced and remains
contained, corporate external debt has risen rapidly in recent years.
66
FIGURE 10
Ratings have improved more than governance (rating and
governance indicators weighted by GDP)
-0.10
FIGURE 11
EM governance remains poorer than in DM (governance
indicators weighted by GDP)
8
-0.12
-0.14
EM ratings ex-China
1.5
9
10
-0.16
-0.18
11
-0.20
12
-0.22
13
-0.24
index (-2.5
min, 2.5
max)
1.0
DM governance
EM governance
0.5
0.0
14
2013
2012
2011
2010
2009
2008
2007
2006
2005
2004
2003
2001
2002
2000
1999
1998
-0.5
1997
2013
2012
2011
2010
2009
2008
2007
2006
2005
2004
2003
2002
2001
2000
1999
15
1998
-0.28
1996
-0.26
Source: Barclays Research, S&P, Moodys and World Bank. DM countries include:
Australia, Canada, France, Germany, Italy, Japan, Netherlands, Norway, Spain,
Swede, Switzerland, United Kingdom, and United States. EM countries include:
Brazil, Chile, Colombia, Indonesia, Malaysia, Mexico, Peru, Poland, Russia, South
Africa, Turkey, Philippines, Argentina, Venezuela
24 February 2015
67
FIGURE 12
No clear improvement in governance metrics of EMs: Z-score of ratings and governance (as of
2013)
z-scores
Governance (z-score)
TUR
COL
PER
PHL
CHL
IDN
BRA
CHN
MEX
RUS
POL
MYS
ZAF
VEN
ARG
2.5
2.0
1.5
1.0
0.5
0.0
-0.5
-1.0
-1.5
-2.0
-2.5
Ratings (z-score)
Source: Barclays Research, S&P, Moodys and World Bank. Z-scores calculated using a sample from 1998-2003.
The lack of progress on structural reforms means that potential output in EM economies is
likely to be constrained. This makes GDP growth more vulnerable to external headwinds,
especially in those countries that depend heavily on external demand and commodity
exports. The so-called BRICs, once the flagship of the EM boom, have slowed significantly.
In Brazil, GDP growth has slowed from close to 8% to almost 0% in just five years
(Figure 13). Such outturns have consistently surprised the consensus to the downside
(Figure 14). Disappointing growth has been the main driver of lower asset returns,
especially in equity markets, where returns are tightly linked to expectations of earnings
growth, which in turn depend on prospects for activity growth.
FIGURE 13
GDP growth has slowed rapidly in the BRICs*
12
Real GDP
growth YoY
2010
2011
2012
2013
FIGURE 14
BRICs* growth has surprised the consensus to the downside
in recent years (GDP growth year-ahead-forecast)
2014
ppt
2010
2011
2012
2013
2014
10
-1
-2
-3
-4
Brazil
India
Russia
China
Source: Barclays Research and Bloomberg. BRICs are Brazil, Russia, India and
China.
24 February 2015
Brazil
India
Russia
China
68
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
*Corporate external debt is total external debt minus government and central bank debt. It therefore includes the debt of deposit-taking corporations. Source:
World Bank Quarterly external debt data, Barclays Research
The charts below show that the governments of Hungary, Poland, Mexico, Malaysia, and South Africa have large external debt
(more than 15% of GDP). Note, however, that not all that debt is in foreign currency. Indeed, a large fraction of South Africas
external debt is in local currency. The corporate sectors of Hungary, Chile, Czech Republic, Malaysia and Poland have much
larger external debt outstanding (more than 40% of GDP). Hungarys corporate sector is the most indebted even though, as a
percentage of GDP, its external debt has fallen to 90%, from 107% at end-2008. By contrast, corporate sector external
borrowing as a percentage of GDP in the Czech Republic, Chile and Turkey has risen by more than 10pp since 2008. Note,
however, that these data include not only bond issuance but also loans that international banks may have extended to EM
corporates and financial institutions.
The corporate sector in EM economies is increasingly borrowing by issuing offshore debt securities. EM corporations borrow
abroad via offshore affiliates and repatriate the proceeds. Some authors have mentioned that this sort of corporate borrowing
is partly responsible for the massive expansion in EM corporate issuance in international bond markets in the past few years.
This has probably increased foreign exchange risk exposure in EM. This expansion also means that indicators of vulnerability
that are based only on international bank credit expansion do not fully capture financial system risks.4
The increase in EM corporates foreign exchange risk is not necessarily a big negative, however. A stronger USD environment is
likely to weigh on companies that earn revenues in local currency but incur costs in USD. By contrast, companies earning
revenues in USD but incurring costs in local currency are likely to be less risky. This is the case with many exporters, which
usually issue debt in the currency of their export receipts (see Emerging Markets Weekly: Dont drop your guard, 15 January
2015). Other risks from the external environment include rising USD interest rates and higher rate volatility.
In sum, EM corporate external debt has soared in recent years and, unlike sovereign debt, has risen materially relative to GDP
since the mid-2000s. The build-up of corporate debt varies by country, with CEEMEA looking most exposed. A rise in the USD
and higher US interest rates are important risks that need to be monitored, especially given that markets do not appear to have
discriminated by the size of corporate debt in recent episodes of risk aversion (Figure 16). Overall, given the existing buffers at
the sovereign level, we see the risk of corporate stress as a pocket of vulnerability rather as a source of systemic risk of EM.
See The global long-term interest rate, financial risks and policy choices in EMEs, Philip Turner, BIS working paper 441
and BIS Quarterly Review, December 2014 - Non-financial corporations from emerging market economies and capital flows.
24 February 2015
69
FIGURE 16
EM corporate debt* is generally higher than sovereign
external debt (% GDP Q3 2014)
% GDP
%
18
16
General government
Corporate
14
12
10
8
6
Philippines
Mexico
Argentina
Colombia
Indonesia
Peru
India
Brazil
Korea
Thailand
South Africa
Russia
Poland
Turkey
Malaysia
Chile
Czech Republic
4
Hungary
90
80
70
60
50
40
30
20
10
0
FIGURE 17
EM corporate and sovereign (OAS) external debt spreads
have converged since mid-2011
2
0
-2
03
Source: World Bank- Quarterly external debt statistics, Haver, Barclays Research
*Corporate external debt is total external debt minus government and central
bank debt. It therefore includes debt by deposit taking corporations.
04
05
06
07
08
09
10
11
12
13
14
Click here to view an interactive Barclays Live Chart. Source: Barclays Research,
Bloomberg
EM equities
Vulnerable to external factors
EM equities have been
resilient to the less benign
external conditions
EM market equities clearly capture the transitions from bust to boom and landing cycles in the
past two decades. In fact, their path is notably similar to our measure of external
supportiveness (Figure 18). The external drivers we identified supported EM GDP and earnings
growth in the boom years, so the close relationship between the measure and EM equities is
not entirely surprising. Similarly, when those drivers became less supportive starting in 2011,
EM equities fell. Given our view that the external environment will become even more
challenging, there seems to be room for some downside risk to EM equities.
However, looking at EM equities by region, it is apparent that the resilience of the overall
index has been driven by EM Asia. LatAm has already priced in the more challenging
backdrop (Figure 19). That is not surprising given that the region benefited hugely from
China's growth and demand for a wide range of commodities, including bulk commodities
(iron ore), crude oil, base metals and agricultural products.
CEEMEA equities have also priced in the bad news (Figure 20). Despite the region being a
net commodity importer, it has a large exposure to the euro area via trade and financial
links. So perhaps the region has been penalised more for its dependence and links with the
euro area and the negative effects of the ongoing crisis.
Our analysis includes a group of 16 EM economies (Brazil, China, India, Mexico, Poland, Russia, South Africa, Turkey,
Korea, Indonesia, Malaysia, Thailand, Hungary, Colombia, the Philippines, Chile) and a group of 11 DM (US, Canada,
Japan, UK, Germany, France, Italy, Spain, Netherlands, Sweden, Norway).
24 February 2015
70
FIGURE 18
EM equities: Overall resilience masks regional differences
Boom Years
Measure of external supportiveness
EM equity (RHS)
Z Score
1.4
Index
FIGURE 19
LatAm seems to have priced in the more challenging backdrop
already
1,400
1.4
1.0
1,200
1.0
0.6
1,000
0.6
0.2
800
0.2
-0.2
600
-0.2
-0.6
400
-0.6
-1.0
200
-1.0
-1.4
-1.4
96
98
00
02
04
06
08
10
12
Index
5,500
4,500
3,500
2,500
1,500
500
-500
96
14
Boom Years
Measure of external supportiveness
LatAm equity
Z Score
98
00
02
04
06
08
10
12
14
The gap between the EM Asia equity index and our measure of external supportiveness
widened further in 2014 (Figure 21) as equities markets in the two regional heavyweights
China and India rallied strongly. Chinese equities had posted a weak H1 14 but surged in
H2 14, despite monthly economic indicators indicating easing domestic demand. Poor
economic data fuelled investor expectations that the PBoC would ease policy. Favourable
terms of trade as a result of lower oil prices, as well as increased retail participation and the
beginning of a Shanghai-Hong Kong cross-trading link, boosted Chinese equities in H2 15.
Structural reform progress under the Xi-Li government, expectations of further benchmark
rate and reserve requirement ratio (RRR) cuts and low valuations of large-cap stocks fuelled
another c.30% rally in Chinese equities between November 20 and end-December 2014,
pushing the SHCOMP index up c.50% in 2014.
In India, expectations of a Narendra Modi-led government began to spur investor interest in
Indian equities at the start of 2014. A landslide win for the BJP pushed Indian equities to touch
record highs in May 2014. More important, the steep drop in oil prices turned key
macroeconomic indicators favourable last year, further boosting investor interest in India.
Lower oil prices resulted in easing inflation, shrank the current account deficit and prompted
the government to undertake fiscal consolidation measures. Moreover, GDP growth improved
FIGURE 20
CEEMEA also seems to have priced in the bad news
FIGURE 21
EM Asia shows some structural resilience
Boom Years
Measure of external supportiveness
CEEMEA equity
Z Score
1.4
Index
1.0
500
1.4
400
1.0
0.6
Boom Years
Measure of external supportiveness
EM Asia equity
Z Score
Index
600
500
0.6
0.2
300
-0.2
200
-0.6
100
-1.0
-1.4
0
96
98
00
02
04
24 February 2015
06
08
10
12
14
400
0.2
300
-0.2
200
-0.6
100
-1.0
-1.4
0
96
98
00
02
04
06
08
10
12
14
71
and earnings growth accelerated. All these factors, combined with a better policy
environment, resulted in a significant rise in fund flows into India, boosting Indian equities by
c.30% in 2014. See Chapter 6, India: A step change for a thorough discussion of the positive
economic changes that India has recently experienced as well as its bright prospects.
Ongoing and expected structural changes in China and India should help equity markets in
these countries to continue to attract investor attention. China still has significant control over
its capital account, but gradual liberalisation has been pushed forward in recent years. Further
opening of Chinas capital account will be appealing to investors (see Macro Daily Focus:
What causes divergent performance of onshore and offshore China-related stocks? 20 January
2015). In India, a sustained decline in inflation and inflation expectations, improved consumer
spending and increased capacity utilisation are likely to work in favour of equities.
EM Asias resilience could
also reflect the benefits of
lower oil prices and sizable
exports to the US
EM Asias resilience could also stem from the fact that most Asian economies export a
sizable fraction of their GDP to the US. Moreover, falling commodity prices (notably oil
prices) have been a positive terms of trade shock for EM Asia as it is the main oil importer
and consumer within EM. Lower inflation gives central banks more scope to keep policy
accommodative.
Figure 22 shows that EM earnings yields were higher than in DM in the early 2000s. This
gap closed as EM equities outperformed DM into and after the global financial crisis of
2008-09. In 2011, both metrics started drifting apart, with EM equities underperforming
DM, in line with a tougher external environment. The taper tantrum of 2013 opened that
gap further and there has since been a gradual fall in the EM earnings yield. Note also that
the recent acceleration in the fall of the EM earnings yield has been mainly because of the
rally in Chinese equities. Figure 23 shows that the fall is more gradual and the gap to DM
still wide when China is excluded. Overall, this suggests that EM equities offer more
attractive earnings yields, but we also know that this is for good reason, as the external
environment has been a lot tougher for EM since mid-2011.
FIGURE 22
EM vs DM equity earnings yields (%) - gap closing largely
because of recent rally in China
14
(%)
12
EM-DM diff.
16
DM
14
EM
10
FIGURE 23
EM ex-China vs DM equity earnings yields (%) - gap remains
wide when China is excluded
(%)
EM ex-China - DM diff.
DM
EM ex-China
12
10
-2
-2
03
04
05
06
07
08
09
10
11
12
13
14
03
04
05
06
07
08
09
10
11
12
13
14
Note that a common measure of the equity risk premium is the earnings yield minus real risk free rates. We do not
use this metric in this article. Note, however, that the difference between earnings yields and real bond yields is
currently roughly the same for EM and DM.
24 February 2015
72
FIGURE 24
EM vs DM equity earnings yields based on SCAPE (%) is wide
FIGURE 25
EM vs US equity earnings yields based on SCAPE (%) is even
wider
2
06
07
08
09
EM
10
11
12
13
14
07
08
09
DM
10
11
EM
06
12
13
14
US
Many analysts dislike aggregating EM and DM countries and making comparisons between
them because of big sector differences. Indices like the DAX, the argument goes, include highvalue-added corporates and virtually no commodity producers, whereas the opposite is true in
countries like Russia. How can their earnings yields be compared? Our in-house Sector and
Cycle-Adjusted P/E ratios (SCAPE) take into account such sector differences.7 We use the
inverse of these metrics (earnings over price) to compare the earnings yields in EM and DM
(Figure 24). The picture does not really change much. If anything, it shows that the gap
remains wide and is even wider when comparing EM and US earnings yields (Figure 25).
Of course, one reason why earnings yields are higher in EM versus DM is because real
interest rates are higher there, too, so the compensation is not just for a risky earnings
profile; it is also because of higher bond risk premia.
We define the risk premium in local debt markets as the real yield on 5y local bonds (5y
yields minus CPI inflation). Figure 26 shows a GDP-weighted average of these metrics for
selected EM and DM economies. During the boom years, EM had average real rates of
around 1.5% excluding the global financial crisis period. These were not very different from
FIGURE 26
Real bond yields are higher in EM vs DM (5yr real yields, GDP weighted)
4
(%)
EM-DM diff.
DM
EM
3
2
1
0
-1
-2
03
04
05
06
07
08
09
10
11
12
13
14
24 February 2015
For details see Introducing the SCAPE: why US equities are less expensive than they seem, Equity Gilt Study 2014.
73
FIGURE 27
Real bond yields in EM reflect high real EM policy rates (GDP-weighted)
4
3
2
1
0
-1
-2
06
05
07
08
09
10
11
12
13
14
those of DM in that period. However, this pattern changed significantly in the period from
2011 to 2014. DM real yields went deeply negative, while EM real yields initially fell but then
rose to about 2% in 2014.
High real yields in EM are
hard to ignore when those in
DM are close to zero
From an asset allocation point of view, this difference is hard to ignore. The universe of
positive real yields is becoming scarcer as central banks in the developed world continue to
hold policy rates close to zero, with some even venturing into negative policy rates. This
contrasts with policy rates in EM, where the zero bound is further away. Indeed, real policy
rates in EM are close to 3%, the highest since 2007 (Figure 27). The rise in real policy rates
since 2011 has more to do with falling inflation than higher nominal policy rates: while EM
policy rates remained broadly stable, inflation has gradually fallen since 2011. Looking
ahead, the growth environment is likely to remain challenging and external conditions to
deteriorate further, while inflation is likely to remain subdued given the fall in energy prices.
In this context, local bonds in EM are an interesting proposition. Cyclical headwinds are
likely to keep central banks in check even as US yields start turning higher. In addition, carry
is still generally high and in some countries (eg, India and Brazil) it should compensate for
weaker currencies.
FIGURE 28
Real bond yields 5y minus CDS spreads plenty of potential left in some EMs after
stripping out credit risk
Japan
Russia
Turkey
Indonesia
Chile
Norway
France
Mexico
Italy
South Africa
Sweden
Germany
Philippines
Netherlands
Malaysia
UK
Colombia
US
Thailand
Spain
Korea
Poland
Hungary
bp
Brazil
4
3
2
1
0
-1
-2
-3
-4
-5
-6
24 February 2015
74
As EM real yields embed credit risk premia on local bonds, it is helpful to gauge the risk
premium left in bonds after stripping out such credit risk. Although these risk premia are
difficult to measure, metrics of credit risk on external debt, including CDS, can be used as
proxies. On our calculations, these premia are high in such countries as Brazil, Hungary and
Poland (Figure 28).
Sovereign external debt is perhaps the asset class within EM that received most attention
during the boom years, largely because issuance had been extensive during and after the
EM sovereign crises as most sovereigns had to access international financial market via
foreign-currency-denominated debt. We study credit risk on sovereign external (USDdenominated) debt via sovereign CDS spreads.
Figure 29 shows the GDP-weighted average of these spreads since 2003 for our groups of EM
ex China and DM economies.8 The first observation is that EM ex-China CDS spreads reached
a low in 2007, before the global financial crisis. The second is that these spreads have moved
broadly sideways but above the 2007 tights since 2011. We attribute this to the role that risk
appetite and global macro volatility play in EM credit risk. Figure 30 shows that risk appetite,
proxied by the VIX, has a high correlation with EM credit risk. Global risk appetite has
strengthened recently but has not returned to pre-2008-09 levels. However, it has remained
strong enough to contain the rise of EM credit risk in the face of a tougher external backdrop
since 2011.
Other external factors have also influenced EM credit spreads recently. For example, in
Figure 30, the large spike in mid-2011 coincides with the slowdown in Chinese activity and
the widening in 2013 stemming with the taper tantrum. At the same time, at the country
level, the domestic fundamentals discussed earlier also play a big role. We study the
importance of external and domestic influences in the Box below.
FIGURE 29
EM vs DM CDS spreads*: Higher EM spreads mainly reflect
tougher external environment
700
bp
700
FIGURE 30
EM CDS spreads* and VIX: External drivers matter
DM
600
bp
EM ex China (lhs)
VIX (rhs)
70
600
60
500
50
400
40
300
30
200
20
100
10
EM ex-China
500
400
300
200
100
0
03
04
05
06
07
08
09
10
11
12
13
14
0
03 04 05 06 07 08 09 10 11 12 13 14
We exclude China as the sovereign has very little external debt. Indeed, Chinas weight in our sovereign external debt
index is close to zero.
24 February 2015
75
T-Stat
Coefficient
T-Stat
1772.2
7.5
1594.1
6.4
77.9
30.3
71.3
24.1
6.6
15.8
6.9
15.9
-23.7
-10.2
-21.6
-8.8
Commodity prices
-0.9
-5.5
-0.8
-5.0
-327.5
-7.5
WB governance indicators
Adjusted R-squared
Cross-sections included
Total panel (unbalanced) observations
0.69
0.67
16
16
2417
2225
Source: Barclays Research. *Statistically significant at less than 1%. When standard errors are corrected for
autocorrelation, the coefficients remain statistically significant at the 1% level, with the exception of the
governance indicator, which is significant at the 15% level.
All the estimated coefficients of the variables considered are statistically significant
and their signs are the ones we expected. Countries with better sovereign credit
ratings (lower numerical rating) tend to have tighter spreads. In particular, a onenotch credit upgrade (coded as a decline of 1 in the sovereign credit variable) is worth
a 70-80bp spread compression, depending on the specification. Countries with better
governance/institutions (higher World Bank governance indicators) also tend to have
tighter spreads: a typical (1 std dev) improvement tightens spreads by 25bp. This is
important because it suggests that institutional progress matters even after considering
the improvement in macro fundamentals captured by ratings and other countryspecific factors captured by the country dummies. The country fixed effects explain
23% of the CDS spreads variation. See Figure 31 for details of the model estimates.
On the external front, a typical (one standard deviation) rise in the China leading
indicator is worth a spread compression of 38-41bp depending on the specification. A
10% rise in commodity prices tightens spreads by 8-9bp, while a typical rise in the VIX
widens spreads by 56-59bp, also depending on the model specification on Figure
31.These results, combined with our view that Chinese growth will continue to slow,
macro volatility will likely rise, commodity prices will not surge, EM macro fundamentals
will not improve materially and institutional reform will remain challenging, mean that a
sustained compression of spreads based on these factors is unlikely. Country
differences may present interesting opportunities rather than directional views for the
asset class as a whole. For example, countries that are directly linked to Chinese
demand for commodities like Peru and Chile may remain challenged, while those with
prospects of deeper reform, like Mexico or Indonesia, are in a better position.
24 February 2015
76
Our model does not incorporate many other variables that affect sovereign spreads,
such as structural changes in financial institutions demand for EM assets, or
forward-looking metrics of country fundamentals (eg, prospects for institutional
reforms, political outcomes or growth outlooks). With those caveats in mind, we use
the model to assess fair value spread estimates (Figure 32). Mexico, for example,
ended 2014 with tighter spreads than the model predicts, suggesting that markets
expect US growth and domestic structural reforms to anchor credit risk. Hungary is
another interesting example, where the model predicts much wider spreads,
reflecting worsening credit ratings and a market perception that euro area policies
will indirectly support Hungary. Russian spreads are wide versus the model but our
ratings metric misses the S&P and Moodys downgrades early in 2015 and likely
further downgrades this year, while sanctions are probably limiting some investors
ability and willingness to engage in Russian risk.
FIGURE 32
Actual CDS vs fair value estimates (December 2014)
600
bp
500
400
300
200
100
Actual
POL
CHN
PHL
CHL
MEX
MAL
PER
COL
IDN
HUN
TUR
SAF
BRA
RUS
Model
Source: Barclays Research, Bloomberg and IMF. Model included the following explanatory variables: credit ratings,
China leading index of activity, commodity prices and VIX. Argentina and Venezuela are not shown in the chart.
Their CDS were truncated at 1500bp estimation purposes as they trade at distressed levels. The values predicted by
the model are 1397bp and 1092bp.
Overall, a more challenging external environment may provide a difficult backdrop for EM
sovereign credit. And although macro vulnerabilities are generally lower than they
were15-20 years ago, growth prospects are limited by lack of structural reforms. But
rather than systemic crises in EM, we are likely to see differentiation against a more
difficult external backdrop. There are also positives for some EM countries on the
horizon, such as low crude prices and a stronger and more solid US economy, so
sovereign credit in countries that benefit from that backdrop, mainly in EM Asia, is likely
to outperform.
FX risk premia
We explore two key areas related to EM FX risk premia: first, whether EM currencies have
adjusted enough to the negative shift in underlying fundamentals in recent years; second,
whether EM FX carry remains an attractive investment.
EM currencies (ex-China) are
now cheap relative to
fundamentals, but not
markedly so
24 February 2015
As best we can judge, the answer to the first question is yes, but only just. Figures 33- 35
look at the percentage misalignment between real effective exchange rates and our FX
teams behavioural equilibrium exchange rate model (BEER) for an EM aggregate and EM
regional aggregates (see Currency valuation from a macro perspective, 14 June 2011). Back in
2002, when the EM boom began, the aggregate FX valuation was signalling very cheap EM
currencies perhaps unsurprising following years of EM crisis episodes. Over the
subsequent 10 years, EM valuations climbed, peaking around mid-2011, which marked the
start of Chinas structural slowing and the end of the commodity super cycle. Admittedly,
77
FIGURE 33
Aggregate EM FX (ex China) are now cheap relative to
fundamentals, but not markedly so
GDP-weighted misalignment between real effective FX
rate and Barclays BEER model
FIGURE 34
Outside of China, Asian currencies remain below
fundamental fair value, mainly because of India
10
10
-5
-5
-10
-10
-15
-15
02
04
06
08
EM
10
12
14
EM ex China
-20
02
04
Asia
06
08
Asia ex China
10
12
14
the overall EM FX misalignment with fair value has continued to grow in the past few
years the GDP-weighted EM currency basket now stands a little over 5% above fair value,
having been 10% cheap to fair value in 2002. But todays overvaluation of EM currencies is
mainly a function of Chinas real effective exchange rate, which is close to 20% overvalued,
according to our BEER model. Once we remove China, EM currencies have shifted to mildly
cheap over the past year or so.
Outside of China, Asian
currencies remain below
fundamental fair value, mainly
because of India
Regionally, the picture is an interesting one. Asia excluding China has been cheap to fair
value for much of the past 15 years only in 2005-06 did Asian currencies move slightly
above fair value. Today, Asian currencies are still more than 5% below fair value, although
this is almost exclusively a function of a still-cheap Indian rupee. Most other Asian
currencies are at or above fair value the Philippine peso is, according to our BEER model,
the most expensive currency in EM (Figure 36). Latin Americas FX valuations have largely
followed the ups and downs of the regions terms-of-trade cycles. Having started the EM
boom period close to 30% cheap to fair value, LatAm currencies had moved to 20%
expensive by 2011. Today, despite persistent currency weakness, LatAm currencies remain
some 5% above fair value mainly a function of the still expensive Brazilian real (11%
expensive). CEEMEA currencies have shifted from significantly expensive just a few months
FIGURE 35
Outside of Asia, EM currencies have moved from significant
overvaluations to fair value in recent years
GDP-weighted misalignment between real effective FX
rate and Barclays BEER model
30
FIGURE 36
Largest EM FX misalignments according to Barclays BEER
model
% gap between real effective FX rate and Barclays BEER model
30
Five cheapest EM
currencies
20
20
10
10
-10
-10
-20
-20
-30
-30
02
04
LATAM
06
08
CEEMEA
24 February 2015
10
12
14
CEEMEA ex Russia
IDR ZAR
2002 average
78
ago, to near fair value today almost exclusively a function of the Russian rouble closing its
near-40% gap to fair value. According to the BEER model, the rouble is now below fair value
for the first time in more than 10 years. Elsewhere, the Turkish lira is one of the more
expensive EM currencies, while the South African rand is the cheapest.
Positive carry has been a big
support for EM currency
returns
Another way of thinking about EM currencies is to ask whether they have been a good
investment choice, irrespective of fundamental valuations. In other words, do EM currency
risk premia compensate investors for the risk they are taking? Here, we would say the
evidence is compelling. One simple exercise is to track the performance of the five highyielding EM currencies (TRY, INR, IDR, ZAR and BRL) that have received significant attention
in recent years because of a mix of a deteriorating global backdrop and poor domestic
fundamentals. Figures 37 and 38 look at the average BEER misalignments and the total
returns of holding an equally weighted basket of these currencies, with and without carry.
On a fundamental basis, this basket of five EM currencies began the EM-boom period some
15-20% undervalued and by the middle of 2011, the start of the landing period, were
10% expensive. Since then, valuations have dropped and these five high-yielding EM
currencies are now, on average, mildly cheap, though neither notably nor universally so.
During this almost 15-year period, this basket of high-yield EM currencies has dropped by
around 30% against the US dollar. Still, once the carry is added back into the basket (risk
premia), the returns of holding these high-yielding EM currencies have been more than
150%, or around 7.5% per annum. Indeed, despite a sense of crisis in EM currencies in the
past few years, total returns in the basket are down just 13% from the peak and virtually flat
from May 2013, when then-Fed Chairman Ben Bernanke delivered his famous tapering
speech. Net-net, it would seem that in the universe of higher-yielding EM currencies,
investors have been broadly well rewarded for being long.
Carry in high-yielding EM
currencies relative to volatility
has picked up in recent years,
making them an attractive
investment
FIGURE 37
EMs vulnerable 5 * have moved from expensive to modestly
cheap in the last few years
15
10
FIGURE 38
Returns in high-yield EM FX * from mid 2011 to present
index, 30-Jun 2011 = 100
120
110
100
90
0
80
70
-5
60
50
-10
mid 2011
peak
40
-15
30
01
-20
02
04
06
08
10
24 February 2015
12
14
03
05
excluding carry
07
09
11
13
15
including carry
79
FIGURE 39
Average high-yield EM FX carry-to-volatility ratio *
implied FX carry / implied 3-month volatility
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
03
04
05
06
07
08
09
10
11
12
13
14
Source: Barclays Research, Bloomberg. * index is 12m implied FX carry / 3m implied FX volatility for TRY, BRL, INR and ZAR
References
Barclays
Navigating the new EM landscape: Where to find the best returns, Equity Gilt Study 2011.
Introducing the SCAPE: Why US equities are less expensive than they seem, Equity Gilt Study
2014.
Macro Daily Focus: What causes divergent performance of onshore and offshore Chinarelated stocks?, 21 January 2015.
Cross commodity themes and strategy: Plumbing the depths, 10 February 2015.
Emerging Markets Weekly: Dont drop your guard, 15 January 2015.
Currency valuation from a macro perspective, 14 June 2011.
Other
Avdjiev, Stefan, Michael Chui and Hyun Song Shin (2014), Non-financial corporations from
emerging market economies and capital flows, BIS Quarterly Review, December 2014.
Calvo, Guillermo (2005), Emerging Capital Markets in Turmoil: Bad Luck or Bad Policy?
Cambridge, MA: MIT Press, 2005.
Felices and Wieladek (2012), Are emerging market indicators of vulnerability to financial crises
decoupling from global factors? Journal of Banking & Finance, 2012, vol. 36, issue 2, pages
321-331.
IMF Global Financial Stability Report, April 2006, Chapter 3: Structural changes in emerging
sovereign debt and implications for financial stability.
IMF World Economic Outlook, April 2014, Chapter 4: On the receiving end? External
conditions and emerging market growth before, during and after the global financial crisis.
Reinhart, Carmen and Kenneth Rogoff (2004), The Modern History Of Exchange Rate
Arrangements: A Reinterpretation, The Quarterly Journal Of Economics; Vol. Cxix February
2004 Issue 1.
Roubini, Nouriel and Brad Setser (2004), Bailouts or Bail-Ins: Responding to Financial Crises
in Emerging Markets, Institute for International Economics, 2004.Turner, Philip (2014), The
global long-term interest rate, financial risks and policy choices in EMEs, BIS working paper
441.
24 February 2015
80
Appendix
A ranking of EM vulnerability to external headwinds
We build a ranking of EM vulnerability to external headwinds by combining the rankings of
vulnerability to the five external drivers discussed in this chapter Figures 40-44. We do this
by ranking each measure from 1 (most resilient) to 16 (most vulnerable). Figure 45 shows
the average of the five rankings. Not surprisingly, the most vulnerable countries are
commodity exporters exposed to weaker Chinese and euro area growth, such as Malaysia
and Russia. The most resilient are those that are relatively insulated from China and
European growth and benefit from carry and expected FX resilience, such as Mexico and
India. For more details see Macro Daily Focus: A ranking of EM vulnerability to a more
challenging external environment, 9 January 2015.
FIGURE 40
EM vulnerability to China growth slowdown (exports to
China over GDP, higher = more vulnerable)
FIGURE 41
EM vulnerability to euro area growth slowdown (exports to
euro area over GDP, higher = more vulnerable)
18
16
14
12
10
8
6
4
2
0
60
50
40
30
20
10
FIGURE 42
EM vulnerability to lower commodity prices (net commodity
exports over GDP, higher = more vulnerable)
FIGURE 43
EM vulnerability to higher US rates (correlation of EM
assets* and US rates vol, higher = more vulnerable)
Hungary
Russia
Poland
Turkey
Malaysia
Thailand
Chile
South Africa
Korea
Taiwan
Colombia
Philippines
India
Brazil
Indonesia
Mexico
Taiwan
Korea
Malaysia
Chile
Thailand
Philippines
South Africa
Brazil
Indonesia
Russia
Hungary
Colombia
India
Mexico
Poland
Turkey
30
20
15
25
10
5
20
0
-5
15
-10
-15
24 February 2015
Turkey
Mexico
Philippines
Korea
Malaysia
Indonesia
Brazil
South Africa
Hungary
Colombia
India
Poland
Russia
Taiwan
Chile
Thailand
Russia
Colombia
Malaysia
Chile
Indonesia
Brazil
South Africa
Mexico
Taiwan
Poland
Turkey
Hungary
Thailand
India
Philippines
Korea
10
81
FIGURE 44
EM vulnerability to stronger USD (12m EM FX appreciation
vs USD + 5y bond yields, lower = more vulnerable)
FIGURE 45
Ranking of EM vulnerability (higher = more vulnerable)
%
15
14
10
12
Higher =
more
vulnerable
10
-5
24 February 2015
Malaysia
Chile
Russia
Korea
Hungary
Taiwan
South Africa
Turkey
Poland
Philippines
Colombia
Indonesia
Brazil
Thailand
India
Poland
Hungary
Korea
Chile
Russia
Thailand
Philippines
Taiwan
Malaysia
Turkey
South Africa
Colombia
Indonesia
India
Brazil
0
Mexico
-15
Mexico
-10
82
Average inflation
(% YoY)
19982001
20022010
20112014
19982001
20022010
20112014
19982001
20022010
20112014
19982001
20022010
20112014
Argentina
-1.2
5.0
2.7
-0.6
10.9
10.1
Brazil
1.5
3.9
1.6
5.5
6.7
6.1
-2.8
2.7
-0.6
-3.3
-3.1
-3.3
-4.1
-0.2
-2.9
-4.7
-3.4
Chile
2.6
4.2
4.3
4.0
3.2
-3.1
3.1
-1.8
1.3
-2.5
-0.7
2.2
China
8.0
10.7
8.0
-0.3
-0.1
2.3
3.2
2.0
5.6
2.0
-2.2
-1.4
Colombia
0.2
4.4
5.0
-0.3
11.7
5.3
2.9
-1.0
-1.8
-3.3
-3.8
-1.7
Czech Republic
2.2
3.5
-1.1
0.6
5.3
2.3
1.8
-3.5
-3.6
-1.5
-4.4
-3.9
Hungary
3.8
-2.5
1.8
1.0
10.8
5.2
2.9
-7.4
-5.9
1.7
-3.7
-6.4
India
-0.8
5.9
7.8
5.5
6.2
6.5
9.2
-0.5
-0.9
-3.2
-8.3
-8.2
-7.4
Indonesia
-1.1
5.4
5.9
23.5
8.3
5.4
4.4
1.9
-2.3
-1.8
-0.6
-1.7
Korea
4.6
4.4
3.2
3.7
3.1
2.3
4.4
1.8
4.4
2.3
1.4
1.1
Malaysia
2.0
5.1
5.4
2.8
2.3
2.5
11.5
13.4
6.4
-3.8
-4.0
-3.9
Mexico
3.0
2.1
2.9
12.1
4.5
3.8
-2.7
-1.2
-1.6
-4.3
-2.3
-3.8
Peru
1.7
6.0
5.4
4.1
2.4
3.3
-3.6
-0.5
-3.7
-2.1
0.2
1.2
Philippines
2.5
5.0
6.0
6.9
4.6
3.8
-1.6
2.6
3.0
-2.7
-1.9
-0.4
Poland
3.7
4.2
2.8
8.7
2.5
2.2
-5.2
-4.3
-2.9
-3.7
-5.0
-4.1
Russia
4.0
4.9
2.3
38.9
11.6
6.9
10.4
7.5
3.2
-1.3
2.8
-0.1
South Africa
2.4
3.6
2.3
5.8
6.1
5.7
-0.5
-3.6
-4.8
-1.4
-1.4
-4.4
Taiwan
3.4
4.5
2.8
0.8
1.1
1.4
3.3
8.1
10.8
-4.6
-3.8
-3.4
Thailand
0.2
4.6
2.6
2.9
2.7
2.8
8.7
2.7
1.1
-4.7
-0.4
-1.3
Turkey
0.2
5.1
4.5
64.7
14.5
8.0
-0.3
-4.0
-7.4
n.a.
-5.0
-1.4
Venezuela
0.4
3.5
2.0
22.0
23.2
38.0
2.3
10.1
6.1
-1.0
-2.4
-14.3
Credit ratings
WB governance
19982001
20022010
20112014
19982001
20022010
20112014
19982001
20022010
20112014
19982001
20022010
20112014
Argentina
37.8
75.6
40.9
42.8
60.0
22.7
13.7
17.3
17.2
0.0
-0.3
-0.3
Brazil
68.7
69.0
66.2
37.4
25.4
19.1
14.0
12.3
9.3
0.0
0.0
0.0
Chile
13.3
8.1
12.4
46.9
41.3
43.5
7.3
6.3
4.2
1.1
1.2
1.2
China
37.2
34.9
38.5
13.5
11.1
9.4
7.8
6.3
4.0
-0.5
-0.5
-0.5
Colombia
35.0
38.2
34.4
35.5
26.3
22.9
11.2
11.6
9.8
-0.6
-0.5
-0.3
Czech Republic
18.0
30.1
44.5
35.0
38.0
52.5
7.4
6.0
4.6
0.6
0.9
0.9
Hungary
57.3
67.0
80.1
61.5
113.9
152.5
8.3
7.1
11.1
0.9
0.9
0.7
India
72.6
77.4
63.8
21.5
17.6
20.7
11.8
10.6
10.0
-0.2
-0.3
-0.3
Indonesia
87.6
43.6
25.2
109.2
43.3
27.9
16.8
14.2
10.7
-0.8
-0.7
-0.4
Korea
16.5
26.3
33.3
28.7
25.9
33.0
9.7
6.5
4.9
0.5
0.7
0.8
Malaysia
37.7
45.2
56.2
51.3
48.3
60.9
9.0
7.3
7.0
0.3
0.4
0.3
Mexico
43.3
41.4
45.2
26.4
21.1
30.9
11.3
8.7
8.3
-0.1
-0.1
-0.1
Peru
44.0
36.7
20.9
56.5
38.1
27.5
12.5
11.9
9.0
-0.3
-0.3
-0.2
Philippines
55.6
53.8
39.3
73.4
50.7
24.7
11.0
12.7
11.0
-0.2
-0.5
-0.4
Poland
38.2
47.5
54.6
37.2
51.7
69.3
8.9
6.9
6.5
0.7
0.6
0.8
Russia
68.8
17.6
13.5
79.0
35.8
31.7
16.9
9.7
8.7
-0.8
-0.7
-0.7
South Africa
43.4
32.9
43.5
19.2
24.7
35.5
10.2
8.3
8.2
0.4
0.3
0.2
Taiwan
26.4
34.5
40.5
11.5
21.3
28.3
3.1
3.9
4.0
0.8
0.9
1.0
Thailand
55.5
45.3
45.2
74.0
33.8
34.5
10.3
8.4
8.0
0.3
-0.1
-0.3
Turkey
77.9
52.1
36.3
44.5
42.6
43.1
14.5
13.4
11.2
-0.3
-0.1
-0.1
Venezuela
31.8
40.2
46.9
40.4
30.1
31.1
14.6
14.8
15.1
-0.6
-1.1
-1.3
24 February 2015
83
CHAPTER 4
Michael Gapen
with lost output and slower potential growth. More than five years after the end of
the global recession, we feel enough time has passed to assess the extent of the
destruction of output in developed economies.
In applying a uniform framework across seven developed economies that account for
nearly half of world output, we estimate that potential growth in these economies has
fallen by 1.5pp since 1999 and, in turn, has reduced global potential growth by 0.7pp.
Our finding that slower growth in developed economies could slow global growth by
0.7pp is of similar magnitude to the effect of a slowing China on global growth.
Slower potential growth in developed economies and a decelerating Chinese
economy have reduced global potential growth by 1.5pp a significant deceleration.
We estimate that the effects of the recession accounted for about two-thirds of the
1.5pp decline in potential growth in developed economies, with the remaining onethird pre-dating the global recession. Policymakers efforts to stem the tide have been
effective, but we doubt policy can fully reverse the slowing in trend output growth
before the end of the decade.
Economic downturns that coincide with severe financial crises destroy output and lower
potential growth. In this chapter, we examine the experience of seven large developed
economies that comprise nearly half of world GDP based on purchasing power weights
France, Germany, Italy, Japan, Spain, the UK, and the US to estimate the damage to output
and trend growth from the recent recession. The recession hit when many of these countries
were already experiencing a deceleration in trend growth related to demographic factors and
the fading of the effects of the technology revolution. The slowing of population growth and
rising dependency ratios across much of the economically advanced world was a subject we
took up in last years Equity Gilt Study (see Economic implications of demographic change,
Equity Gilt Study 2014, 13 February 2014). We also take up the importance of demographic
trends in boosting saving rates and asset prices in Chapter 1 of this years Equity Gilt Study,
FIGURE 1
Peak-to-trough decline in real GDP
FIGURE 2
Trough-to-peak rise in the unemployment rate
%
24 February 2015
Germany
France
US
UK
Germany
Spain
Japan
Italy
-12
Japan
-10
UK
-8
France
-6
US
-4
Italy
-2
Spain
20
18
16
14
12
10
8
6
4
2
0
84
The destruction of output and slowing of potential growth in the developed world comes just
as growth outside the developed world is also slowing. We do not find this surprising, given
the extent of globalization and linkages among developed and emerging economies. We
expect potential GDP growth in China to slow from about 9-10% in the 1990s to about 6.0%
in the coming 5-10 years as it transitions from investment-led to consumption-led growth.1 If
realized, this would lower the growth rate of potential global GDP by another 0.6-0.7pp, given
Chinas burgeoning share of world output.
Taken together, and assuming policy cannot significantly reverse the effects of the global
recession, slower potential growth in developed economies and a decelerating Chinese
economy could reduce global potential growth by 1.5pp annually. Growth in emerging market
economies outside of China are also slowing in part because of the rebalancing of the global
economy following the recession and financial crisis, which helped to narrow the current
account deficit in the US. For our view on how this will impact risk premia and asset returns in
emerging markets, see Chapter 3, EM is still an attractive asset class. Outside of India, where
the growth outlook appears more promising, we see the bulk of the evidence as pointing to a
significant deceleration in potential growth. Much that once was, now appears lost.
See China: Beyond the miracle The complete series, 1 March 2013.
Carmen M. Reinhart and Kenneth S. Rogoff, This time is different: Eight centuries of financial folly, Princeton NJ:
Princeton University Press, 2009.The statistics cited herein are from eighteen post-war financial crises, including five
severe cases (Spain, 1977; Norway, 1987; Finland, 1991; Sweden, 1991; and Japan, 1991) along with other examples
from East Asia and Latin America.
24 February 2015
85
The duration of the decline in real output varied substantially, with Germany and Japan
experiencing relatively short four-quarter declines in output, while the downturns in Italy
and Spain have lasted much longer. The duration of the declines in Italy and Spain is open
to debate. In Italy, real output peaked in Q1 2008 and fell for seven quarters through the
end of 2009, before rising for six consecutive quarters thereafter. However, beginning in Q3
2011, real output has fallen steadily and, 27 quarters since 2008 Q1, output in Italy has yet
to stage a convincing turnaround. Spain posted a similar double dip, with output falling for
six quarters before staging a brief rebound, only to fall further in 2011-12. More recently,
Spain has achieved five consecutive quarters of positive growth through 2014 Q3.
There are several channels through which potential output is lost following severe
economic downturns. The most common include:
Slower capital accumulation and distortions to the efficient allocation of capital. Weak
profitability reduced the ability of firms to self finance in an environment of tighter credit
standards. Sluggish economic growth and heightened uncertainty stemming from the
severity of the downturn also weighed on business sentiment and suppressed capital
accumulation. Finally, tighter credit conditions and a reluctance to lend reflect increased
risk aversion. We find features of this in each of the developed economies investigated.
For the US, our approach follows Charles Fleischman and John M. Roberts, 2011, From many series, one cycle:
Improved estimates of the business cycle from a multivariate unobserved components model, Finance and Economics
Discussion Series 2011-46. The US framework includes nine variables: real gross domestic product, real gross
domestic income, real nonfarm business output, real nonfarm business income, nonfarm business employment, the
work week, labor force participation rate, the employment rate, and core CPI inflation. For the remaining countries, a
scaled down model and six variables are used: real gross domestic product, real gross domestic income (if available),
employment, working hours, output per hour, employment, participation, and inflation. Variables in both models are
detrended by population growth. See the appendix for further details.
4
See Jun Ma and Mark Wohar, An unobserved components model that yields business and medium-run cycles,
Journal of Money, Credit, and Banking, 45(7), October 2013, for further discussion on the benefits of the unobserved
components model.
24 February 2015
86
models are estimated using quarterly data from 1963 Q1-Q1 2014 for the US, 1975 Q1-Q1
2014 for the UK, 1975 Q1-Q1 2014 for France, 1973 Q1-Q1 2014 for Germany, 1993 Q1Q1 2014 for Italy, 1996 Q1-Q1 2014 for Spain, and 1981 Q1-Q1 2014 for Japan.
We see several advantages to using a multivariate approach. Although it is more difficult to
implement, academic research has shown that multivariate analysis improves the accuracy
of cycle estimates and using a single system means the framework uniformly accounts for
trade-offs between alternative signals.5 Applying the framework across countries also
ensures that trade-offs between competing signals are treated in similar fashion. Our
common framework, detailed in the appendix to this chapter, makes several important
assumptions. First, we assume that each measure of economic activity and labor markets can
be represented as the sum of cyclical and trend components. Second, we assume that the
cyclical component is common across all the inputs, with the understanding that a wider set
of data should enable estimation of the trend with improved accuracy. Third, the cyclical
component is allowed to have both contemporaneous and lagged effects to account for
variables that may lag the cycle, yet still inform its estimation. Fourth, while each variable has
a common cyclical component, we permit each variable to have its own unique trend. Finally,
we allow cyclical deviations in output to affect inflation, creating a natural rate interpretation.
Our methodology allows for
cycles in both permanent
(trend) and transitory (cycle)
components
The benefit of the generalized unobserved components model is that it allows for dynamics
in both cyclical and trend components. Many traditional frameworks assume a smooth
trend and view recessions as temporary events that only inform the cycle. In other words,
volatility in the data is restricted to inform the estimate of the cycle, but not necessarily the
trend. Our methodology allows for cycles in both permanent (trend) and transitory (cycle)
components. Academic research has shown that this generalized framework is more
appropriate for capturing both short-term and medium-term cycles, where the latter may
be more suitable when dealing with movements in technology, research and development,
and efficiency of resource utilization.6 Our preference is to let the data speak for themselves
about whether volatility is related to transitory outcomes or structural phenomenon.
Arabinda Basistha and Richard Startz, 2008, Measuring the NAIRU with reduced uncertainty: A multiple-indicator
common-cycle approach, Review of Economics and Statistics, 90, 805-11. Also see James H. Stock and Mark W.
Watson, 1989, New indices of coincident and leading economic indicators, NBER Macroeconomics Annual 1989,
Oliver Blanchard and Stanley Fischer, eds., 351-394.
6
See Diego Comin and Mark Gertler, Medium-Term Business Cycles, American Economic Review, 96, 523-551.
24 February 2015
87
Figure 5 presents the results of the estimation across each economy related to the
estimation of the trend. Where possible, we present the results in decade-averages to
smooth through the variability in annual estimates. As the figure shows, potential growth
has decelerated in recent years in five of the seven economies in our sample, with Japan and
Germany the exceptions to the deceleration trend.
Beginning with the US, trend growth is estimated at 3.0-3.4% for the three decades ending
in 1999, with the decades of the 1970s and 1980s buoyed by trend growth in both hours
and productivity. In the 10 years ending in 2009, however, potential growth began to slow
as the trend labor input slowed. This slowing was initially offset by faster productivity
growth, which we attribute to the technology revolution that began in the US in the mid1990s and supported faster rates of productivity growth (Figure 3).
We estimate that trend growth in output in the US began to slow in 2001, falling from 2.5% to
1.5% by 2009, as the benefits of technological progress began to fade and the workforce
aged. Our US economics team has written frequently about US demographic trends and their
contribution to slower potential growth.7 In our view, the decline in labor force participation
since its peak in the early 2000s mainly reflects the ageing of the baby boomers. While labor
force participation among the 55+ age cohort has risen during this time period, it nonetheless
is half of the participation rate for the prime working age population (those aged 25-54).
Therefore, the ageing of the population naturally reduces aggregate labor force participation
despite the upward trend in participation among older people, leading to a structural decline
in potential growth. The model estimates that the participation rate dropped 2.5pp from 2007
to 2013, accounting for around three-quarters of the 3.2pp decline in the actual participation
rate (Figure 4) during the same period. This is consistent with the view that most of the
decline in the participation rate is structural and unlikely to be reversed.
In addition to the above, the US has been in a gradual transition from a goods-oriented
economy to a services economy, the latter of which is associated with more part-time
employment and a shorter average work week.8 Altogether, we estimate that these factors
caused US potential GDP growth to slow to 2.5% in the 10 years ending 2009 and 1.2% in
the post-recession period from 2010 through Q1 2014.
FIGURE 3
Trend growth in US output per hour
FIGURE 4
Trend US labor force participation
% y/y
3.0
68
67
66
65
64
63
62
61
60
59
58
2.5
2.0
1.5
1.0
0.5
70
74
78
82
86
90
94
98
02
06
10
14
70
74
78
82
86
90
94
98
02
06
10
14
See Beyond the cycle: Weaker growth, higher unemployment, 15 December 2010 and Dispelling an urban legend: US
labor force participation will not stop the unemployment rate decline, 1 March 2012
8
Employment in the goods sector in the US was nearly 40% of total private employment in 1965. The share has fallen
to around 15% in recent years, leaving the remainder (85%) in services. Since average weekly hours in the service
sector averages about 33 hours, compared to 41 hours for the goods sector, the relative shift into services has caused
average weekly hours for the overall US private sector to decline from 39 in 1965 to 34 today. See U-6
unemployment may not reach normal, 11 July 2014.
24 February 2015
88
FIGURE 5
Potential growth and its trend components
United States (% saar)
Potential output
1970-79
1980-89
1990-99
2000-09
2010-Q1 2014
3.4
3.1
3.0
2.5
1.2
Total hours
2.2
1.4
1.1
0.8
0.2
Population
2.0
1.2
0.1
1.3
1.0
LFPR
0.6
0.4
0.1
-0.2
-0.8
Employment rate
0.0
0.0
0.0
-0.1
-0.1
0.1
-0.5
-0.3
-0.1
-0.2
Non-farm productivity
1.6
1.8
1.8
2.3
1.1
GDO to NFBO
-0.4
-0.2
-0.3
-0.3
-0.3
0.1
0.1
0.5
-0.2
0.2
1972-79
1980-89
1990-99
2000-09
2010-Q1 2014
1.8
2.7
2.3
1.8
1.6
-0.8
0.0
0.1
0.1
1.5
Population
0.1
0.1
0.3
0.6
0.7
LFPR
-0.1
0.2
-0.2
0.0
0.1
Employment rate
-0.4
-0.1
0.2
-0.2
0.3
-0.4
-0.3
-0.2
-0.3
0.3
Productivity
Germany (% saar)
Potential output
2.6
2.7
2.2
1.7
0.1
1972-79
1980-89
1994-99*
2000-09
2010-Q1 2014
2.3
1.7
1.2
1.1
1.2
Total hours
-0.8
-0.1
-0.1
-0.2
0.3
Population
0.0
0.1
0.2
0.0
0.1
LFPR
0.3
1.1
0.5
0.4
0.3
Employment rate
-0.1
-0.4
0.1
0.1
0.4
-1.0
-0.9
-0.8
-0.6
-0.5
3.1
1.8
1.3
1.2
1.0
France (% saar)
1980-89
1990-99
2000-09
2010-Q1 2014
Potential output
2.7
2.1
1.3
0.9
Total hours
-1.0
-0.5
0.3
0.1
Population
0.5
0.4
0.6
0.5
LFPR
-0.4
0.0
0.1
-0.1
Employment rate
-0.3
-0.1
0.0
-0.2
-0.8
-0.8
-0.4
-0.1
Productivity
Productivity
Italy (% saar)
Potential output
3.7
2.6
1.0
0.8
1994-1999
2000-2009
2010-Q1 2014
2.0
0.4
-0.7
Total hours
1.1
1.1
-0.3
Population
0.0
0.4
0.5
LFPR
1.1
0.5
0.3
Employment rate
-0.1
0.2
-1.2
0.1
0.1
0.1
0.8
-0.7
-0.5
1996-1999
2000-2009
2010-Q1 2014
Productivity
Spain (% saar)
Potential output
4.1
2.2
-0.3
Total hours
4.5
1.9
-2.4
Population
0.4
1.4
0.6
LFPR
1.6
1.1
-1.0
Employment rate
2.3
-0.5
-1.7
0.2
-0.1
-0.3
-0.4
0.2
2.1
Productivity
24 February 2015
89
FIGURE 5, CONTD.
Potential growth and its trend components (continued)
Japan (% saar)
1981-1989
1990-1999
2000-2009
2010-Q1 2014
Potential output
3.8
1.3
0.2
0.4
Total hours
0.2
-0.9
-0.9
-0.5
Population
0.6
0.3
0.1
-0.1
LFPR
0.0
-0.1
-0.5
-0.1
Employment rate
0.0
-0.3
0.0
0.3
-0.3
-0.7
-0.5
-0.7
3.5
2.2
1.2
0.9
Productivity
Note: The reunification of West and East Germany in the early 1990s is omitted since the event creates an artificial recession in model estimates. The population
surge boosts potential GDP growth via a stronger labor contribution. The business cycle framework accounts for this by estimating a positive output gap prior to
reunification and a negative output gap immediately afterward. In terms of the effect on the trend, the reunification pushes trend output per hour down discretely in
1991 and the series resumes its trend growth thereafter. We omit the 1990-93 model estimates for this reason. Source: Barclays Research
broad-based capital
mismatch is cited as the most
likely explanatory factor
In the UK, the slowing in the rate of trend output is clearly related to a slowdown in trend
productivity growth. We find that productivity growth in terms of output per hour grew
between 2.2% and 2.7% per year in the three decades ending 1999. We then estimate that
productivity growth fell steadily from 2.6% in 2002, down to zero by 2008, and has stayed
near this level through Q1 2014 (Figure 6). This fall in labor productivity growth, or the
productivity puzzle, has been heavily investigated and several factors put forward to explain
the slowdown. A report from the Bank of England points to labor hoarding during the early
stages of the recession, reduced investment in physical and tangible capital, and misallocation
of resources in low to high productivity sectors.9 A higher cost of capital would encourage
firms to substitute less expensive labor for capital, but this explanation is often discounted
because aggressive monetary policy kept the cost of capital low for a portion of the postrecession period and modest rates of investment have meant the aggregate stock of capital
has not fallen enough (as a share of GDP) to fully account for the productivity slowdown.
Broad-based capital mismatch is cited as a more likely explanatory factor. As discussed by
Ben Broadbent, external member of the Monetary Policy Committee, data from the UK
Office of National Statistics show that the dispersion of output and relative prices across
sectors widened markedly following the recession.10 A reallocation of capital and labor
would reduce the dispersion across sectors, but this process takes time and, in the interim,
FIGURE 6
Trend growth in UK output per hour has slowed sharply
FIGURE 7
amid modest growth and a surge in trend employment
% y/y
4.0
% y/y
2.0
3.5
1.5
3.0
1.0
2.5
0.5
2.0
0.0
1.5
-0.5
1.0
-1.0
0.5
-1.5
0.0
-2.0
-0.5
-2.5
72 75 78 81 84 87 90 93 96 99 02 05 08 11 14
72 75 78 81 84 87 90 93 96 99 02 05 08 11 14
See The UK productivity puzzle by Alina Barnett, Sandra Batten, Adrian Chiu, Jeremy Franklin, and Maria SebastiaBarriel of the Bank of Englands Monetary Analysis Directorate, Bank of England Quarterly Bulletin, 2014 Q2.
10
See Productivity and the allocation of resources, Ben Broadbent, External Member of the Monetary Policy
Committee, Bank of England, 12 September 2012.
24 February 2015
90
productivity growth stalls. The financial sector is often cited as one that is likely to have
persistently slower productivity growth following the recession. A tighter regulatory
environment and higher capital requirements have raised the cost of capital and
necessitated more spending to cover infrastructure, system, and regulatory requirements.
The new regulatory environment will likely mean trend productivity growth in the financial
services sector will be persistently lower relative to pre-2008 levels. Research from the Bank
of England estimates that slower financial sector productivity growth could account for
about half (eg, 1pp) of the decline in trend productivity.
UK potential growth has been
supported by a surge in trend
hours worked
The other piece of the productivity puzzle, in terms of estimating the net effect on trend
potential growth, is the contribution from labor. The trend growth of hours worked in the UK
has provided an important offset to the slowing in trend productivity. Growth in the labor
force, due to a steady trend participation rate and growing population, along with a rapid
boost in trend employment following the recession (Figure 7), has provided important
support for trend output growth. In addition, trend growth in average working hours has
turned positive for the first time since the mid-1970s. Together, these have caused trend
hours to rise to 1.5% in the post-crisis period. However, growth in trend hours has not been
enough to fully offset the sharp slowing in productivity, and trend output growth fell to 1.8%
between 2000 and 2009 and to 1.6% in the post-recession period from 2010 to Q 2014.
The model results for Germany clearly show the effects of reunification in the early 1990s:
data prior to 1990 are from West Germany and post-reunification data include both East
and West Germany. The surge in population from reunification leads to an artificial
recession. The burst in potential labor contribution boosts trend growth and the business
cycle framework accounts for this by estimating a positive output gap prior to reunification
and a negative output gap immediately afterward. In terms of the effect on the trend,
reunification pushes trend output per hour down discretely in 1991 and then the series
resumes its trend growth thereafter. We suggest interpreting the artificial recession and
the results for 1990-93 with caution. We exclude these years from the data and focus our
attention on the remaining sample period.
Like other countries in our developed economy sample, output per hour in Germany has
slowed in recent years, but we find that trend productivity growth did not decelerate as
sharply in Germany as it did in the US and UK in 2001 and 2002, respectively. Productivity
growth was estimated at 1.8% annually in 1980-89, 1.2% per year from 2000-09, and 1.0%
from 2010-14 Q1.11 Output per hour slowly accelerated from just under 1.0% per year in the
FIGURE 8
German labor market reforms boosted trend employment
% y/y
%
98
2.5
2.0
1.5
1.0
0.5
0.0
-0.5
-1.0
-1.5
-2.0
-2.5
-3.0
96
94
92
90
88
93
FIGURE 9
with consistent employment rate growth since 2005
96
99
02
05
08
11
14
72 75 78 81 84 87 90 93 96 99 02 05 08 11 14
Germany: Employment trend growth
Our estimates of the decomposition of potential growth closely match those of the Council of Economic Experts.
See Peter Bofinger, Lars Feld, Christoph Schmidt, Isabel Schnabel, and Volker Wieland,Mehr Vertrauen in
Marktprozesse, Jahresgutachten 2014-2015.
24 February 2015
91
early 1990s to 1.4% per year in 2000. Thereafter, trend productivity growth slowed and
reached 0.9% y/y in recent years. Overall, the timing of the productivity slowdown matches
that of the US and UK, but the amplitude of the peak to trough decline has been more muted.
As in the UK, however, the deceleration in productivity has been matched by a rise in trend
employment growth (Figure 8). The trend employment rate (eg, one minus the long-run
unemployment rate) has trended steadily higher since 2005 and now stands at a multidecade high. Except for a brief period during 2009, likely an effect of the global recession,
year-on-year growth in trend employment has remained in positive territory for the past
decade (Figure 9). In addition, and in contrast to many of its developed economy peers,
Germanys trend labor force participation has been on a steady upward path (Figure 10),
rising by just over 2pp since end-2001. Together, faster growth in trend employment and
participation added about 0.6pp to potential growth in Germany over the past decade.
Extensive labor market reform
boosted trend employment
and participation in
Germany
In our view, the model results likely reflect the Hartz reforms to the German labor market
enacted between 2002 and 2005. In response to a steadily rising unemployment rate over
several decades, Germany implemented a series of wide-ranging reforms to improve the
efficiency of labor markets with the aim of lowering unemployment, reducing the duration of
unemployment, and curbing unemployment benefits as part of an overhaul of the benefit
system. The Hartz reforms are generally credited with boosting employment and participation
rates, particularly among women, while leading to a reduction in long-term unemployment.
Our model estimates confirm these findings. As Figures 8 and 10 show, the acceleration in the
growth rates of trend employment and participation occurred after Hartz reforms were
implemented. We find that, on net, long-term unemployment (NAIRU) fell by 2pp by end2013, in line with estimates from other sources (Figure 11).12
Labor market reform that boosted trend employment and participation, however, was
unable to cause potential GDP growth to accelerate because average working hours in
Germany have been on a steady decline. We find that trend working hours subtracted 0.6pp
and 0.5pp from potential growth in 2000-09 and 2010-Q1 2014, respectively. This, together
with a gradual slowdown in productivity, left potential GDP growth largely unchanged.
FIGURE 10
German labor market reforms boosted trend participation
% of total population
55
FIGURE 11
with a surge in participation between 2002 and 2005
%
12
54
10
53
52
51
50
49
4
48
90
93
96
99
02
05
08
11
14
93
96
99
02
05
08
11
14
Germany: NAIRU
Source: Barclays Research
12
See Tom Krebs and Martin Scheffel, Macroeconomic Evaluation of Labor Market Reform in Germany, IMF
Working Paper 13/42, February 2013. The authors find that long-term (non-cyclical) unemployment was reduced by
1.4pp due to the Hartz IV reforms. Our results indicate that long-term unemployment initially rose in Germany after
reforms were implemented, but then fell after 2005. On net, we find NAIRU fell by about 2pp relative to late 1990s
levels.
24 February 2015
92
Following the recession, a likely contributing factor to soft productivity growth in France has
been lackluster business investment. Since 2010, gross fixed capital formation in France,
which includes public, private (financial companies and nonfinancial corporations), and
household entities, has grown by only 1.2% per year on average and contributed less than
0.1pp to real GDP growth. In level terms, gross fixed capital formation still stands nearly 10%
below the pre-recession peak in Q4 2007. Standard economic theory suggests the behavior of
business investment is influenced by long-run factors like potential GDP growth and short-run
cyclical economic factors, including the rate of growth in economic activity, credit conditions,
and uncertainty.13 The sluggish domestic economic recovery and heightened uncertainty
stemming from the episodic concerns about sovereign debt sustainability in Europe are likely
to have weighed on business sentiment, as have poor corporate profitability. As Figure 13
shows, corporate profitability in France has declined steadily following the recession, reducing
the ability of the nonfinancial corporate sector to engage in internally financed investment.
Declining corporate profitability has also been a feature of Italys economy in the past decade,
whereas trends in corporate profitability in Germany and Spain have been more favorable.
FIGURE 12
France: Trend productivity growth
FIGURE 13
Weak corporate profitability has constrained investment in
France and Italy
%, y/y
1999 = 100
7.0
6.0
5.0
4.0
3.0
2.0
1.0
0.0
-1.0
-2.0
-3.0
140
130
120
110
100
90
80
75 78 81 84 87 90 93 96 99 02 05 08 11 14
France: Output per hour trend growth
99 00 01 02 03 04 05 06 07 08 09 10 11 12 13
France
Spain
Italy
Germany
See France: IMF Selected Issues, IMF Country Report No. 14/183, July 2014 for discussion of business investment in
France. Also, see Eugenio Pinto and Stacey Tevlin, Perspectives on the recent weakness in investment, FEDS Notes,
May 21, 2014 for an analysis of accelerator and long-run growth models applied to US investment
24 February 2015
93
FIGURE 14
France: year-on-year growth in trend average working hours
%, y/y
FIGURE 15
France: Modest rise in NAIRU following the recent recession
%
12
1.5
1.0
0.5
0.0
-0.5
-1.0
-1.5
-2.0
-2.5
-3.0
0
75 78 81 84 87 90 93 96 99 02 05 08 11 14
76
80
84
92
96
00
04
08
12
France: NAIRU
88
Weak productivity growth, however, is not the sole factor behind Frances slowing in trend
GDP growth. Since 1980, average working hours have been on a downward trend (Figure
14), dragging potential growth 0.8pp lower in the two decades ending 1999, and somewhat
less since then. The downtrend in average working hours may also be related to the
structural shift away from goods production and toward services, where part-time
employment and shorter-work weeks are more prevalent. We also find that a modest
increase in structural unemployment has occurred, with NAIRU rising from an average of
8.4% in 2006-07 to 10.5% now. This rise in structural employment and decline in trend
working hours has, on average, offset population growth and meant that total hours have
been approximately neutral in terms of contribution to GDP potential. However, we find the
rise in structural unemployment in France has been much more modest than in either Spain
or Italy, as discussed further below.
A high labor tax wedge in
France remains a constraining
factor on growth
France also has a fairly high tax wedge, or the difference between before-tax and aftertax wages. A high tax wedge translates into high labor costs for employers and low net
take-home pay for employees. High tax wedges are generally associated with higher
structural rates of unemployment, lower hours worked, and lower productivity. 14
According to OECD estimates, the tax burden in France has risen from 49.6% in 2000 to
50.1% in 2005, well above the 37.3% average for OECD countries and higher than the
European average of 42.1% as of 2005. 15 The European average, however has drifted
modestly lower in recent years.
Our estimates of potential GDP and its components in France are similar to those found
elsewhere, including in two recent IMF studies that find potential output grew at an average
rate of more than 2% during the 1980s and 1990s, but decelerated to around 1.7-1.8% in the
2000s before the crisis.16 During and after the crisis, IMF staff found that potential output fell
to below 1%. Across both exercises, Fund staff use a variety of methodologies, including
statistical filters, production function approaches, and a multivariate approach similar to the
one used in this analysis. The authors conclude that a multivariate approach provides more
robust estimates than the remaining approaches, although none of the approaches is fully
robust to data revisions and uncertainty about the true level of potential output should be an
accepted fact of life for policymakers and investors.
14
Hong Ding, Can tax wedge affect labor productivity? A TSLS fixed model on OECD panel data, International
Journal of Applied Econometrics and Quantitative Studies, Vol. 5-1, 2008.
15
See Tax wedges on earning vary sharply in OECD countries, OECD. The tax burden is measured as income tax plus
employee and employer contributions, less cash benefits, as a % of labor costs. Data is for single persons without
children at 100% of average earnings.
16
See France: Selected Issues, IMF Country Report No. 11/212, July 2011 and France: Selected Issues, IMF Country
Report No.13/252, August 2013.
24 February 2015
94
FIGURE 16
Housing-related employment surged in Spain before 2007
%
% y/y
14
13
12
11
10
9
8
7
6
5
4
FIGURE 17
boosting trend employment and total hours
Spain: Construction and real estate
employment growth
15
10
5
0
-5
-10
-15
-20
1995 1997 1999 2001 2003 2005 2007 2009 2011 2013
-25
1995 1997 1999 2001 2003 2005 2007 2009 2011
Construction/real estate
Remaining sectors
FIGURE 19
and structural unemployment in Spain has trended higher
4.0
%
30
3.0
25
2.0
20
1.0
15
0.0
10
-1.0
-2.0
0
95
98
01
04
07
10
13
96
99
24 February 2015
02
05
08
11
14
Spain: NAIRU
Source: Barclays Research
95
others argue that sectors outside of housing exhibited poor productivity and Spains rigid dual
labor market limited flexibility and kept inefficiencies high.17
Following the recession, we
find that potential growth in
Spain has fallen to -0.3%
Following the recession, we find that potential growth has fallen to -0.3%, modestly below
IMF and OECD estimates and in line with estimates from the European Commission.18
Population growth has slowed in recent years and is likely to be a feature of the Spanish
economy in the years ahead, as the working age population is projected to decline. In
addition, we find that NAIRU has risen substantially and, together with demographic trends,
mean labor force participation and trend total hours are major constraints on potential
GDP. We find that NAIRU increased sharply from just under 10% in 2008 to more than 25%
currently (Figure 19). Only recently has the rise in trend structural unemployment begun to
moderate. The shedding of employment, in our view, is the main reason trend productivity
exhibited a medium-term bounce in 2007-13. As Figure 18 shows, productivity growth rose
to 3.5% in 2010 before falling back to 0.7% in 2013.
In contrast to the more dramatic turn of events in Spain and elsewhere that reflect more of a
boom-bust phenomenon, Italy shows the signs of an economy limited by structural rigidities
and inefficiencies. We find that potential growth in the second half of the 1990s was a modest
2.0%, with half coming from a trend increase in labor force participation and half from
productivity gains. Post-recession, we find a fairly sharp reduction in both productivity growth
and hours, with productivity growth negative, on average, since the beginning of the last
decade. This result is similar to estimates of trend growth in total factor productivity from the
OCED, which shows productivity in Italy declining by 0.4% per year in 2001-10.19
Labor force participation has trended steadily higher throughout the sample period, rising
from nearly 56% of the total population to around 63% in 2013, although most of this
increase took place prior to the recession (Figure 20). Previous labor market reforms the
Treu reform in 1997 and the Biagi reform in 2003 provided for non-standard work
arrangements and part-time employment. Data indicate that these reforms were most
helpful in boosting participation among workers in the 15-24 age group and among
women. According to data from Eurostat, youth employment increased from 25% 1997 to a
high of 27.6% in 2004, while employment among women rose from 36.5% in 1999 to a high
of 47.2% in 2008. Although the reforms boosted participation, they also tended to reduce
FIGURE 20
Labor market reforms in Italy helped boost participation,
particularly among youth and women
FIGURE 21
but attachment was low and the downturn sent structural
unemployment higher
64
63
62
61
60
59
58
57
56
55
54
14
13
12
11
10
9
8
7
6
5
94
96
98
00
02
04
06
08
10
12
14
94
96
98
00
02
04
06
08
10
12
14
Italy: NAIRU
Source: Barclays Research
17
See Lpez-Garca, P., Puente, S. and A. L. Gmez, 2007, Firm Productivity Dynamics in Spain,
Documento de trabajo No. 0739 (Madrid: Bank of Spain).
18
See Spain: Selected Issues, IMF Country Report No. 14/193, July 2014.
19
See Italy: Selected Issues, IMF Country Report No. 12/168, July 2012.
24 February 2015
96
FIGURE 22
An ageing population sent participation rates lower in
Japan
FIGURE 23
and a shift to part-time employment reduced average
working hours
%
65
1981=100
102
64
101
100
63
99
62
98
61
97
60
96
59
95
58
94
81
84
87 90 93 96 99 02 05 08 11
Japan: Trend labor force participation rate
14
81
84
87
90 93 96 99 02 05 08
Japan: Average working hours trend
11
14
average weekly hours because of the increase in temporary and part-time employment. As
in Spain, Italian labor markets exhibit a dual structure, with the core of the labor market
more rigid and inflexible and the margins youth and female employment in Italy more
susceptible in downturns. As a result, we find that structural unemployment moved sharply
higher beginning in 2007, more than doubling from 6.0% to 12.7% at present.
Japans economic performance has been widely studied and our findings correspond with
others, including official sources.20 We find that potential growth slowed significantly from
nearly 4.0% in the 1980s to around 1.0-1.5% heading into the recession. In 1990-99, the
slowdown in potential growth came mainly from a reduction in trend hours driven by softer
participation (Figure 22) and a trend decline in average working hours (Figure 23).
Demographics in Japan are a clear factor in the slowing of potential GDP as the labor force
participation rate began to turn sharply lower in the mid-1990s, similar to the behavior of the
labor force participation rate in the US after 2000. We find that the trend participation rate fell
FIGURE 24
The recession led to a modest slowing in trend productivity
growth in Japan
% y/y
FIGURE 25
The trend employment rate in Japan has bounced back to
pre-crisis levels
4.5
%
99
4.0
98
3.5
97
3.0
2.5
96
2.0
95
1.5
94
1.0
93
0.5
0.0
92
81
84
87
90
93
96
99
02
05
08
11
14
81
84
87
90 93 96 99 02 05 08
Japan: Employment rate trend
11
14
See Measuring potential growth in Japan: Some practical caveats, Bank of Japan Review, February 2010. For
further decomposition of labor markets and the effect of demographics on potential growth in Japan, see The new
estimates of output gap and potential growth rate, Bank of Japan Research Review, May 2006. Our findings are also
similar to results presented by Ms. Sayuri Shirai, Japans economic activity, prices, and monetary policy
relationships between the output gap, prices, and wages, Okinawa, May 29, 2014.
24 February 2015
97
from 63.5% in the mid-1990s to 60.5% heading into the recession, roughly in line with official
Bank of Japan estimates. At the same time, Japan underwent a shift toward part-time
employment, as have most developed countries in recent decades. We view demographics,
the transition to a more service-based economy, and a decelerating trend in average weekly
hours as explaining most of the fall in potential growth heading into the recession.
Following the recession, potential GDP growth slowed further as productivity growth fell to
below 1.0%. Like many of the countries in our developed market sample, capital
accumulation slowed during the recession and trend employment and participation
declined for a relatively brief period between 2007 and 2009. Since then, however, we find
that the trend employment rate has rebounded to pre-crisis levels (Figure 25) and, in the
process, reversed the rise in NAIRU. We also find that the trend participation rate has
ticked higher since late 2012. Our multivariate framework estimates structural
unemployment at only 4.0%, down from a peak of 5.8% in 2009. Altogether, while we
estimate that potential growth has averaged only 0.4% between 2010 and Q1 2014, there is
evidence in recent years that the Japanese economy is shaking off some of the adverse
effects of the recession on employment and hours, and we do not find evidence that the
recession has severely affected trend productivity growth.
To put this number further into perspective, our finding that slower growth in developed
economies could slow global growth by 0.7pp is of similar magnitude to the effect that a
slowing China has on global growth. China, which accounts for 18.6% of world GDP on a
PPP basis, is expected to see its potential GDP growth slow from about 9-10% in the 1990s
to about 6.0% in the coming 5-10-year period as the country transitions from its previous
investment-led growth strategy to a consumption-led economy. If realized, the slowing in
Chinas potential growth would lower the growth rate of potential global GDP by 0.6-0.7pp.
The developed economies in our sample plus China account for nearly 62% of world GDP
on a PPP-adjusted basis; slower potential growth in developed economies and a
decelerating Chinese economy constitute a significant drag on global growth. Taken
together, the two forces may slow potential global growth by 1.5pp.
21
This implies that potential growth across the countries in our sample slowed by about 1.5pp. The five-year
centered moving average IMF purchasing power parity adjusted weights in 2014 were: US: 22.3%, Japan: 6.1%,
Germany: 4.2%, UK: 3.2%, France: 3.0%, Italy: 2.4%, and Spain: 1.8%. To estimate the effect on global potential, we
multiply these weights by the 1990-99 estimate for potential growth in each economy less the 2010-14 period. We
then sum across countries to yield the full estimate.
24 February 2015
98
Approximately two-thirds of
the slowdown in potential
growth occurred after 2007
We compute the change in our estimate of potential growth between 1990-99 and Q4 2007
to form an estimate for the fraction of the slowdown attributable to the recession. Taking
Q4 2007 as the cut-off date, we estimate that the pre-recession slowing in potential
developed economy output was about 0.5pp, or one-third of the total decline over the full
sample period. This represents about a 0.2pp drag on global growth based on relative PPP
weights. Consequently, about two-thirds of the decline in developed market potential
output growth came after the onset of the recession. Although it is difficult to fully isolate
the effect of the recession on trend growth from the slowing already in place prior to the
recession, we present these results as a useful starting point.
All the countries in our developed economy sample implemented countercyclical policies,
although these were mainly implemented through conventional and unconventional easing
of monetary policy.23 Advanced economy central banks responded to the crisis by lowering
target interest rates to zero (or below), providing abundant liquidity to traditional and nontraditional counterparties at various maturities, and initiating asset purchase programs in an
FIGURE 26
US output gap
FIGURE 27
UK output gap
% of potential output
% of potential GDP
3
2
-2
-1
-4
-2
-6
-3
-4
-8
67 70 73 76 79 82 85 88 91 94 97 00 03 06 09 12
US output gap
75 78 81 84 87 90 93 96 99 02 05 08 11 14
UK output gap
+/- 2RMSE
In addition, structural reforms intended to boost long-run potential growth often make near-term outcomes worse.
The example of Germany following labor market reforms of 2002-05 illustrates how labor market outcomes initially
deteriorated before later improving. Any full assessment of structural reforms must include the netting of short-term
losses against long-term improvement.
23
For a more complete listing of the policy response by global central banks to the recession, see Global themes: A
quantum shift in central bank communication, 12 September 2013.
24 February 2015
99
effort to lower interest rate term premia on safe assets and risk premiums on risky assets.
Following the ECBs most recent announcement in late January that it would launch
outright QE that included government bonds, every central bank in our sample has now
engaged all three of these policy tools to a significant degree. In addition, the extensive use
of unconventional policy tools required central banks to enhance their communication
efforts to achieve greater transmission of monetary policy into the real economy.
Accommodative fiscal policies
were more modest in scope
and ultimately reversed
Expansionary fiscal policy was used to a much lesser degree, particularly in Europe, where the
rules of monetary union prohibit significant swings in the budget balance and fears over debt
sustainability were more pronounced. Fiscal policy in the UK was countercyclical during 200809, but policy reversed course sharply in 2010 on concerns about deficits and the
sustainability of government debt.24 Expansionary fiscal policy was used early on in the
recovery in the US, but the size of the effort was relatively modest and ultimately reversed
through sequestration and the expiration of some upper income tax rate cuts. In Japan, the
first arrow of Abenomics consisted of a large fiscal stimulus bill, which policymakers
described as part of an offensive strategy to boost growth. That said, the policy framework
also used an increase in the consumption tax as part of a defensive strategy to preserve the
medium-term sustainability of the budget. Altogether, countercyclical policies in advanced
economics were generally small and front-loaded, and were either reversed or offset by other
actions in later years. In our view, the lack of coordination between fiscal and monetary policy
has limited the ability of policy to mitigate the effects of the recession on potential output.
Whether countercyclical policies are effective at facilitating the reallocation of labor and
capital across different sectors of the economy, improving the efficiency of matching
available jobs and properly skilled workers. and limiting the rise in structural unemployment
depends, in part, on how responsive the underlying economy is to the incentives created by
accommodative policies. In the context of our analysis, economies that are more dynamic
and flexible will be better able to absorb shocks and, as a result, will likely exhibit greater
cyclical amplitudes and stable trends. In other words, recessions and shocks cause the
economy to deviate from potential in the short run, but the rate of potential growth is
generally undisturbed over the long run. In contrast, economic shocks will be transmitted to
trend variables more quickly in economies that are less dynamic and inflexible. These
economies will have smaller business cycles and more volatile trend variables.
FIGURE 28
Germany ouput gap
FIGURE 29
France output gap
% of potential GDP
% of potential GDP
6.0
1.6
1.2
4.0
0.8
2.0
0.4
0.0
0.0
-0.4
-2.0
-0.8
-4.0
-1.2
-1.6
-6.0
93
96
99
02
05
11
+/- 2RMSE
14
77 80 83 86 89 92 95 98 01 04 07 10 13
France output gap
+/- 2RMSE
24 February 2015
08
The fiscal tightening in the UK was later paused due to concerns that it was choking off the recovery.
100
FIGURE 30
Spain output gap
FIGURE 31
Italy output gap
% of potential GDP
% of potential GDP
0.8
0.8
0.6
0.6
0.4
0.4
0.2
0.0
0.2
-0.2
0.0
-0.4
-0.2
-0.6
-0.4
-0.8
-0.6
-1.0
96
99
02
05
08
11
+/- 2RMSE
14
96
99
02
05
11
08
14
+/- 2RMSE
In Figures 27-31, we present the estimates of the output gaps for six of the seven developed
economies in our sample. We find that the amplitude of the business cycle in the US, UK,
and Germany is larger than in France, Spain, and Italy, including in the most recent
recession.25 One interpretation of these results is that the recession was characterized by
large reductions in aggregate demand and smaller reductions in productive potential in the
US, UK, and Germany. When applied to the other three economies, the model estimates
could be interpreted as suggesting either the large cyclical shortfall in aggregate demand
was quickly transmitted to trend variables and lower potential output, or the shock itself
was a supply-side disturbance that could be immune to countercyclical policies.
The results have significant implications for the ability of policy to mitigate a recession-related
decline in potential GDP. Conventional countercyclical monetary and fiscal policies are likely to
be more effective in the US, UK, and Germany if they are successful in quickly reversing the
decline in aggregate demand. Dynamic economies with more flexible labor and product
markets are likely to be more responsive to activist policies. This argument has been made
explicitly by the Federal Reserve to justify its aggressive policy stance as a way to limit the
amount of supply side-damage that occurred initially following the downturn, and potentially
to help reverse a portion of the damage at a later stage.26 That the output gaps in these three
economies have closed suggests policy has had success in reversing the shortfall in aggregate
demand and ameliorating some of the damage done to long-term productive potential.
In the remaining economies, including Japan, where the model estimates indicate a more
rapid transmission of the economic downturn into trend variables, the results validate the
emphasis on appropriate structural reforms to complement countercyclical policy. If
successful, these policies would reduce structural unemployment, raise participation rates
to boost the size of the labor force, increase hours, encourage capital accumulation,
rebalance capital and labor to more efficient uses, and boost productivity. Significant reform
agendas are already under way in several countries, including:
25
Our findings for the size of the output gap in the US, UK, and Germany are similar to those of the Federal Reserve,
IMF, and OECD. However, the European Commission, IMF, OECD, ECB, our Barclays European economics research
team, and others find wider output gaps in France, Italy, and particularly Spain, where other studies find output gaps as
large as 6% during the boom and -4% thereafter (see Borio, C., P. Disyatat, and M. Juselius, Rethinking potential
output: Embedded information about the financial cycle, BIS Working Papers No. 404, 2013. The differences may be
methodological in that traditional HP filters, bandpass filters, and other similar techniques used to estimate potential
growth often assume a smooth trend, whereas the multivariate approach we apply in this chapter does not. We offer
interpretations for what our findings could imply without seeking to validate one estimation approach over another.
26
See Aggregate supply in the United States: Recent developments and implications for the conduct of monetary policy,
David Reifschneider, William Wascher, and David Wilcox, Finance and Economics Discussion Series 2013-77, 2013.
24 February 2015
101
Spain. A series of structural reforms has been implemented with the goal of
strengthening the financial system, increasing the efficiency of public services,
improving competitiveness, and lowering regulatory barriers, among others.27 A highly
fragmented labor market remains an issue, as does low productivity.
Italy. Reforms to liberalize product markets and improve competitiveness began in 2011
and 2012 in energy, transportation, professional services, and public services.28 Current
reform proposals are aimed at making labor markets more flexible. As in Spain, a dual
labor market structure remains an obstacle.
France. In 2013, France passed a labor reform law intended to improve mobility, allow
for more flexibility to adjust pay and hours in response to changes in the business cycle,
and streamline the dismissal procedure. In addition, pension reform is anticipated to
raise labor force participation rates over the long run. If realized, it would help offset a
less favorable demographic environment where growth in the labor force is expected to
slow to 0.2% per year between 2021-2030.
Japan. Reforms comprise the tri-arrow policies of aggressive monetary easing (1st
arrow), expansionary fiscal policies (2nd arrow), and structural reforms (3rd arrow), with
the last including efforts at electricity sector reform, governance and investment
reforms at the Government Pension Investment Fund, coordinated wage setting, and
other changes to increase participation and reduce fragmentation in labor markets.
Although structural reforms
hold much promise
Efforts on the structural reform front are bearing fruit, particularly in Spain, where real
output grew for five consecutive quarters through Q3 2014 and the unemployment rate has
fallen 2.5% from its peak. Despite this progress, the legacy of the recession persists, with
the unemployment rate at 23.7% and approximately 3.5m persons (15% of the labor force)
unemployed for over a year. Even under a decidedly optimistic scenario of productivity
growing at twice its pre-crisis rate and NAIRU falling to 14% by 2019, the IMF finds that the
unemployment rate would still be 16.0%.29 Turning to Italy, IMF staff estimate that a
simultaneous implementation of product and labor market reforms would lift potential
growth by about 0.8-9pp annually relative to baseline assumptions, recovering about half of
our estimate of the decline in Italys potential growth since 1994-1999. Finally, IMF staff
estimate that potential growth in France could rise by 0.7pp if appropriate structural
reforms are enacted.
In Japan, the first and second arrows have supported economic activity and inflation, but
progress on the third arrow has been slower. IMF staff estimate that potential growth is likely
to remain below 1.0% through 2017. Against our estimate of potential output growth of
0.4pp per year between 2010 and Q1 2014, IMF estimates imply that full adoption of third
arrow policies may improve trend growth by 0.5-0.6pp on a 5-10-year horizon.
27
28
29
24 February 2015
See Spain: Article IV Consultation, IMF Country Report No. 14/192, July 2014.
See Italy: Selected Issues, IMF Country Report No. 12/168, July 2012.
See Spain: Selected Issues, IMF Country Report No. 14/193, July 2014.
102
= + + 2
= + 10 + 3
= + 10 + 4
where GDO* represents the common trend component of GDP and GDI (eg. potential
output) and NFBO* the common trend between NFBP and NFBI. 32,33 The framework
assumes the residuals are measurement errors which can be decomposed into a sum of a
common component and idiosyncratic components.
30
24 February 2015
Our approach follows Charles Fleischman and John M. Roberts, 2011, From many series, one cycle: Improved
estimates of the business cycle from a multivariate unobserved components model, Finance and Economics
Discussion Series 2011-46; and Jun Ma and Mark Wohar, An unobserved components model that yields business and
medium run cycles, August 2012.
31
Arabinda Basistha and Richard Startz, 2008, Measuring the NAIRU with reduced uncertainty: A multiple-indicator
common-cycle approach, Review of Economics and Statistics, 90, 805-11. Also see James H. Stock and Mark W.
Watson, 1989, New indices of coincident and leading economic indicators, NBER Macroeconomics Annual 1989,
Oliver Blanchard and Stanley Fischer, eds., 351-394.
32
The cycle is assumed to be a stationary AR(2) process equal to = 1 1 + 2 2 + .Typically 1 > 0
and 2 < 0 which implies the cycle is hump-shaped in response to a shock. The sum of the coefficients is assumed
to be close to 1, but less than 1, meaning the business cycle is persistent.
33
Since NFBO is not the same as GDO (since it exclude the farm and public sectors), =1 cannot be assumed for a
contemporaneous, normalized cycle. We estimate = 10 and assume it is the same across both variables with the
prior that nonfarm business output likely has larger amplitude than GDO since the latter includes the public sector.
103
Potential output and nonfarm business output can be further broken down into
= +
= +
= +
= +
= +
where OSR* is the output sector ratio between gross domestic output and nonfarm
business output, HNFB* is the trend of total working hours, OPH* is the trend of output per
hour or productivity, ENFB* is the trend in total employment, WW* is the trend of average
working hours, ER* is the employment rate, and LP* is the labor force participation rate.
ECPS* is the trend in employment from the current population survey and ESR* is the
employment sector ratio between total employment and the current population survey.
The observed data on employment, the work week, the employment rate, and participation
are broken down into the sum of a trend and cyclical components
= + 20 + 21 1 + 22 2 + 5
= + 30 + 31 1 + 32 2 + 6
= + + 40 + 41 1 + 42 2 + 7
= + 50 + 51 1 + 52 2 + 8
where the framework allows for some deviation between shocks to output and the response
of employment hours and labor force participation. The rationale for this specification would
include adjustment costs; whereby firms find it costly to adjust the factors of production so
that changes in labor market activity may lag changes in output.34 The introduction of state
emergency and extended benefits (EEB) following the rise in long-term unemployment during
the crisis is allowed to influence employment and participation, but not the cycle.35
Finally, the Phillips curve is given by
= ()1 + 11 ()1 + 12 ()85 1 + 2 ()
+( [ + + 1 ]) + 9
where DCPIX is core CPI inflation, drpe is the relative change in consumer energy prices,
drpi is the change in the relative price of imports, d85 is a dummy from 1985 to the present
to account for rising share of the import ratio in consumer spending, and (L) represents
lagged values.36 The inflation equation also assumes that cyclical deviations in output from
its trend affect inflation and the employment rate gap is adjusted to account for extended
and emergency unemployment benefits.
34
As referenced in Fleischman and Roberts (2011), previous research suggests the unemployment rate, and therefore
the employment rate, should be a lagged indicator of the business cycle whereas employment is considered a
contemporaneous variable. These results could be imposed as explicit model assumptions, but we choose to let the
data show whether this is the case.
35
Like Fleischman and Roberts (2011), we measure EEB as the ratio of total quarterly payments of federal and state
emergency and extended benefits programs to the four-quarter moving average of total private wages and salaries.
36
We use ten lags of core CPI and six lags on the relative price of energy. We constrain the sum of the coefficients on
lagged inflation to be equal to one and, in doing so, impose a unit root process to pin down the trend inflation rate.
24 February 2015
104
0
0
0
0
0
1
= 1
0
0
0
0
0
0
0
0
0
0
0
()
1
1
1
1
1
0
0
0
1
1
1
1
0
1
0
0
1
1
1
1
0
0
1
0
1
0
0
0
0
0
0
0
1
1
1
10
1 10
1 20
+
0 30
0 40
0 50
40
0
0
0
0
1
0
0
0
0
0
0
0 0
0
0
0
1 + 0
0
0
0
85 0
0
0
0
0
0
0
0
0
0
11 () 12 () 2 ()
0
0
0
21
31
41
51
41
1 1
1 0
0 0
0 0
0 0
0 0
0 0
0
0
0
22
+
32 1
2
42
52
42
0
0
1
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
1
0
1
0
2
0
3
0
4
0
5
0
6
7
0
8
1
9
The US model is estimated using quarterly data from 1963 Q1-Q1 2014 using maximum
likelihood techniques in the state-space model estimation framework in Eviews, which uses
the Kalman filter to estimate model coefficients while using numerical methods to ensure
fitted values are close to observed data.
= + + 2
Where GDO* is potential output and cyc is the output gap. As before, the measurement
errors can be decomposed into the sum of a common component and idiosyncratic
components.
Potential output is further decomposed into
= +
= +
= +
Where HGD* is the trend of total working hours for the economy, OPH* is the trend in
output per hour (eg, productivity), EGD* is the trend of total employment, WW*, is the
trend of average working hours, ER* is the trend employment rate, and LP* is the trend
labor force participation rate.
The observed data on the work week, the employment rate, and participation are
decomposed into the sum of a trend and cyclical components
= + 10 + 11 1 + 12 2 + 3
= + 20 + 21 1 + 22 2 + 4
= + 30 + 31 1 + 32 2 + 5
24 February 2015
105
and the cycle is assumed to be a stationary auto-regressive process.37 Finally, the Phillips
curve is for non-US developed economies is given by
= ()1 + (20 + 21 1 + 22 2 ) + 6
which has a similar interpretation to the US specification in that it is assumed that cyclical
deviations in output are allowed to affect inflation, creating a natural rate interpretation.
The six equations can be represented in the following measurement equation
1
1
0
+
0
0
0
1
1
1
0
0
0
1
1
0
1
0
0
1
1
1
1
0 10
0 20
1 20
0 20
0
0
11
21
31
21
0
0
12
22
32
22
1
0
0
0
0
0
1
0
0
1
1 0
0 0
3
0 + 0 1 + 4
0 0
5
6
0 ()
The models are estimated using quarterly data for the UK (1975 Q1-Q1 2014), France
(1975 Q1-Q1 2014), Germany (1973 Q1Q1 2014), Italy (1993 Q1-Q1 2014), Spain (1996
Q1-Q1 2014), and Japan (1981 Q1-Q1 2014), using maximum likelihood techniques in the
state-space model estimation framework in Eviews.
37
24 February 2015
106
CHAPTER 5
Banking regulation has intensified since the financial and sovereign crises in a global
effort to improve the safety and stability of the financial system. Regulators have
forced banks to change their capital structures and their business models to enhance
the safety of the banking system and make future financial crises less likely.
These new regulations have materially improved the stability of the financial system.
However, in an effort to reduce the risk of future fire-sales financed by short-term
debt, they have also reduced the supply of safe, short-term, liquid assets such as
repurchase agreements, causing them to trade at lower yields (and, by extension,
higher prices).
brian.monteleone@barclays.com
Eric Gross
+1 212 412 7997
eric.gross@barclays.com
The reduction in the supply of short-dated safe assets and associated fall in the
liquidity of fixed income markets has created incentives for investors to look to nontraditional sources of liquidity, such as ETFs and mutual funds. In turn, this may
result in a transfer of fire-sale risk into assets such as leveraged loans and
investment grade and high yield bonds, as liquidity in the underlying investments of
these funds deteriorates, exposing end-investors to run risk.
Conor Pigott
+1 212 412 3441
conor.pigott@barclays.com
Joseph Abate
+1 212 412 7459
joseph.abate@barclays.com
A changing landscape
Before the crisis that erupted in 2007, many banks operated with too little equity and were
overly reliant on short-term wholesale financing, such as repo, or repurchase agreements, to
fund illiquid investments. When the crisis began, these banks did not have the capacity to
absorb losses, given their limited capital base. Regulators have addressed this by forcing all
banks to significantly increase their capital ratios, which are now higher than at any time since
World War II. Excessive reliance on short-term financing exposed some banks to destabilizing
runs when investors pulled their financing as the crisis began to mount, contributing to
failures. More important from a systemic point of view, this precipitated the fire-sale of assets
financed by short-term debt, driving down the prices of specific assets. This contributed to
system-wide funding issues, even for banks with relatively strong balance sheets. To reduce
the risk of future fire-sales, several of the new initiatives have targeted repo and other shortterm liabilities, resulting in a more than 50% reduction in repo balances relative to their peak.
In particular, the Volcker Rule was introduced to address illiquid and riskier investments that
had burgeoned in the banking sector before the crisis.
Whether these steps will be sufficient to curb future crises remains an open question. But it is
clear that the new regulatory environment has materially improved the stability of the system.
The best evidence of the effect of new regulations on banks probably comes from the credit
market, where the spreads of bonds issued by the largest banks have narrowed significantly
and, in many cases, are now tighter than industrial spreads. In other words, bond investors
believe bank safety has improved so much that they are once again willing to accept low
spreads for bank risk1.
Less well understood are the broader effects of improved stability on investors and the
economy. Last year, we wrote that decreased bank lending was one potential implication (see
The cost of evolving bank regulation, 13 February 2014). This year, we focus on the
implications of two separate, but related, changes in financial markets.
This argument is bolstered by the fact that banks arguably benefited from implicit government support (ie, bail-outs
in the event of a disruption) pre-crisis, causing their credit spreads to be artificially low. Subsequent changes to
regulation have likely reduced or eliminated the extent to which banks will benefit from bail-outs in any future crisis,
which would bias bank spreads wider absent the improvements in credit quality that we cite.
24 February 2015
107
The reduced size of the repo market. This large, but relatively esoteric, part of the
financial market is used by hedge funds and banks to finance securities and by money
market funds to invest cash.
We believe there are two broader implications that are more likely to be disruptive, particularly
once (if) interest rates begin rising. First, the decline in repo has reduced the supply of safe,
short-term assets. Relatively few assets fit this description: Treasury bills, bank deposits, and
repo. The reduction in repo is happening as Treasury bill supply is shrinking and banks are less
willing recipients of deposits, given lackluster loan demand. As overall supply of such assets
declines, we believe investor demand for them is relatively inelastic and a function of financial
wealth, which has been rising. We expect excess demand for short-dated safe assets to cause
them to trade at lower yields (ie, higher prices), even as and when interest rates begin to
normalize. This applies to deposit rates, which we believe will lag any rate hikes, such as we
expect in the US later this year, as investors remain willing to accept low interest rates to
maintain a base of liquid assets. Similarly, money market funds may need to accept lower
rates to remain invested.
Second, reducing the supply of these safe, short-dated assets creates incentives for investors
to look to non-traditional sources of liquidity. Migration to seemingly liquid alternatives has
happened before: in the pre-crisis period, safe short-dated assets were in limited supply
(relative to financial wealth) because of the tremendous run-up in equity prices. The result
was a massive spike in CP, repo on structured assets such as ABS and CDOs, and auction rate
securities, all of which purported to offer the daily liquidity investors were seeking. But this
liquidity dried up once the crisis began.
For various reasons, the same alternatives will not be chosen this time around: the changes in
regulations, investors collective experiences with those investments, and the simple fact that
many of them no longer exist. However, there have been increased flows in other vehicles that
offer daily liquidity, such as ETFs and mutual funds. The desire for liquidity may also be
limiting demand for closed-end fixed income funds, which would seem a natural response to
the decline in fixed income liquidity.
The inflows into ETFs and mutual funds are happening just as liquidity in the underlying
investments that these funds purchase is deteriorating. This has raised new concerns about
retail runs and fire-sale risks in such assets as leveraged loans and investment grade and
high yield bonds, where either liquidity has dropped most severely and/or where the funds
offering daily liquidity have grown the most. Ironically, these new fire-sale risks have arisen in
part because the risks of a repo-driven fire-sale have fallen. The well intentioned and arguably
successful efforts to make the banking system more robust and less susceptible to runs have
transferred fire-sale risk out of the banking system and into the hands of end-investors.
Repo 101
A repurchase transaction (repo) is effectively a collateralized short-term (often overnight)
loan. For example, an investor looking to borrow money pledges a security (eg, a Treasury)
as collateral, and receives cash. The next day, the investor pays back the cash plus interest,
24 February 2015
108
and receives his or her collateral back in return. A reverse repo is the same transaction but
viewed through the lens of the lender.
Repurchase transactions have several important aspects. First, although much of the repo
market is overnight, term repo, which can be measured in weeks or even months, is also
possible. The structure is the same, but the collateral is not returned (and the loan paid off)
until the end of the term. The second aspect is the interest rate of the transaction, which
depends on the term and the specific collateral involved. For various reasons, some collateral
may be specifically desirable to lenders and thus command lower interest rates. The final key
dimension is the haircut which defines just how much cash the borrower gets for the
collateral. This is quoted in terms of a percentage of market value. Higher-quality collateral,
such as Treasuries or agency debt, typically requires the lowest haircuts, eg, 2%. This means
that it is possible to borrow $98 for every $100 of Treasuries that the borrower pledges as
collateral. Lower-quality collateral (eg, corporate bonds) typically requires higher haircuts.
Banks engage in repo transactions for two related reasons. First, repos match cash-rich
investors (such as money market funds) with investors (such as hedge funds) who own
securities but need financing. This is done via a matched book banks engage in reverse
repo transactions with hedge funds, lending them money collateralized by securities. Banks
then borrow from money market funds via repo transactions, collateralized by the same
securities. The banks effectively act as middlemen, with the cash flowing from the money
funds to the hedge funds, and the collateral moving in the opposite direction. The second
reason banks engage in repo is to finance their own portfolio of securities, essentially playing
the role of the hedge funds in the matched book example above.
However, the same features that make reverse repo a safe asset for money funds make repo
a risky liability for leveraged investors and banks. At the slightest hint of trouble with either
the collateral or the borrower, the funding can be withdrawn, which is as simple as not
renewing an expiring contract. For example, if the collateral is downgraded, it may become
harder to borrow against. Similarly, if the borrower (eg, hedge fund or bank) deteriorates in
some way such that money funds or other lenders question its credit quality, borrowers
may have a harder time securing funds regardless of the quality of their collateral.
Essentially, borrowers reliant on repo need to continually roll over their financing, and are
exposed to the risk of a run as a result, similar in concept to a deposit run.
This presents two concerns for regulators. First, banks finance significant securities
portfolios via repo, and thus there is risk that an individual bank would need to liquidate
assets in response to being locked out of the repo market a pre-default fire-sale. This is
problematic because the highest-quality assets are the easiest to sell Treasuries, agencies,
etc. A bank that was overly reliant on repo financing of lower quality securities and faced a
repo run could be forced to sell assets quickly to raise liquidity, potentially driving down
their market valuations and leading to asset write-downs that would impair capital, and
increase the banks risk of default or downgrade. It might also need to sell assets or draw
down on its cash holdings to meet increases in haircuts on the collateral it is pledging.
24 February 2015
109
Such a run could affect multiple (or even all) banks at once if bank credit quality deteriorated
across the board, or if the entire repo market experienced a disruption. This could be caused
by a systemic shock leading to a crisis of confidence in the broader financial system. In this
scenario, with multiple borrowers trying to liquidate assets, the market could experience a
fire-sale the prices of certain assets could plummet because of a large number of forced
sellers trying to liquidate at once. This could be exacerbated by money market funds, which
are often legally prohibited from owning the types of collateral underpinning their repo trades
and would be forced to sell quickly if their counterparty defaulted and the fund took
possession of the collateral. The solvency of an individual bank could deteriorate much faster
in this scenario because it would be forced to sell assets at a loss, thereby eroding its capital. In
fact, solvency concerns could spread through the financial system.
Academic studies have described this phenomenon as a funding and liquidity spiral. 2
Asset price shocks in a particular market create funding problems for cash borrowers who
pledged the same or similar collateral. Borrowers reduce their positions by selling some of
their holdings, while their ability to borrow against their remaining assets shrinks as haircuts
increase and the value of these holdings falls in response to selling pressure. This exacerbates
the funding problems and forces more de-leveraging and asset fire-sales the process
becomes self-reinforcing.
Concerns about repo runs are not merely theoretical. Lehman Brothers failure in September
2008 serves as a real world case study. Lehmans repo book accounted for 34% of total
liabilities at 2Q08, a cursory measure of the firms dependence on short-term funding.
During normal times, this was an effective strategy for leveraging returns, but as the firms
crisis reached a climax, repo funding providers suddenly fled. Between September 9, 2008,
and September 15, 2008 (the day of its bankruptcy filing), the number of tri-party
counterparties providing Lehman Brothers with cash in exchange for securities fell from 63
to 16. Given that Lehman Brothers had used repo to fund a material volume of lowerquality, non-governmental securities the prices of which had fallen sharply the firm was
left with assets it could no longer fund in overnight markets or sell without destroying
capital, eventually contributing to the firms bankruptcy filing.
Although the Lehman experience is an important, cautionary tale, it also delineates where
the true run risk lies within the repo market. Interestingly, the financial crisis did not cause
a waterfall of repo runs across the rest of the system. Instead, the deterioration in repo
markets was more focused.
Higher-quality assets were still funded at modest haircuts: Repo haircuts did rise
during the crisis across many asset classes; however, this was generally concentrated in
funding for lower-quality ABS structures.3 Treasuries, agencies, and even investment
grade corporate bonds showed modest if any increases in margin requirements over
this period. For example, a Federal Reserve staff report indicated that U.S. Treasuries
and agencies continued to be funded in the repo market throughout this period at
haircuts of only 3% or less (ie, 97% loan to value).
See Market Liquidity and Funding Liquidity, M. Brunnermeier and L. Pedersen, National Bureau of Economic
Research working paper, December 2008.
3
See Repo Runs: Evidence from the Tri-Party Market, A. Copeland, A. Martin, and M. Walker, Federal Reserve Bank
of New York, July 2011.
24 February 2015
110
Overall, this suggests to us that repo is less flight-prone than might be imagined. Funding
terms were not markedly increased and were not in themselves the transmission
mechanism for forced sales. Furthermore, the markets for higher-quality assets that
typically serve as repo collateral were able to absorb the liquidation of Lehmans large
Treasury and agencies books, which had been funded by repo. This can be naturally linked
to the strong performance of these safe haven assets during times of turbulence,
minimizing the risk of needing to take a loss as positions are closed.
However, we must be careful not to draw too much comfort from the experience of the
crisis, given the unprecedented intervention in markets by the Federal Reserve and other
central banks, which may have helped stem further contagion. The core issues around
funding long-term, price-sensitive assets remain entities using short-term funding (such as
repo) need to mark their assets to market and obtain new funding every day. A temporary
price decline has the potential to wipe out a firms margin and force it to sell its assets. This
could in turn push asset prices lower, forcing other participants to sell and perpetuating the
cycle. We believe that it is this risk of a waterfall of forced sellers destabilizing the broader
system that regulators are attempting to address via repo-targeted reform.
Volcker Rule
The introduction of leverage ratios
SIFI buffers
Haircuts
Volcker Rule
Banks trading operations historically served two main purposes: 1) providing liquidity to
market participants wanting to buy or sell securities in exchange for a bid-ask spread; and
2) using the banks balance sheet to generate profit from price movements. Bank regulators
grew concerned that proprietary trading positions created undue risks on banks balance
sheets. In response, the Dodd-Frank Act created the Volcker Rule, which prohibits
proprietary trading. Among other things, the rule limits banks ability to take trading
positions capped at demonstrated market demand. In a market where demand from
clients, customers or counterparties is expected to diminish, this limits a banks ability to
intermediate the market. Notably, regulators chose to exempt Treasury and municipal
securities from these restrictions.
Leverage ratios
Pre-crisis, the most important (and binding) regulatory capital ratios banks needed to meet
were based on risk-weighted assets. Safe assets, such as repo, were assigned low risk weights,
and thus banks were required to allocate very limited capital to those types of positions. As a
result, there were few practical limitations on the size of bank balance sheets, which expand as
banks increase the size of their matched-book repo positions.
This has changed in both the US and Europe. Regulators in the US have adopted a 5%
supplementary leverage ratio for the holding companies of the systemically important US
banks. 4 This represents a materially stricter requirement than the old US standard, as it raises
the hurdle from 4% and expands the scope to capture some off-balance sheet assets. This rule
4
Technically, the proposed higher supplementary leverage ratio requirement would apply to all banks in the US with
at least $700bn in assets and/or $10trn in assets under custody, which at present captures the eight US G-SIBs: Bank
of America, Bank of New York, Citigroup, Goldman Sachs, JPMorgan, Morgan Stanley, State Street, and Wells Fargo.
24 February 2015
111
complements existing risk-weighted capital measures by ensuring that even low-risk assets
and certain off-balance-sheet exposures are backed by material equity capital if exposures are
large enough (see Leverage ratio: An attack on repo?).
Before the crisis, European
banks were not subject to any
restrictions on balance sheet;
thus, they naturally gravitated
toward lower risk-weighted
assets this has now changed
Prior to the crisis, European banks were not subject to any restrictions on balance sheet. Thus,
they naturally gravitated toward lower risk-weighted assets (eg, repo). This has now changed
for two reasons. First, European regulators have adopted a 3% leverage ratio and several are
moving toward an even higher standard. Second, new regulations on US subsidiaries of
foreign banks will push these banks to manage the balance sheets of their US operations more
conservatively. Previously, foreign banks US intermediate holding companies were not
required to meet US capital standards independently. However, beginning in July 2015, Section
165 of the Dodd-Frank Act will require foreign-domiciled banks to roll up all their US
broker/dealers and bank branches into a single intermediate holding company (IHC). The IHC
will then need to meet risk-based capital requirements, maintain minimum liquidity buffers,
and meet the minimum leverage ratio. The challenge of establishing an IHC is particularly
acute for foreign banks that mainly conduct broker-dealer business in the US, with limited
lending capabilities, because their balance sheets would be naturally skewed toward lower
risk-weight business (Figure 1). Based on recent data, these institutions will be under similar
pressure as their US peers to reduce size and/or increase equity.
FIGURE 1
Foreign banks account for a significant share of US broker-dealer activity
Assets of US broker-dealers ($bn)
500
450
400
350
300
250
200
150
100
50
0
GS
JPM
MS
BAC
BACR
CS
DB
UBS
RBS
BNP
Nomura
Mizuho
HSBC
RY
24 February 2015
112
Haircuts up next
Though somewhat less certain, we expect further rulemaking to address haircuts for repo
transactions. These would likely be designed to cap leverage within the repo market to
levels appropriate for the quality of the underlying collateral, see Squeezing the leverage
out, October 24, 2014. Federal Reserve Governor Daniel Tarullo has repeatedly expressed a
desire to add regulation along these lines over the past year. Most recently, in a speech at
an Office of Financial Research conference (excerpted below), he highlighted his intention
that such rules also apply outside the traditional banking sector to mitigate the risk of nonbanks building up repo leverage as banks pull back.
24 February 2015
113
FIGURE 2
Aggregate repo volumes have contracted sharply from precrisis levels
Repo outstanding ($trn)
FIGURE 3
Most large US banks have responded to SLR requirements by
reducing repo balances
Repo borrowings ($bn)
$trn
$bn
350
3Q10
-26%
3Q14
-35%
300
-8%
250
3
-44%
-37%
200
150
+51
100
1
0
2001
50
0
2003
2005
2007
2009
2011
2013
BAC
GS
JPM
MS
WFC
Although the pace of the reaction to new rules has varied, all bank management teams that
face balance sheet size pressure have taken steps to reduce their low RWA exposures. Most
recently, Goldman Sachs CFO Harvey Schwartz highlighted the companys focus on
reducing its balance sheet in response to increased regulatory clarity.
24 February 2015
114
100%
1,400
95%
1,200
Volume
Amount
400
65%
200
60%
2014
2013
2012
2011
2010
2009
2008
2007
2006
2005
70%
120%
100%
80%
Volume
Turnover
Amount
2014
1,000
600
2013
75%
140%
2012
2,000
800
2011
80%
160%
1,000
2010
85%
3,000
180%
2009
90%
200%
2008
4,000
$bn
1,600
2007
5,000
105%
2005
$bn
6,000
FIGURE 5
Volume, market size, and turnover in high yield credit
2006
FIGURE 4
Volume, market size, and turnover in high grade credit
Turnover
Transaction costs have risen at the same time. Figure 6 contains pre- and post-crisis
transaction costs in the IG and HY markets, estimated using our Liquidity Cost Score (LCS)
methodology5. Transaction costs have increased in both markets. Although the change in
HY is notable, at more than 20%, the change in IG is more marked. We think this is the
result of the substantial strength of pre-crisis liquidity in that market. Note that the change
in LCS is more severe than that in bid-offer this is driven by an increase in the average
duration of the IG market over the past several years. The same average bid-offer spread
corresponds to a higher transaction cost for a longer-duration bond.
FIGURE 6
Transaction costs, today versus the pre-crisis period
1/31/2007
LCS (%) Bid-Offer
1/31/2015
Change
LCS
Bid-Offer
LCS
Bid-Offer
US Credit Corporate
0.531
8.5 bp
0.951
13.2 bp
+79%
+55%
1.276
1.28 pts
1.550
1.56 pts
+21%
+21%
The changes in the drivers of volumes and turnover at the individual bond level provide
further evidence of the decline in liquidity. In Figure 7, we present regressions of turnover in
high yield bonds against size, age, and volatility in 2006 and 2014. In 2006, the two main
determinants of turnover were the age of a bond and its volatility. We interpret the
relevance of age as a halo effect around new issue bonds tend to trade in meaningful size
in the months immediately after issuance. Turnover increases with volatility because price
changes force investors to re-evaluate their holdings in a particular bond. Corporate
actions, earnings, upgrades and downgrades are all possible sources of volatility that could
lead credit investors to reposition their portfolios.
By 2014, a few things had changed. First, the coefficients on age and volatility were both
sharply lower. The new issue effect was much reduced, and it took much more volatility
to drive the same level of turnover. More interesting, size became a much more important
determinant of turnover. This suggests to us that investors pooled liquidity in the largest
bonds, which became proxy trading vehicles for the market. This is exactly the type of
reaction we would expect from investors struggling to position portfolios in a lowerliquidity environment the little liquidity that does exist is concentrated in a smaller number
of issues, rather than dispersed across the market.
24 February 2015
115
FIGURE 7
Cross-sectional regressions of annual turnover on size ($bn), age (yrs), and volatility (%)
2014 (R 2 18%)
Beta
2006 (R 2 20%)
Size
Age
Vol
Alpha
Size
Age
Vol
Alpha
0.15
-0.06
3.80
0.89
0.08
-0.17
11.40
1.20
Standard Error
0.03
0.01
0.11
0.02
0.04
0.01
0.34
0.04
t-statistic
6.10
-9.13
35.62
38.36
1.82
-14.82
33.16
32.30
$trn
Short-term Assets
24 February 2015
FIGURE 9
Non-financial corporates have also maintained a steady
proportion of assets in short-term funds
Non-financial corporate financial assets
ST Assets %
Total
20
50%
18
45%
40%
16
14
35%
12
30%
10
25%
8
20%
6
15%
4
10%
2
5%
0
0%
4Q89 4Q92 4Q95 4Q98 4Q01 4Q04 4Q07 4Q10 4Q13
$trn
Short-term Assets
116
2008, when sharp equity market declines reduced the value of stocks. In fact, absolute
holdings of short-term assets have increased in 19 of the past 20 years, by an average of
6.2%. In other words, the pace of growth in short-term assets has steadily tracked the longterm growth rate of household and corporate accounts. Indeed, Gorton et al report that
their safe asset share of all US assets has remained steady at around 33% since 1952.6
FIGURE 10
Following periods of strong equities performance, household exposure to equities peaks
and short-term assets reach a local low as a percentage of total assets
US Household Exposure to Equities and Short-term Assets
30%
24%
25%
22%
20%
20%
18%
15%
16%
10%
14%
5%
0%
Dec-89
12%
10%
Dec-94
Dec-99
Equity Exposure % Total Assets
Dec-04
Dec-09
Short-Term Assets % Total Assets (RHS)
Faced with the prospect of shrinking bank-provided repo, what alternatives are available to
investors seeking short-duration liquid assets? A survey of similar low-risk, short-term
options suggests investors may struggle to deploy their growing allocation to this category.
The main low-risk alternatives Treasury bills, bank deposits, and the Federal Reserves
new reverse repo program (RRP) all have their limitations.
24 February 2015
See The Safe-Asset Share, G. Gorton, S. Lewellen, A. Metrick, NBER working paper, January 2012.
117
FIGURE 11
Upon the contraction of the repo market in 2008, MMFs
redeployed capital into Treasury bills
40%
35%
% of Total
MMF
Holdings
$trn
FIGURE 12
However, bill volume has since declined and is unlikely to be
able to absorb incremental demand from declining repo
2.0
$trn
2.0
30%
1.5
25%
1.0
20%
15%
1.9
-22%
1.8
1.7
1.6
1.5
10%
0.5
1.4
1.3
5%
0%
0.0
2005
2006
2007
2008
2009
2010
2011
1.2
1.1
1.0
2008
2009
2010
2011
2012
2013
2014
Deposits are a natural alternative, but rates could lag if demand increases out of
step with lending opportunities
Deposits are clearly a safe, liquid asset and are one of the main areas where households and
corporates deploy short-term funds. Uninsured bank deposits (above the $250k insurance
maximum) do represent incremental credit risk versus government obligations; however,
money market funds already deploy roughly 20% of their holdings into wholesale bank
deposits, suggesting that deposits form a reasonable investment avenue for these entities.
However, we do not see much demand from banks for this incremental funding. Banks are
already awash in deposits, as demonstrated by an average loan-to-deposit ratio of roughly
70%. Although banks will continue to take deposits, away from pockets of demand for retail
deposits driven by the new Liquidity Coverage Rule (LCR), we believe banks interest in further
inflating their balance sheets is limited. Their appetite is constrained both by the new SLR rule
and limited lending opportunities. Thus, if money market funds boost the supply of deposits
to banks, banks in turn may be less inclined to raise the interest paid on deposits.
FIGURE 13
Deposits already form the core of households safe, short-term assets and a material share
of money market holdings, suggesting they are a likely alternative investment avenue
Mix of household financial assets (% of total)
Deposits
13%
Other Short-term
2%
Other Assets
85%
Source: Federal Reserve, Barclays Research
24 February 2015
118
The Feds reverse repo program (RRP) is a close substitute for shrinking private sector repo
and is available to a wider range of counterparties, including large money market funds and
the GSEs. For these investors, Treasury repo from the Fed supplements what is available to
them from the private sector. Since program testing began in September 2013, average daily
balances in the RRP have been roughly $125bn (and considerably higher at quarter-ends,
when bank and dealer balance sheet scarcity increases and few private sector repo assets are
available for money market funds to invest in). This has largely offset the decline in private
sector repo volume in recent years (Figure 14). In turn, mutual fund repo holdings have
remained relatively stable in aggregate (actually increasing as a % of total holdings) as they
have redeployed funds into the RRP.
FIGURE 14
Fed RRP has offset much of the fall in private-sector repo volume, helping MMFs keep total
repo holdings fairly constant
$trn
3.0
2.5
2.0
1.5
1.0
0.5
0.0
4Q11
2Q12
US Banks Repo
4Q12
2Q13
Fed RRP
4Q13
2Q14
However, the Feds stated concerns about the program have led it to put a hard cap of
$300bn on the RRP; thus, we expect its capacity to replace shrinking private sector
collateral supply to be limited. The Feds concern stems from its discomfort with directly
funding money market funds and the fact that even with the $300bn cap, its market
presence in the repo market is nearly as large as the top three dealers combined. Moreover,
it worries about the potential for the program to dis-intermediate bank funding during a
financial crisis. Notably, in the January FOMC minutes, most participants accepted that the
Fed might have to temporarily increase the cap on the overnight RRP program to
strengthen its control over the fed funds rate. Officials are concerned, however, that the
market might attach more significance and permanence to the RRP program if the size was
increased so it is far from certain the Fed will provide money funds with extra repo.
24 February 2015
In aggregate, we expect demand for safe, short-term assets to grow steadily. However, the
supply of these assets from the avenues listed above will likely be constrained. When we
factor in an expected decline in repo, we project an increased imbalance between supply
and demand. This imbalance more investors looking to deploy cash in the short end than
safe borrowers needing that cash should lower the relevant interest rates paid. For
example, the available data suggest that bank deposits have historically had a 60-80% beta
to short-term interest rate changes. We expect deposits to exhibit a lower beta once the
Federal Reserve begins hiking rates, as funds that short-term investors previously allocated
to repo assets flow into bank deposits. Through this indirect mechanism, forced declines
in repo volumes could keep the interest earned by deposits or government-focused money
market funds closer to the zero bound, even as other rates rise. In fact, it is exactly
the concern about substantial demand for short-dated assets that is leading the Fed to
119
question whether the RRP program may need to be increased. Otherwise, the actual frontend rates used in the economy may not track Fed funds as closely, limiting Feds control of
interest rates.
3%
20%
2%
10%
1%
0%
4Q90 4Q92 4Q94 4Q96 4Q98 4Q00 4Q02 4Q04 4Q06 4Q08 4Q10 4Q12
0%
-10%
-1%
-20%
-2%
-30%
-3%
Note: Short-term assets and deposit assets for all entities. Median short-term asset rate represents aggregate of
household and non-financial corporate data. Source: Federal Reserve Flow-of-Funds data, Barclays Research
24 February 2015
120
FIGURE 16
Holdings of cash-like substitutes grew dramatically pre-crisis and have shrunk since
Commercial paper and bankers acceptances outstanding ($trn)
$trn
2.5
2.0
1.5
1.0
0.5
0.0
4Q89 4Q91 4Q93 4Q95 4Q97 4Q99 4Q01 4Q03 4Q05 4Q07 4Q09 4Q11 4Q13
Source: Federal Reserve Flow of Funds, Barclays Research
FIGURE 18
ETFs have gained a significant foothold in the management
of fixed income assets
1,400
3.5%
1,200
3.0%
1,000
2.5%
800
2.0%
600
1.5%
400
1.0%
200
0
Jan-09
IG
0.5%
Mar-10
May-11
Balanced
Source: Lipper
24 February 2015
Jul-12
Flexible
Sep-13
Nov-14
HY
Other
0.0%
Jan-06
Jun-07
Nov-08 Apr-10
Sep-11
Feb-13
Jul-14
121
expect versus the (poor) liquidity available in the underlying bonds. ETFs function as a trading
vehicle, aided by their increasing liquidity, such that portfolio managers can meet daily inflows
and redemptions without actually needing to trade bonds7.
Investors are increasingly using
the ultra-liquid CDX indices to
satisfy their daily liquidity
needs
Similarly, portfolio managers have increased their trading in other related products. For
example, investors are increasingly using the ultra-liquid CDX indices to satisfy their daily
liquidity needs. In the high yield market, where the on-the-run CDX index trades nearly as
much as all TRACE bonds combined, the correlation between large fund flow events and
positioning data shows that portfolio managers use the derivatives index as a source of
liquidity in periods of high fund flow volatility (Figure 19).
However, these alternative sources of liquidity come at a cost, even if such cost is not
immediately apparent in bid-offer prices. With CDX, the price of liquidity comes in the form of
basis risk, which can be very significant in times of market stress (Figure 20). This risk comes as
a result of mismatches in rates exposure (CDX has virtually none) and differing credit exposure,
among other potential mismatches. With the ETFs, the costs include non-trivial management
fees and a market that can dislocate significantly from its underlying asset value. Holding more
cash to fund potential liquidity events is an alternative whose risks are better understood, but
the consequent performance drag can make this the least appealing option to managers.
The increased use of these tools to manage the disparity between the provision of daily
liquidity to end-investors and poor liquidity in the underlying fixed income assets is itself
evidence of the tension that the influx into mutual funds has caused. Fund managers have
found that they need to use these tools already, in relatively calm markets. In the event of a
market disruption, these tools may no longer be effective if outflows exceed the extent to
which fund managers have built in flexibility to meet them, they would have no choice but to
turn to the underlying markets to meet their liquidity needs. This could become selfperpetuating if the corresponding price declines in the underlying led to further outflows.
Thus, regulations aimed at bolstering stability at the core of the financial system, combined
with a growing demand for liquidity, may eventually lead to increased instability and fire-sale
risk in the periphery (eg, the secondary markets for investment grade, high yield, leveraged
loans, and emerging markets). The fragile new equilibrium comes not only from the reduced
tradability of these asset classes, but also from deep liquidity mismatches between the assets
themselves and the instruments being used to manage daily liquidity needs.
FIGURE 19
Changes in investor positioning in HYCDX (OTR) are
consistent with liquidity needs ($mn)
200
1,000
750
500
-100
-200
-250
-750
-1,000
100
0
250
-500
FIGURE 20
Basis between the HYCDX index and the Barclays US High
Yield Very Liquid Index (bp)
-300
-400
-500
-600
Mar-07
Apr-09
May-11
Jun-13
24 February 2015
-700
Feb-05
Institutional Investors Turning to Fixed-Income ETFs in Evolving Bond Market, Greenwich Associates, 2014.
122
CHAPTER 6
We expect Indias real GDP to grow at 7-8% annually in the next 5-10 years very
strong for an economy exceeding USD 2trn and with about a 3% share of global GDP.
Against a backdrop of generally subdued global growth, our forecasts imply that India
could be the worlds fastest-growing large economy in the years ahead.
Indias central bank is also enjoying a fresh credibility boost under Governor Rajan.
Amid other tailwinds, Indias twin deficits are improving quickly and inflation is
softening materially. India is one of the biggest beneficiaries of lower commodity
prices, oil in particular, which we believe can remain low over the medium term.
We expect India to enjoy multi-notch rating upgrades to high BBB by 2017 and we
remain positive on INR assets (ie, bonds, equities) over a multi-year horizon.
mitul.kotecha@barclays.com
Dennis Tan
The recent turnaround in Indias economy has been remarkable. In the wake of concerns
about a potential balance-of-payments (BoP) crisis in mid-2013, there has been a sharp
turnaround in sentiment and asset prices, reflecting a marked improvement in Indias
macroeconomic fundamentals. After years of sub-par economic performance, the Indian
economy looks set to gain further strength in the coming years, buoyed by multiple cyclical
and structural tailwinds.
First, the mid-2014 general election ushered in a reformist government with the strongest
mandate in nearly 30 years. The new government has moved quickly to pursue an
ambitious programme of macroeconomic and policy reforms.
Second, the governments policies, apart from boosting growth and development, seem
poised to maintain an explicit focus on generating jobs. India enjoys a strong demographic
dividend, with a high percentage of the population of working age. Thus, an explicit focus
on job creation should help India foster development and materially combat poverty.
Third, the central bank, the Reserve Bank of India (RBI), under Governor Raghuram Rajan, has
further strengthened its inflation-fighting commitment recently, which could help India
overcome a perennial stumbling block.
Fourth, in the near to medium term, India looks set to be among the biggest beneficiaries of
softer commodity prices, which could translate into several macroeconomic pluses: falling
inflation, stronger current account balance, potentially even a surplus in some years, a smaller
fiscal deficit, lower interest rates, and faster growth.
123
electoral mandates, which had weighed on governments ability to make decisions and initiate
reforms. This issue was in sharp focus with the previous government, which, facing pressure
from major corruption allegations and challenging alliance partners, put economic reforms on
the back-burner for most of its time in office. We think the 2014 election result has
significantly improved Indias ability to take on difficult but necessary policy reforms.
Narendra Modi also emerges as Indias first prime minister to head a government that is
right of centre in terms of economic policies. Identifying economic growth as the fastest
path to reducing absolute poverty, the government has moved quickly to restore business
and consumer confidence. Several key economic sectors were liberalised for foreign direct
investment (FDI) in the early days of the governments tenure. The government is pursuing
an ambitious agenda of key fiscal reforms, financial inclusion, boosting the manufacturing
sector, and facilitating planned urbanisation. The government has curtailed subsidies on a
medium-term basis by deregulating domestic diesel prices; undertaken fiscal reforms; and
pursued initiatives to ease labour laws and land acquisition norms to facilitate
manufacturing and de-clogging infrastructure bottlenecks. Overall, it has committed to
improving the ease of doing business. India currently languishes at 142nd place in the
World Banks Ease of Doing Business rankings, but Prime Minister Modi has set his sights
on reaching the top 50 in the next 3-5 years.
At the same time, more state governments have aligned themselves with the centre, either
through political affinity or through federal programmes, which should help in terms of
project implementation. The ruling BJP now directly controls eight of Indias 29 states, and
has formed alliances with regional parties in another three. These states are systemically
important and together contribute nearly half of Indias GDP. With eight state elections
scheduled by 2016, the BJPs political control could strengthen. In sum, the political
backdrop, which has long hobbled Indias growth story, is now a clear tailwind, in our view.
Political inertia seems to have disappeared, and further reforms, targeting growth,
employment and improving the ease of doing business will likely remain the focus of the
government in the coming years. Key fiscal reforms such as the implementation of an
integrated goods and service tax (GST) has been delayed for years, largely because of
differences between state governments and the centre over revenue sharing, but we expect
this initiative to move forward over the next 1-2 years. Furthermore, alongside the opening
up of more sectors to larger FDI caps in the coming years, we also expect a number of
critical business hurdles to be lowered, especially with regard to land acquisition, labour law
reforms and the ease of doing business.
FIGURE 1
Narendra Modis BJP won the 2014 election with a landslide
350
300
282
FIGURE 2
giving India its first single-party majority govt since 1984
100
90
80
250
70
200
60
150
50
100
54
50
44
37
15
46
34
20
11
Others
24 February 2015
Others
TRS
BJD
TMC
AIADMK
UPA allies
NDA allies
BJP
INC
UPA
30
20
NDA
40
10
0
84
89
91
Congress
96
98
BJP
99
Left parties
04
09
14
Others
124
FIGURE 3
India is set to remain a young country for decades to come
60
FIGURE 4
fuelling a surge in its labour force
1,600
Median age
(years)
1,400
50
1,200
40
1,000
800
30
600
20
400
10
200
0
1991 1997 2003 2009 2015 2021 2027 2033 2039
0
IN
BR
CN
2000
ID
2015
KR
US
Residual
2040
FIGURE 5
Labour productivity has fallen in recent years, but is still far
above the long-term average
9
10
3
2
0
1982
1986
1990
24 February 2015
2
2014
FIGURE 6
Disposable income has been trending upward since the late
1990s
25
20
15
10
5
0
FY98
FY03
FY08
FY13
Per capita personal disposable income (% y/y)
125
Growth: Stepping up
Despite a slowdown in recent years, Indias growth in the past 10 years has averaged above
7.5% pa, with the size of the economy expanding from ~USD700bn to more than USD2trn.
India continues to benefit from a large, young population, a deep savings pool to finance
capital investment, and productivity growth. Although infrastructure bottlenecks persist, the
government is taking steps to resolve them, setting the stage, we think, for better growth
dynamics. We estimate that India can grow at an average of 7-8% pa over the next decade.
We feel that the political transition in 2014 has been the main catalyst for the revival of
expectations for the Indian economy. The government looks poised to capitalise on the recent
surge in confidence and optimism in the economy and seems committed to combating some
of the persistent challenges it faces, including generally entrenched inflation expectations, a
sub-optimal fiscal situation, a stagnating agricultural sector, corruption and poor delivery of
government services. We think the government is currently prioritising three areas: 1)
manufacturing, via the Make in India campaign; 2) planned urbanisation; and 3) enhancing
the effectiveness and efficiency of government services.
Planned urbanisation
Another key focus area is urbanisation, which has been a strong driver of growth in India
since the 1990s. Urban areas now generate nearly two-thirds of Indias GDP up from
c.45% in 1990. The policy focus is to make the countrys urbanisation plans more coherent
24 February 2015
126
with its industrialisation needs by creating new smart cities and dedicated industrial
corridors, as well as upgrading existing urban infrastructure. If the government succeeds,
we think this would be a major factor supporting industrial growth. Even if past trends are
maintained, we estimate that urban India could make up 35% of the countrys population
and contribute 70-75% of GDP by 2020. According to Urbanisation in India, a report by the
McKinsey Global Institute in 2010, India could have 68 cities with a population of more than
1m by 2030 up from 42 in 2010.
FIGURE 7
Improvements in productivity and capital formation could help India average 7-8% growth over the next 5-10 years*
Investment rate (GFCF as a % of GDP)
ICOR
27%
30%
33%
36%
39%
42%
3.50
7.7%
8.6%
9.4%
10.3%
11.1%
12.0%
3.80
7.1%
7.9%
8.7%
9.5%
10.3%
11.1%
4.10
6.6%
7.3%
8.0%
8.8%
9.5%
10.2%
4.40
6.1%
6.8%
7.5%
8.2%
8.9%
9.5%
4.70
5.7%
6.4%
7.0%
7.7%
8.3%
8.9%
5.00
5.4%
6.0%
6.6%
7.2%
7.8%
8.4%
Note: *Values indicate real GDP growth rate for a certain combination of incremental capital output ratio (ICOR) and investment rate. Source: Barclays Research
24 February 2015
127
FIGURE 8
Indias saving improvement is in line with Chinas experience post-liberalisation
Change in savings
rate from the year
of takeoff (%)
20
15
10
5
0
-5
0
10
15
20
Year of takeoff
CN
IN
25
30
35
ID
Take-off years for China, India and Indonesia are defined as 1979, 1991 and 1973, respectively.
Source: Haver Analytics, Barclays Research
A number of Indias macro parameters have a degree of resemblance with those of China 12 decades back, when China had taken off for a spell of very high growth. For example,
Indias per capita GDP crossed the USD1,000 mark in 2007, a milestone that China crossed
in 2001. On a purchasing power parity basis (PPP), Indias per capita GDP in 2014 stood at
around USD5,700; a similar level was achieved by China in 2006 (China in 2014: about
USD13,000). The current differential in per capita income between the two countries
remains large given Chinas protracted period of high growth and strong currency.
However, a steady near-8% growth rate in India in the coming years if maintained could
help India to narrow this gap in the coming years.
The manufacturing sector and an uptick in exports played a critical role in boosting Chinas
economy during the high growth years. Recently, this particular area has been a relatively
weak point for India. Nevertheless, Indias current export intensity remains similar to that of
China at the turn of the century. Indian policymakers are currently emphasising boosting
the countrys manufacturing industry with a view to eventually boosting net exports, which
could be a critical part of Indias growth strategy in the coming 5-10 years. A somewhat
similar pattern can be observed in case of a few other important macro parameters, such as
the rate of gross capital formation.
FIGURE 9
Indias current per capital GDP is close to that of China about
a decade back
FIGURE 10
while Indias current export intensity is broadly in line
with that of China at the turn of the century
China
45
3,000
35
2,000
1,000
24 February 2015
2010/2019
2009/2018
2008/2017
2007/2016
2006/2015
2005/2014
2004/2013
2003/2012
2002/2011
2001/2010
2000/2009
1999/2008
1998/2007
1997/2006
1996/2005
1995/2004
India
40
China
India
4,000
Exports % GDP
5,000
30
25
20
15
10
5
0
-
1,000
2,000
3,000
4,000
128
25
15
5
-5
-15
Jan-05
Apr-06
Jul-07
Oct-08
Jan-10
Apr-11
Jul-12
Oct-13
Jan-15
FIGURE 12
Current account balance eyeing a surplus
20
FIGURE 13
Current account ex-gold gives a much stronger picture
2.5
40
20
0
0.5
-20
0
-20
-40
-1.5
-60
-3.5
-80
-40
-60
-80
-100
-5.5
FY91
24 February 2015
-100
FY04
FY06
FY08
FY10
FY12
FY14
FY16F
Gold imports
129
FIGURE 14
Foreign reserves back on a rising track
350
20
10
300
250
-10
200
-20
150
-30
100
-40
Jan-05 Jan-06 Jan-07 Jan-08 Jan-09 Jan-10 Jan-11 Jan-12 Jan-13 Jan-14 Jan-15
Foreign reserves (USDbn)
FIGURE 15
Oil bill accounts for almost two thirds of total trade deficit
0
-50
40
-3
50
-4
60
70
80
-6
90
-7
-150
-250
Dec-09
-2
-5
-100
-200
FIGURE 16
Fall in crude oil price a major tailwind for Indias BoP
100
-8
Oil
Gold
Trade balance ex oil ex gold
Dec-10
Dec-11
24 February 2015
Dec-12
12m rolling
sum, USD
bn
Dec-13
Dec-14
110
-9
-10
2008
120
130
2010
2012
2014
130
80
1,800
75
1,600
70
1,400
65
1,200
60
1,000
55
800
50
600
400
45
40
Jan-12
200
Jul-12
Jan-13
Jul-13
Diesel (INR/ltr)
Source: Bloomberg, Barclays Research
Jan-14
Jul-14
Petrol (INR/ltr)
Jan-15
0
FY05
FY07
FY09
FY11
FY13
FY15E
FY17E
Liberalising FDI rules remains a key government focus. We expect policy initiatives to keep
capital inflows into India strong, resulting in a persistent balance of payments (BoP) surplus.
We think Indias economy can sustain FDI levels of USD40-50bn (about 2.0% of GDP),
which would play a key role in financing the current account gap. We believe the proposed
opening of strategic sectors, including defence, insurance, railways and aviation, to foreign
investment signals the governments willingness to attract more foreign investment to
improve Indias industrial capacity and benefit consumers. We expect other policy initiatives
to boost foreign inflows, including raising the limits on foreign investment in the domestic
debt market, and the potential issuance of (quasi-) government bonds in foreign currencies
and/or offshore INR bonds.
24 February 2015
131
FIGURE 17
Inflation continues to surprise to the downside
FIGURE 18
on the back of a broad-based softening
16
% y/y
14
12
12
10
10
8
0
-2
1999
2001
2004
2006
2009
2011
2014
WPI (% y/y)
CPI-combined (% y/y)
Repo rate (%)
Source: GoI, RBI, CEIC, Barclays Research
0
-2
Jan-09
Jan-10
Jan-11
Jan-12
Core CPI
Jan-13
Jan-14
Jan-15
Core WPI
We expect the overall BoP dynamics to lead to a healthy accumulation of foreign exchange
reserves at the Reserve Bank of India (RBI). Moreover, following the 2013 experience of
uncertain BoP dynamics and rapid depreciation of the INR, the new leadership of the RBI
seems determined to build foreign reserves to defend the currency, if need be. We think the
RBIs forex reserves will likely reach USD450bn by FY 2018-19, even on what we consider to
be a conservative assessment. Indias import cover has averaged close to 11 months over
the past two decades and around nine months over the past five years, above the IMFs
recommended safe level of six months.
UPA-I (CPI
food average:
7.3%)
UPA-II (CPI
food average:
11.5%)
20
FIGURE 20
leading to a major drop in inflation expectations recently
15
(%, y/y)
12
9
15
10
5
0
Sep-06
-5
1998 2000 2002 2004 2006 2008 2010 2012 2014
CPI: Food inflation (%, y/y)
Source: GoI, CEIC, Barclays Research
24 February 2015
HH inflation expectation
Jun-08
Mar-10
Dec-11
Sep-13
132
FIGURE 21
India has delivered significant deregulation of fuel prices since 2010
June 2010 Government benchmarks
Petrol prices to market level
Petrol
Diesel
LPG
Kerosene
Note: The horizontal bars in the chart above denote the indicative time period during which (since 2010) fuel price reforms took place. Source: PIB, Barclays Research
tends to be more volatile, but the new government has enjoyed some early success in tackling
rising food prices. The ongoing weakness in commodity prices, which seems unlikely to be
temporary, is also playing an important role in this context. The recent softening in inflation
momentum thus appears quite broad-based. The central bank has also recently flagged
downside risks to its early 2016 CPI inflation forecast of 6%. The material softening in inflation
momentum triggered the start of rate cuts by the RBI in early 2015. We expect the RBI to cut
the repo rate by a cumulative 75bp during H1 2015.
133
FIGURE 22
Indias fiscal deficit on a gradual consolidation path
FIGURE 23
Primary deficit set to be close to zero by FY17
-1
-2
-2
-3
-4
-4
-2.5
-3.0
-3.3
-5
-4.8
-6
-6.0
-7
FY07
FY09
-4.9
-4.5
-4.1
-6
-3.6
-8
-10
FY77
-5.8
-6.5
FY11
FY13
FY15F
FY17F
FY82
FY87
FY92
FY97
FY02
FY07
FY12 FY17F
The recent fall in global commodity prices, particularly oil prices, and the governments
initiatives to gradually rationalise domestic energy prices, will likely be a material benefit for
the fiscal balance over the coming years. A materially lower subsidy bill is likely not only to
help reduce the fiscal deficit, but also to improve the quality of government expenditure by
shifting it toward capital spending.
On the revenue side, the central governments level of tax collection has been stagnant for the
past eight years. Tax cuts introduced to stimulate growth in 2009-11 had a long-lasting
impact on the fiscal balance as the governments reliance on non-tax revenue rose, while
fixed expenditure (eg, subsidies, interest payments, defence) kept rising along with spending
under various welfare schemes between 2008 and 2012. The government is also looking to
boost revenues. A key initiative in this regard is the implementation of an integrated GST,
potentially from 2016. Apart from broadening the tax base, we think a GST would trigger
significant productivity gains, given the move to a single tax platform.
134
We believe the undervaluation of the INR is likely associated with two factors. First, the INR
was previously under pressure from the widening current account deficit, which in turn was
associated with the failure of domestic production capacity to keep up with the rapid
growth in domestic demand. Factors such as higher CPI and wage inflation in India relative
to trading partners, infrastructure bottlenecks and rigid land and labour laws limited the
improvement in competitiveness of the manufacturing and export sectors. Consequently,
import growth outpaced export growth during the boom years. Second, the RBI has tended
to limit INR appreciation during times of strong capital inflows. According to our estimates,
FX reserve accumulation in Indias amounted to around 4% of GDP during 2001-14.
Although this is a comparatively slower pace than in other countries in the region, such as
Singapore (8.3%) and China (5.8%), over the same period, it was higher than in Malaysia
(3.2%) and Indonesia (0.5%).
However, macroeconomic fundamentals in India are improving and should help to lessen
the pressure on the INR. Importantly for the rupee, Indias current account deficit should
narrow significantly on lower oil prices and an improvement in savings/investments ratio.
Lower inflation, as a result of cheaper fuel, should also enable the RBI to cut interest rates
further to boost growth and investment. Structurally lower consumer price and wage
inflation, along with reforms to regulations on infrastructure development, land acquisition
and labour, should also help strengthen the competitiveness of the manufacturing and
export sectors. Above all, India is embarking on a higher growth trajectory, which in turn
should raise the attraction of and demand for the rupee among foreign investors.
Even if the current account balance eventually weakens on stronger domestic demand, we do
not expect India to face difficulties funding the deficit, given the likelihood that the country will
continue to attract both short- and long-term inflows as Prime Minister Modis government
implements its reform agenda. The government also may gradually lift the limits on foreign
investors investments in local debt, and FDI regulations are being relaxed gradually. Indeed,
the prospects of long-term structural investment inflows should reduce the reliance on shortterm, hot money flows, leaving the currency less susceptible to capital flow gyrations.
Despite improving fundamentals and strong capital inflows, USD buying by the RBI to
rebuild FX reserves allowed limited INR appreciation in the recent past. However, we think
the healthy increase in FX reserves has strengthened RBIs capacity to defend the INR in
times of market stress and currency weakness. Although we expect a likely continued
increase in foreign exchange reserves over coming years to limit INR appreciation as the RBI
absorbs dollars, this will in turn provide a higher degree of confidence and stability in the
rupee over the long term.
21%
24%
21%
17%
NZD
CNY
16%
9%
ILS
16%
8%
AUD
TRY
7%
THB
HUF
6%
PLN
15%
6%
CZK
14%
5%
BRL
RON
3%
GBP
12%
2%
IDR
20%
USD
SGD
40%
CHF
FIGURE 24
The INR is among the most undervalued currencies
-4%
-3%
-2%
-2%
-2%
0%
TWD
MXN
MYR
HKD
KRW
-6%
CLP
-5%
-8%
COP
EUR
-10%
DKK
-12%
INR
-16%
CAD
-23%
-27%
-32%
-40%
-36%
-20%
-19%
0%
PHP
SEK
RUB
ZAR
NOK
JPY
ARS
-60%
24 February 2015
135
Overall, we expect the INR to come under less pressure than in previous years, even though
the currency could still depreciate against the USD in a strong-USD environment. Since the
1990s, much of the appreciation of the INR effective exchange rate has occurred in real
terms due to relatively faster inflation in India; the INR NEER has been on a downward path.
However, we see scope for the currency to appreciate in NEER terms, given the
improvement in macro fundamentals, with NEER appreciation helping to play a bigger role
in correcting some of its undervaluation over the medium term. We also think the INR will
likely remain an attractive proposition to yield-seeking investors in a world of depressed
interest rates.
Additionally, Indias currency framework will undergo further structural reforms over the
coming years, and this should help to strengthen confidence in the rupee. India is set to
continue to gradually move towards capital account convertibility after making the current
account convertible in 1994. Currently, a limited amount of capital account transactions are
permitted. The RBI has taken more steps in recent years to relax regulations around foreign
exchange transactions. However, we expect foreign investors to be allowed increased
access to Indias capital markets, especially Indias bond markets where there is a cap on
such inflows. Importantly, India is also likely to allow its huge pool of domestic savings to
increasingly access foreign markets. However, caution over possible disruptions to the
domestic financial sector means that this process will likely continue to be gradual.
FIGURE 26
Indias external balance is improving, with capital inflows
picking up while the current account deficit is narrowing
4-qtr sum, % fo GDP
120
14%
REER
100
NEER
Current account
Net portfolio investm't
Basic Balance
9%
80
4%
60
40
-1%
20
0
70
75
80
85
90
95
00
05
10
15
Source: Bruegel, Barclays Research (Click here to view the INR REER in Barclays
Live)
24 February 2015
-6%
Q1-03
Q1-05
Q1-07
Q1-09
Q1-11
Q1-13
136
FIGURE 27
Household financial savings have declined since 2010
17%
15%
FIGURE 28
however, with a sharp rise in real rates, we expect the
2010-13 trend in household financial savings to reverse
2
13%
11%
-2
-4
9%
-6
7%
5%
2001
-8
02
2003
2005
2007
2009
2011
04
06
08
10
12
14
16
18
2013
Source: CEIC, Barclays Research
Sustained positive real rates, resulting in a rise in household financial savings to 11% of
GDP by FY18 from 7.1% of GDP in FY13. This should boost demand for bonds from
insurance companies and pension funds.
However, given falling inflation, the governments renewed focus on small savings and the
RBIs anti-inflation stance, we think positive real rates are here to stay and that the trend in
household financial savings is set to reverse (see Chapter 1, Population Dynamics and
the (soon-to-be-disappearing) global savings glut). Moreover, given poor returns on
valuables and real estate, the backdrop looks favourable for a rotation from physical
savings to financial savings. Finally, a broader choice of financial savings products and a
savings boost from wider pension access and revitalization of small savings programs
should help speed the process (see Asia Macro Themes: India: A step change for details.)
In our view, pension funds, insurance and equity/debt instruments are set to attract
substantially larger flows given increased product penetration and focus. We expect this
flow increase to outpace the issuance of bonds; hence, pension and insurance demand
for equities and corporate debt should increase. Although deposits are likely to attract a
smaller portion of the expected rapid expansion of the household savings pool, we
expect them to still grow in line with nominal GDP. Banks can fund strong credit growth
through bond issuance and by reducing their government bond holdings.
24 February 2015
137
FIGURE 29
Higher savings can kick-start virtuous cycle for asset markets
Higher
HH Fin
Savings
Indian household
financial savings
have fallen a lot...
-3.7%
7.2%
From
of GDP in FY14
led by small
savings, which
have dropped a lot
As fiscal deficit
falls, govt bond
demand likely to
outstrip supply
RBIs holding of
a rise to
in FY18 is possible
surge to
of HH fin savings
by FY18
of
outstanding
bonds can shrink
with ample
household savings
to buy equity/debt
issuance
11%
75%
LDR is
Expect SLR cuts.
Banks also likely
to sell bonds
10%
Retail demand to
rise
4x by FY18
12.5%
Lower bond
yields
Strong
domestic bid
for equities
24 February 2015
138
FIGURE 30
Gold is less attractive as an investment/inflation hedge
INR
FIGURE 31
Mix likely to change with more small savings
Currency
20%
35,000
50%
15%
40%
30,000
20%
20,000
10%
10%
Provident &
Pension
Funds
30%
25,000
2000-05 Avg
FY14
FY18
Share &
Debentures
5%
0%
0%
15,000
-10%
Life
Insurance
Funds
Jun-14
Jun-13
Dec-13
Jun-12
Dec-12
Jun-11
Dec-11
Dec-10
Jun-10
Dec-09
Jun-09
Jun-08
-20%
Dec-08
10,000
Claims on
Govt
FIGURE 32
Demand/supply divergence
FIGURE 33
Bond yields set to decline gradually against a backdrop of
sustained disinflation
16
14
27%
12
22%
10
Demand
17%
6
12%
Supply
7%
2012
2
2013
2014
2015
2016
2017
2018
0
03 04 05 06 07 08 09 10 11 12 13 14 15 16
FIGURE 34
India has one of the highest EM risk premia
FIGURE 35
Significant current account adjustment would accelerate a
sharp decline in FX risk premia
Note: For India, we use 5y SBI CDS. Source: Barclays Research, Bloomberg
24 February 2015
Russia
Turkey
Mexico
Indonesia
France
Italy
SA
Germany
Malaysia
Thailand
US
Korea
India
China
Poland
Brazil
6.00
Hungary
4
3
2
1
0
-1
-2
-3
-4
-5
-6
-6.00
5.00
-5.00
4.00
-4.00
3.00
-3.00
2.00
-2.00
1.00
-1.00
0.00
0.00
-1.00
1.00
05
06
07
08
09
10
11
12
13
14
139
Equities: Indian stock market could have multiple years of growth in the mid to
high teens80
Strength in Indias GDP growth should also be reflected in a higher growth trajectory for
the corporate sector. Even if the sectors top line moves only in line with nominal GDP
growth rates, we would expect top-line growth in the low teens for Indian corporates.
Looking at the bottom-up estimates of our research analysts, we think sectors including
financials, healthcare, consumer, autos and infrastructure could exhibit strong doubledigit growth over the next decade. New sectors could also emerge, with e-commerce
and alternative energy (largely solar) two areas that we believe could return 20%+
annual growth over the next 10 years.
On various metrics, including market cap to GDP and market-implied growth rates, we
find Indian valuations reasonable. Furthermore, during periods of strong growth, market
multiples usually increase. We thus believe that the Indian market could have multiple
years of returns in the mid to high teens.
FIGURE 36
Sectoral growth trends in India over the past decade
%
10Yr
25
20
15
20.9
14.9
15.1
17.1
14.4
17.1
13.5
15.6
13.6
13.2
10
12.8
6.0
Total
Utilities
Telecom
Materials
IT
Industrials
Health Care
Financials
Energy
Consumer
Staples
Consumer
Discretionary
Nominal GDP
Note: Listed companies sector growth rates are based on an analysis of 640 standalone companies financials.
Source: Planning Commission, Reserve Bank of India, Prowess, Barclays Research
FIGURE 37
Barclays analysts expectations of growth in sectoral market size over the next decade
Consumer
- Organized Retail (Penetration)
Healthcare
- Diagnostics
Now
2025E
CAGR
US$37bn
US$160-220bn
14-18%
6%
20%
US$100bn
US$350-380bn
13-15%
Rs230bn
Rs834bn
27%
- Domestic Pharma
US$15bn
US$55-65bn
14-16%
- Health Insurance
Rs192bn
Rs464bn
16%
E-commerce
US$13bn
US$150bn
28%
Internet Penetration
21%
46%
Internet
Financial Services
- Credit growth
16-17%
80
This section summarises the views of our equity analysts published in Asia Themes: India in the next decade, 19
January 2015.
24 February 2015
140
- Credit penetration
Now
2025E
CAGR
55%
71-76%
150-180bp/year
7.50%
12-15%
18-22%
43%
51%
15%
-Cement
220mt
465mt
7.0%
- Steel
82mt
175mt
6.5%
7,000BPkm
20,000BPkm
11.0%
1,400BTKm
3,400BTKm
8.3%
- Solar Infrastructure
2,800MW
20,000MW
Materials
Infrastructure
Autos
- Cars
-
22%
12-14%
2.5mn
11.3mn
15%
14mn
40mn
10%
- Oil Demand
3.7mbd
5.4mbd
3-4%
51bcm
107bcm
7-8%
Two-wheelers
Note: mbd is million barrels per day; bcm: billion cubic metres BPkm: Billion person kilometres, mt: million tones.
Source: Reserve Bank of India, Government of India, BP Statistical Review, Barclays Research estimates
FIGURE 38
Indias market capitalization as % of GDP still below
historical highs
FIGURE 39
Across major nations, stock market capitalization as % GDP
tends to rise in periods of high growth
% GDP
Dec-14
Oct-13
Jun-11
Market Capitalization (%
GDP)
Aug-12
Apr-10
Feb-09
Oct-06
Dec-07
Aug-05
Jun-04
Feb-02
Apr-03
Dec-00
Oct-99
Jun-97
Aug-98
20
High
25
x avg GDP
growth
4.0
30
3.0
2.0
15
1.0
10
-1.0
-2.0
Low
160
140
120
100
80
60
40
20
0
-3.0
-4.0
Low
0.5
1.0
1.5
x avg Market Cap/GDP
High
2.0
Note: We used 10 major stock exchanges and countries to construct the above
chart. The market capitalization as a % GDP and GDP growth rate of each country
for the past four years are represented as x times avg over the past decade. Low
market cap/GDP thus implies a value lower than that countrys last 10 year avg.
Source: World Bank, United Nations, Bloomberg, Barclays Research.
141
Fiscal consolidation driving the central government deficit toward 3% of GDP. Improved
quality of spending through subsidy reforms and progress on implementing a GST to
strengthen long-term fiscal health.
Banking sector reform is set to increase the capitalisation levels of public-sector banks,
in line with Basel III guidelines, improve corporate governance and reduce the level of
non-performing assets.
Rating
positive
trends needed
between now
and 2017
Sensitivity of
our rating
view to
downside
risks
Fiscal and
government
debt
dynamics
Growth
dynamics
Institutional
effectiveness/
governance
Monetary policy
effectiveness/
credibility
Banking
system
Central
government
fiscal deficit
moves
towards 3%
by FY17
Growth
continues to
accelerate and
reaches near
7% by FY17
Capitalisation
levels increase
across the
system with a
buffer to Basel
III ratios
Debt/GDP
decreases by
over 4% in
FY17 vs. FY14
levels
Growth is not
funded by a
surge in credit
Improvement in
governance
indicators and
competitiveness
indicators to
2003 levels.
General
agreement on
policy direction
with
predictability
Higher degree of
perceived
independence for
the central bank
Stressed
assets
decrease to
2011 levels
Slippage in
deficit by 2030bp or in
debt/GDP by
2% can be
absorbed
Low 6%
growth
unlikely to
change
trajectory
Critical to
upgrades and
any slippage
can constrain
ratings
Significant
slippage could
have an effect
on ratings
24 February 2015
142
CHAPTER 7
Hamish Pepper
volatility has risen recently. We think a trend rise in volatility may be forthcoming in
a highly asynchronous global economic recovery with elevated macroeconomic
uncertainty related to demographic and structural changes across major economies.
An increase in foreign exchange market volatility has the potential to erode returns
nikolaos.sgouropoulos@
barclays.com
We think FX volatility is likely to rise in trend from the low levels of recent years. In this
chapter, we examine how high FX volatility affects an international portfolio and whether or
not FX hedging helps to improve the risk-adjusted portfolio performance of a balanced
international portfolio. We also examine the impact on an international portfolio, hedged
and unhedged, of a trend rise in the USD, which we expect in coming years. Specifically, we
FIGURE 1
Financial market volatility has picked up recently
FIGURE 2
... particularly in FX markets
150
400
Financial
crisis
350
Euro area
crisis
300
140
130
120
250
110
200
100
90
150
80
100
70
50
60
0
2004
2006
2008
Equity IV
2010
FX IV
2012
2014
Swaption IV
24 February 2015
50
Jan-14 Mar-14 May-14
Equity IV
Jul-14
FX IV
143
Our results show that FX hedging the bond portfolio raises the negative correlation between
equity and bond returns, driving the volatility of the overall portfolio lower and improving riskadjusted returns. Our analysis suggests that this feature is particularly pronounced during
periods of heightened market volatility (eg, the 2008-09 global financial crisis and the euro
area debt crisis of 2011-12). Our results also hold more generally and suggest significant
benefits from hedging the FX exposure of the bond portfolio even in normal periods.
Return to volatility
Between mid-2012 and mid-2014, volatility across asset classes declined to historically low
levels (Figure 1). Extremely accommodative global monetary policy, increased financial
regulation, a decline in macroeconomic volatility, greater synchronicity of global economic
cycles and, perhaps, auto-correlation of volatility, all appear to have dampened volatility in
recent years (see Three Questions: Gone fishin, 4 August 2014, for a detailed discussion of
some of the factors behind low realized volatility during this period).
China is likely to remain a
source of uncertainty as it
attempts to rebalance its
economy toward consumption
while managing a structural
slowdown
We expect the highly unsynchronised global economic recovery, broad demographic trends
(see Chapter 4, The great destruction) and structural economic change in major economies,
including China and India, to lead to a sustained rise in macroeconomic uncertainty, creating an
environment of higher financial market volatility. We forecast the euro area economy to grow at
less than half the pace of the US over the next two years. Strong growth and an improving
labour market should support a multi-year process of Fed policy normalisation, which is likely
to begin in June, in our view. In contrast, we expect several small, open-economy central banks
to introduce further stimulus this year in response to weak inflation outlooks and unwanted
exchange rate appreciation. In some cases, this could include the introduction, or expansion, of
extraordinary measures, such as negative deposit rates and quantitative easing (see Three
Questions: Quantum Evolution, 27 January 2015). Elsewhere, China is likely to remain a source
of uncertainty as it attempts to engineer a rebalancing of its economy away from investment
and exports toward consumption while managing a structural slowdown stemming from an
ageing population and declining labour force. Moreover, significant structural change in India is
also likely to take place over the coming years under the Modi government (see Chapter 6,
India: A step change). Additionally, continued political uncertainty in Europe is a reminder of
the unsettled risks around European Monetary Union.
FIGURE 3
Our hedged portfolio results are consistent with existence of positive volatility risk premium
40
7
6
35
30
25
20
15
1
0
10
-1
5
0
2005
-2
-3
2007
2009
2011
2013
2015
24 February 2015
144
One of the most interesting features of the recent pick-up in financial market volatility is that it
has been most pronounced in FX markets (Figure 2). One possible explanation is that
currencies are the most liquid and accessible assets through which to express a view on
market risks or to hedge exposure. Indeed, history suggests that higher FX volatility rarely
occurs in isolation and the prospect of higher volatility in other asset classes adds further risk
to multi-asset portfolio returns. Furthermore, there is a strong theoretical basis for a
relationship between volatility across asset classes. In equities, volatility in FX affects the
earnings of companies with international exposure in either their product or supply chains.
However, causality can also be argued in the opposite direction as changes in the price of
domestic assets, both equities and bonds, will tend to result in changes in demand for local
currency by foreign investors. Indeed, statistical analysis over our sample period proves
inconclusive in this respect, showing two-way causality between FX volatility and that of
bonds and equities.
To examine the impact of rising volatility on portfolio returns, we focus on the response of
risk-adjusted returns. We choose to think about risk and reward via the commonly used
Sharpe ratio. The Sharpe ratio, also known as the reward-to-variability ratio, is a
mathematical construct formally defined by William Sharpe in 1966. It measures a
portfolios predicted performance as the ratio of its expected rate of return per unit of
variability or risk.1 Although it is an imperfect measure of risk-adjusted performance when
returns are not normally distributed, we choose to use (a modified version of) the Sharpe
ratio because of its tractability and ease in ex ante portfolio choice given a particular risk
tolerance and in ex post performance evaluations.
Given a portfolio, the Sharpe ratio is defined as:
=
( )
( )
where E(rp) denotes the expected portfolio return, rf is the risk-free rate and (rp) is the
relevant measure of portfolio volatility. For ex post volatility we use the historical standard
deviation of returns; for the ex ante measure we integrate implied volatilities of underlying
portfolio components using historical co-variation. Realized volatility represents a statistical
measure of variability of the actual return distribution over a specific time horizon. In
contrast, implied volatility represents the markets best estimate of future volatility, given
todays information. It is usually implied from prices of liquidly traded options and will on
average be higher than realized volatility. This stylized fact, often referred to as the volatility
risk premium, discussed below, ensures sufficient compensation to risk-averse sellers of
options for their asymmetric payoffs.
Because we are interested in assessing the impact of FX hedging the bond component of
our market portfolio, we use the following modified version of the Sharpe ratio:
=
( )
( )
where h represents the cost of FX hedging and (0,1) is the share of the bond portfolio
that is hedged. We consider three values for : 1, 0.5 and 0; ie, full hedging, 50% hedging
and no hedging.
In our analysis, we use daily data from 1 January 2004 to 14 January 2015. For the equity
component of our portfolio, we use the MSCI ACWI index, which covers approximately 85%
of the global investable equity opportunity set and includes 23 developed and 23 emerging
See Sharpe, William F. "Mutual Fund Performance," The Journal of Business, Vol. XXXIX, No. 1, Part II, January 1966.
A Sharpe ratio is a sufficient measure of risk-adjusted performance if returns are normally distributed, but may not
fully describe the risk-return tradeoff if returns are not derived from a distribution fully characterized by its mean and
variance.
24 February 2015
145
market indices.2 Ex ante equity returns are calculated as the inverse of the forward-looking
price-to-earnings ratio (inverse P/E, or earnings yield) of the MSCI ACWI index, whereas ex
post returns are computed as the rolling annual return. Because MSCI ACWI options are not
regularly traded in the market, the ex ante volatility of the MSCI ACWI is approximated by
the implied volatility from options on the SPX, SX5e, UKX, NKY, and HIS. To calculate this
measure, we first construct a replicating portfolio consisting of fixed weights in those five
indices with the aim of minimizing a tracking error against the MXWD ACWI index. Using
the estimated weights, we combine the at-the-money-forward volatilities for each index.
We use the three-month point on the volatility curve. It should be noted that our weighted
volatility measure may somewhat overstate the true underlying volatility because
correlation between the five indices is disregarded in our construction.
For the fixed income component of our portfolio, we create a simplified variation of the
Barclays Global Aggregate Bond Index using its USD, EUR, JPY and GBP subindices. Our G4
fixed income portfolio represents approximately 92% of the Barclays Global Aggregate
Bond Index. Using only four currencies greatly simplifies the calculations, particularly in
capturing co-variation between the components, and the high proportion of the Barclays
Agg invested in these four currencies suggests that our proxy is a representative measure of
a realistic global fixed income portfolio. We use the weighted sum of yield to maturity of
each corresponding sub-index as the ex ante return of the bond component of the portfolio
and the weighted sum of annual returns as ex post bond returns. For an ex ante measure of
volatility, we use G4 3-month into five-year at-the-money normal swaption volatilities, and
in contrast to our equity measure we do account for co-variation across currencies as it is
important to our analysis of hedging decisions. We chose 3m5y swaptions since our
Barclays Global Agg index has roughly a five-year average duration over the sample period.
Using the above and three-month USD LIBOR for the risk-free rate, we calculate ex ante and
ex post rolling Sharpe ratios for a fairly typical market portfolio consisting of 60% equities
and 40% bonds. Ex ante Sharpe ratios represent the expected excess return of a market
portfolio per unit of predicted portfolio standard deviation, ie, the portfolios implied
volatility. In calculating this, we create a measure of implied portfolio volatility that
considers not just the variability of each individual portfolio component, equities, bonds and
FX, but also how the three components co-vary. The ex post Sharpe ratio is simply the
realized excess return of the 60/40 portfolio divided by the standard deviation of returns.
Theoretically, we expect
increases in the FX hedging
ratio to improve portfolio
Sharpe ratios by reducing
overall portfolio volatility
We assume a passive FX hedging strategy for a USD-based investor: FX forward contracts are
sold on a rolling basis in proportion to the foreign currency component of the bond portfolio.
Hedging costs, given by in the formula above, are calculated using 12-month forward
exchange rates. In the ex ante case we assume that the cost of the hedge reflects the 12month forward rate relative to current spot. In the ex post calculations, we incorporate the
cash flows resulting from rolling maturing contracts. We assume three different passive FX
hedging strategies: i) no hedging; ii) 50% hedging; and iii) 100% hedging. We assume the
equity component remains unhedged in all cases. International equities often are not hedged
due to the greater volatility of the underlying asset relative to currencies. Additionally, FX
returns and equity returns on average are positively correlated, in contrast to bond and FX
returns. We discuss this point in more detail below. Theoretically, we expect increases in the
FX hedging ratio to improve portfolio Sharpe ratios by reducing overall portfolio volatility.
Results
Figure 4 presents results for the entire sample. Average portfolio return, volatility and
Sharpe ratios are reported for each of the three hedging cases outlined above both in ex
ante and ex post terms. One of the more notable features of Figure 4 is the large
improvement in ex post risk-adjusted performance relative to ex ante expectations. For all
three hedge ratios, the ex post Sharpe ratio is roughly three times the ex ante ratio. The ex
24 February 2015
See http://www.msci.com/products/indexes/tools/index.html#ACWI
146
post improvement is apparent in both the numerator (returns) and the denominator
(volatility) as Figures 3, 5 and 6 show. This divergence also reflects four widely observed
factors: one persistent and three specific to our sample period.
FIGURE 4
Relative portfolio performance: 2004-15
Performance
measure
Hedge ratio
Return
0%
3.73
4.91
1.18
9.07
-2.69
Sharpe ratio
0.33
0.93
0.60
Return
3.86
4.97
1.10
Volatility
100%
Ex-post
11.76
Volatility
50%
Ex-ante
11.52
8.79
-2.73
Sharpe ratio
0.35
0.98
0.63
Return
4.00
5.03
1.03
11.44
8.49
-2.94
0.37
1.05
0.68
Volatility
Sharpe ratio
The persistent factor is the existence of a volatility risk premium that compensates risk-averse
sellers of options for bearing asymmetric payoffs. For this reason, there are few natural sellers
of volatility, but a wide range of buyers. The asymmetric nature of the risk taken by option
sellers is difficult to diversify and operationally intensive to manage. As a result, realized
volatility is, on average, lower than implied volatility (see The FX volatility risk premium:
Identifying drivers and investigating returns, 16 June 2014 and The Lesser Known Risk
Premium - Investing in volatility across asset classes, 19 November 2013).
Three other factors in our sample likely contributed to the ex post improvement in riskadjusted returns. First, there was a persistent trend rally in fixed income as real interest
rates declined almost monotonically through the sample. Second, the negative correlation
between bonds and equities was unusually strong and persistent through the sample,
particularly in times of market stress, as we discuss below.
Our results show a positive
contribution to risk-adjusted
returns from FX hedging, a
decision most managers can
make
Third, US interest rates were generally higher than other G4 interest rates during the period,
making hedging into USD via FX forwards profitable, on average, contrary to theoretical
expectations. Although accommodative monetary policy in the euro area and Japan may
cause this feature to persist in coming years, it is unlikely to be sustained in the long run.
FIGURE 5
Ex-ante portfolio Sharpe ratio for 100% hedging
FIGURE 6
Ex-post portfolio Sharpe ratio for 100% hedging
30
0.8
25
20
0.7
30
0.6
20
0.5
15
0.4
0.3
10
0.2
5
0
2004
2006
2008
24 February 2015
2010
2012
Excess returns
40
4.0
3.0
2.0
1.0
0.0
-10
-20
0.0
-30
2004
Volatility
5.0
10
0.1
2014
6.0
-1.0
-2.0
-3.0
2006
2008
2010
2012
Volatility
2014
Excess returns
147
The size of the ex ante/ex post difference is remarkable, but it has little bearing on the
portfolio decisions of most international asset managers. But our results also show a positive
contribution to risk-adjusted returns from FX hedging, a decision most international
managers can make. As we hypothesized, FX hedging generates higher Sharpe ratios by
reducing portfolio volatility, both ex ante and ex post on average. The average ex ante Sharpe
ratio increases from 0.33 in the case of no FX-hedging to 0.37 in the case of 100% FX hedging
of the foreign bond portfolio. Average ex post increases in Sharpe ratios are more impressive,
from 0.93 in the case of no hedging to 1.05 in the case of 100% FX hedging of the foreign
bond portfolio. Average ex ante portfolio volatility drops from 11.76% in the case of no FX
hedging to 11.44% in the case of 100% FX hedging of the foreign bond portfolio. Average ex
post reductions in portfolio volatility fall from 9.07% in the case of no hedging to 8.49% with
100% hedging of the foreign bond portfolio. Additionally, using rolling Sharpe ratios, we are
able to obtain empirical distributions3 of both ex ante and ex post Sharpe ratio (Figures 7
and 8). In both distributions there is a clear rightward shift in the distribution for the 100% FX
hedged portfolio, and for the ex post distribution a notable skew to higher Sharpe ratios.
A greater degree of FX hedging
can boost the negative
correlation between equity and
bond returns, thereby reducing
portfolio volatility
A key driver of this result is the way in which a greater degree of FX hedging can boost the
negative correlation between equity and bond returns, thereby reducing portfolio volatility as
high returns in one offset low returns in the other. This phenomenon is particularly apparent
during periods of elevated volatility, as in the 2008-11 global financial crisis. That increased
correlation also is noticeable in the period of increasing FX volatility and USD strength since
mid-2014. Figure 10 plots rolling conditional correlations of the residuals on equity, bond and
FX daily returns4. FX hedging the bond component of the portfolio leads to a pronounced
increase in the negative correlation between equity and bond residuals, particularly during
periods of heightened market volatility. Because currencies tend to be negatively correlated
with local bond returns higher interest rates lower bond prices but boost the currency
unhedged bond portfolios returns become less negatively correlated with equity returns.
A strengthening USD also boosted the negative bond/equity correlation. Between October
2007 and March 2009, the MSCI ACWI index fell by almost 60% in USD terms and had
recovered only about half of these losses by end-2011. At the same time, safe-haven demand
for the USD was hurting the returns of foreign bond holdings. Indeed, the USD appreciated by
about 15% against a weighted basket of EUR, JPY and GBP between March 2008 and March
2009 (Figure 11). Balanced portfolios with foreign bond holdings hedged back into USDs did
not suffer a drag on bond returns from USD appreciation, increasing the negative correlation
with equity returns and reducing portfolio volatility.
FIGURE 7
Empirical ex-ante Sharpe ratio distribution
0.007
FIGURE 8
Empirical ex-post Sharpe ratio distribution
Ex-post
fitted kernel
Ex-ante
fitted kernel
0.0010
0.006
0.005
0.0008
0.004
0.0006
0.003
0.0004
0.002
0.0002
0.001
0
-0.15 -0.03 0.09
No hedging
0.0000
-2.83 -1.64 -0.46 0.73
No hedging
1.91
3.10
50% hedging
4.29
5.47
6.66
100% hedging
To obtain the empirical densities we perform non-parametric kernel estimation using a Gaussian kernel.
By conditional correlation we mean the correlation between two variables conditional on fixing the value of another
variable. In our example we compute the conditional correlation of equity and bond innovations by fixing the FX
innovations and similarly for equity and FX and bond and FX.
24 February 2015
148
To establish this more formally we isolate periods of higher market volatility and USD strength
in our sample. We use 2008-11 as our high FX volatility environment and periods of material
USD Index appreciation (Jan 05-Dec 05; Apr 08-Feb 09; Dec 09-Jun 10; and Jun 14-present) as
our strengthening USD environment. Figures 9 and 12 show that in both cases lower portfolio
volatility and higher ex ante and ex post Sharpe ratios are achieved as FX hedging is increased:
i) During periods of higher volatility: The average ex ante Sharpe ratio increases from 0.33
in the case of no FX-hedging to 0.34 in the case of 100% FX hedging of the foreign bond
portfolio. Average ex post Sharpe ratios increase from 0.38 in the case of no hedging to
0.44 in the case of 100% FX hedging of the foreign bond portfolio. Average ex ante
portfolio volatility drops from 15.90% in the case of no FX-hedging to 15.51% in the case
of 100% FX hedging of the foreign bond portfolio. Average ex post portfolio volatility falls
from 13.22% in the case of no hedging to 12.68% in the case of 100% FX hedging of the
foreign bond portfolio.
ii) During periods of a stronger USD: The average ex ante Sharpe ratio increases from
0.33 in the case of no FX-hedging to 0.38 in the case of 100% FX hedging of the foreign
bond portfolio. Average ex post Sharpe ratios increase from 0.91 in the case of no
hedging to 1.10 in the case of 100% FX hedging of the foreign bond portfolio. Average
ex ante portfolio volatility drops from 12.62% in the case of no FX hedging to 12.28% in
the case of 100% FX hedging of the foreign bond portfolio. Average ex post portfolio
volatility falls from 10.06% in the case of no hedging to 9.42% in the case of 100% FX
hedging of the foreign bond portfolio.
FIGURE 9
Relative portfolio performance: Period of high FX volatility*
Hedge ratio
Performance measure
Return
0%
50%
100%
Ex-ante
Ex-post
4.86
0.32
-4.54
15.90
13.22
-2.68
Sharpe ratio
0.33
0.38
0.05
Return
4.86
0.46
-4.40
-2.65
Volatility
Volatility
15.61
12.95
Sharpe ratio
0.34
0.41
0.07
Return
4.85
0.60
-4.25
15.51
12.68
-2.83
0.34
0.44
0.10
Volatility
Sharpe ratio
FIGURE 10
The negative correlation between equity and bond returns
increases with hedging
Equity vs bond
Equity vs bond hedged
Bond vs FX
Equity vs FX
FIGURE 11
and was particularly apparent during the crisis
160
450
150
400
140
350
130
300
120
250
-0.4
110
200
-0.6
100
2002
0.4
0.2
0.0
-0.2
-0.8
2005
2007
24 February 2015
2009
2011
2013
2015
150
2004
2006
2008
Basket FX vs USD
2010
2012
2014
149
FIGURE 12
Relative portfolio performance: Periods of USD strength*
Hedge ratio
0%
50%
100%
Performance measure
Ex-ante
Ex-post
Return
3.78
3.11
-0.67
Volatility
12.62
10.06
-2.56
Sharpe ratio
0.33
0.91
0.58
Return
3.93
3.35
-0.58
Volatility
12.36
9.75
-2.62
Sharpe ratio
0.36
0.99
0.64
Return
4.07
3.59
-0.48
Volatility
12.28
9.42
-2.86
Sharpe ratio
0.38
1.10
0.71
Note: * We define these periods as: Jan 05-Dec 05; Apr 08-Feb 09; Dec 09Jun 10; Jun 14-present.
Source: Barclays Research
The analysis so far suggests that FX hedging is most beneficial in periods of market stress.
To see if this result holds in a more general sense, we examine sub-periods in our sample
that exclude higher volatility or a strengthening USD. These results are presented in Figure
13 and indicate that FX hedging is able to lower portfolio volatility and generate superior
risk-adjusted returns in normal times as well as periods of pronounced market stress.
Through these periods we find that the average ex ante Sharpe ratio increases from 0.31 in
the case of no FX-hedging to 0.36 in the case of 100% FX hedging of the foreign bond
portfolio. Average ex post Sharpe ratios increase from 1.16 in the case of no hedging to
1.30 in the case of 100% FX hedging of the foreign bond portfolio. Average ex ante
portfolio volatility drops from 10.09% in the case of no FX hedging to 9.82% in the case of
100% FX hedging of the foreign bond portfolio. Average ex post portfolio volatility falls
from 7.24% in the case of no hedging to 6.70% in the case of 100% FX hedging of the
foreign bond portfolio.
FIGURE 13
Relative portfolio performance: Periods excluding high FX volatility and USD strength*
Hedge ratio
0%
50%
100%
Volatility
Ex-post
3.10
7.35
-4.54
10.09
7.24
-2.68
Sharpe ratio
0.31
1.16
0.05
Return
3.28
7.36
-4.40
Volatility
9.89
6.98
-2.65
Sharpe ratio
0.34
1.22
0.07
Return
3.45
7.37
-4.25
Volatility
9.82
6.70
-2.83
Sharpe ratio
0.36
1.30
0.10
Note; *We define these periods as follows, Jan 04 Dec 04, Jan 06 Dec 07 and Jan 12 May 14.
Source: Barclays Research.
150
potential to further dampen (or accentuate) volatility in an international equity and bond
portfolio. The implications of active currency management are beyond the scope of this
analysis but should be noted.
Our results show a clear benefit to FX hedging of international bond portfolios in a global
portfolio. While one should not expect a continuation of the sample-specific factors that
boosted the absolute returns in our results (eg, a persistent global bond rally, higher US
interest rates), FX hedging of bonds in a balanced international portfolio should persistently
raise risk-adjusted returns by increasing the negative correlation between bonds and
equities. Furthermore, we show that this effect appears to be amplified during heightened
volatility and periods of USD strength, both of which we expect in coming years.
24 February 2015
151
CHAPTER 8
We analyse returns on equities, gilts and cash from end-1899 to end-2014. Index-linked gilt
returns are available from 1982, while corporate bonds begin in 1999. To deflate the nominal
returns, a cost-of-living index is computed using Bank of England inflation data from 1899 to
1914 and the Retail Price Index, calculated by the Office of National Statistics, thereafter.
FIGURE 1
Real investment returns by asset class (% pa)
Last
2014
10 years
20years
50years
Equities
-0.4
4.1
4.6
5.7
115 years*
5.0
Gilts
Corporate Bonds
16.4
10.7
3.7
2.5
5.1
2.9
1.3
Index-Linked
14.0
3.5
4.4
Cash
-1.2
-0.7
1.1
1.5
0.8
Figure 1 summarises the real investment returns of each asset class over various time
horizons. The first column provides the real returns over one year, the second column real
annualised returns over 10 years, and so on. UK equities had a lacklustre year and
underperformed other developed market indices in 2014. UK nominal total returns were just
1.2%, compared to 2.65% for the German DAX and 10.5% for US equities. The
underperformance occurred despite a reasonable growth backdrop. The UK was one of the
few economies where the consensus growth forecast was actually revised higher last year; US,
European and Global real GDP estimates had all been downgraded over the course of the year.
The Scottish Referendum contributed to some temporary underperformance in the FTSE AllShare index, but the key drag came from the disinflationary impact of the commodity price fall
and, in particular, the 50% decline in the oil price during the second half of the year. Much of
the performance drag on UK equities was driven by exposure to the oil- and mining-related
sectors which accounted for more than 20% of the FTSE All-Share market cap. In comparison,
the worst-performing sectors in the STOXX Europe 600 included oil and gas, and basic
resources, the combined weight of which stood at just 8.5%.
Fixed income and credit had a very strong performance in 2014 as a result of the deflationary
fears fuelled by the oil price decline. Nominal and inflation-linked gilts posted their best
returns since the Euro sovereign debt crisis in 2011. The long end outperformed in both gilts
and treasuries as the curves bull-flattened. Credit returns were the strongest since 2012.
Monetary policy divergence was a key theme driving bond markets in 2014. The prospect of
QE from the ECB caused European government bonds to outperform the US and the UK in
the 10-15 year sector. Cash returns remained weak in the low yield environment.
FIGURE 2
Real investment returns (% pa)
Equities
Gilts
Index-linked
1904-1914
1914-24
1924-34
1934-44
2.1
0.4
9.2
3.0
-0.1
-3.1
11.7
-1.4
1.5
-1.7
5.6
-2.4
1944-54
1954-64
1964-74
1974-84
5.3
7.1
-6.0
17.4
-2.6
-2.6
-6.3
5.6
-2.8
1.4
0.0
-0.3
1984-94
1994-2004
2004-2014
9.4
5.0
4.1
5.8
6.5
3.7
5.3
3.5
Cash
5.5
3.0
-0.7
24 February 2015
152
FIGURE 3
Distribution of real annual equity returns since 1899
FIGURE 4
Distribution of real annual gilt returns since 1899
14
10
9
12
10
7
6
5
6
4
3
1
0
0
-50 -42 -34 -26 -18 -10 -2
14 22 30 38 46 54
14 22 30 38 46 54
FIGURE 5
Distribution of real annual cash returns since 1899
FIGURE 6
Maximum and minimum real returns over various periods
30
Cash
Gilts
Equities
23 year
25
20 year
20
15
10 year
10
5 year
5
1 year
0
-50 -42 -34 -26 -18 -10 -2
14 22 30 38 46 54
0%
20%
40%
60%
80% 100%
Figure 2 breaks down real asset returns for consecutive 10-year intervals. Equities have
outperformed cash and bonds over the past decade, with an average annualised return of
4.1% since 2004. Cash, on the other hand, has delivered the worst returns since the
stagflationary 1970s. Ranking the annual returns and placing them into deciles provides a
clearer illustration of their historical significance. The results for 2014 are shown in Figure 7.
The equity portfolio is ranked in the seventh best decile since 1899, down from the third
decile in 2013, as a result of the poor performance in the second half of the year. Gilts and
linkers are ranked in the first and second deciles, a striking jump from the ninth decile in
2013, as deflationary fears appear to have wiped out the memory of the 2013 taper
tantrum. Cash remained weak as yields were held near zero.
FIGURE 7
2014 performance ranked by decile (1899-2014)
Decile
Equities
Gilts
Index-Linked
Cash
Note: Deciles ranking: 1 signifies the best 10% of the history, 10 the worst 10%. Source: Barclays Research
24 February 2015
153
Figures 3-5 illustrate the distribution of returns over the past 115 years, 2014 is highlighted
within each distribution. They show that equity returns have the widest dispersion, followed
by gilts and then cash. The observed distributions are in accordance with financial theory;
from an ex-ante perspective, we would apply the highest risk premium to equities, given
their perpetual nature and our uncertainty over future growth in corporate profits and
changes in the rate of inflation. For gilts, the uncertainty with respect to inflation remains,
but the risk from the perspective of coupon and principal is reduced, given their
government guarantee. Over the past 30 years, the dispersion of annual gilt returns has
widened significantly; in the 1970s and 1980s, an unexpected increase in the inflation rate
led to significant negative real returns, while in the 1990s, an unanticipated fall in inflation,
in conjunction with lower government deficits, facilitated above-average real returns. The
cash return index has the lowest dispersion. In recent years, the real returns to cash have
been relatively stable, with the move toward inflation-targeting by the Bank of England
stabilising the short-term real interest rate.
10
18
Outperform cash
77
79
81
83
96
97
Underperform cash
37
34
31
28
10
114
113
112
111
106
98
Probability of equity
outperformance
68%
70%
72%
75%
91%
99%
78
84
84
81
84
85
Outperform gilts
Underperform gilts
36
29
28
30
22
13
114
113
112
111
106
98
Probability of equity
outperformance
68%
74%
75%
73%
79%
87%
24 February 2015
154
Equities
Gilts
FIGURE 10
Todays value of 100 invested at the end of 1899, income
reinvested gross
Nominal
Real
14,597
184
Equities
0.75
Gilts
Cash
59
Nominal
Real
2,240,727
28,261
36,197
457
20,444
258
FIGURE 11
Five-year average dividend growth rates
20%
15%
10%
5%
0%
-5%
1945 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010
Source: Barclays Research
Turning to the dividend growth ratio, the FTSE All-Share dividend rose just 0.6% in 2014, the
slowest pace of growth in four years. Figure 11 shows that the five-year average growth rate
had picked up following the steady declines of recent years after corporates began cutting
dividends in 2008. In 1997-2001, dividend income had fallen by a cumulative 15% as
companies cut dividends on the basis that funds would be put to better use by corporates
than by shareholders. In the wake of the dotcom crash, investors actively sought incomeyielding stocks as a way to lower risk.
Figures 12 and 13 illustrate the time series of price indices and total return indices for
equities, gilts and cash over the entire series. These returns are in nominal terms and are
shown with the use of a logarithmic scale.
24 February 2015
155
FIGURE 12
Barclays price indices Nominal terms
FIGURE 13
Barclays total return indices Nominal terms, gross income
reinvested
10,000,000
100000
10000
1000
10,000
100
10
Equities
Gilts
Retail Prices
Equities
Gilts
2009
1999
1989
1979
1969
1959
1949
1939
1929
1919
1899
1909
10
2009
1999
1989
1979
1969
1959
1949
1939
1929
1919
1909
1899
T-Bills
FIGURE 14
Todays value of 100 invested at the end of 1945 without
reinvesting income
FIGURE 15
Todays value of 100 invested at the end of 1945, gross
income reinvested
Equities
Gilts
Source: Barclays Research
Nominal
Real
9,148
261
65
Nominal
Real
179,695
5,118
Gilts
7,773
221
Cash
6,261
178
Equities
FIGURE 16
Todays value of 100 invested at the end of 1990, gross income reinvested
Nominal
Real
Equities
750
379
Gilts
764
386
Index-Linked Gilts
577
291
Treasury Bills
300
151
24 February 2015
156
CHAPTER 9
We have analysed returns on US equities, government bonds and cash using 89 annual
return observations. The construction of the series is explained in more detail in the indices
in Chapter 10 (Barclays Indices). Corporate bond performance is captured using the
Barclays Investment Grade Corporate Long Index, which incorporates bonds with a maturity
of 10 years or more. The Barclays US Inflation Linked 15-year Plus Index is used to represent
the performance of TIPS. The nominal return series are deflated by the change in the
consumer price index, which is calculated by the Bureau of Labor Statistics. The first
holding period covered in this analysis is the calendar year 1926, representing money
invested at the end of 1925 and its value at the end of 1926.
FIGURE 1
Real investment returns (% pa)
Last
2014
10 years
20 years
50 years
89 years*
Equities
9.7
5.5
7.4
5.6
6.7
Government bond
23.0
5.1
6.0
3.4
2.6
0.9
0.5
TIPS
18.3
4.0
Corporate bond
14.9
4.8
5.9
Cash
-0.7
-0.7
0.3
*Note: Entire sample. Source: Centre for Research into Security Prices (CRSP) provided US asset return data for the past
14 years, Barclays Research
Figure 1 provides real annualised returns over various time horizons. US equity returns in
2014 outperformed both developed and emerging markets by a wide margin as domestic
growth remained robust. Despite periodic drags from global growth concerns and
deflationary fears, the upward momentum was maintained throughout the year. Strong
earnings growth, with US corporate profits reaching record highs, helped fuel the
outperformance. M&A activity also accelerated in the US as corporates took advantage of
strong balance sheets and the low rate environment. This, in turn, provided further support
for equity performance.
Fixed income markets followed the trends in the UK: nominal bonds were the best-performing
asset of 2014, producing a 23% real total return, in sharp contrast to the -13% of the previous
year, when investors first digested the prospect of monetary policy normalisation by the Fed.
Treasuries, TIPS and credit produced the best returns since the Euro sovereign debt crisis in
2011. As mentioned in Chapter 8, monetary policy divergence was a key theme of 2014, yet,
despite the prospect of Fed policy normalisation, US bonds still managed a strong performance.
FIGURE 2
Real investment returns (% pa)
Equities
Government bond
Corporate bond
Cash
1934-44
6.4
1.0
-2.6
1944-54
11.5
-1.6
-3.0
1954-64
10.7
0.1
1.0
1964-74
-4.1
-2.6
0.2
1974-84
8.2
-0.2
1.5
1984-94
9.6
7.8
7.6
2.0
1994-2004
9.3
7.0
7.0
1.4
2004-2014
5.5
5.1
4.8
-0.7
24 February 2015
157
Equities only marginally outperformed Treasuries and corporate bonds in the most recent
decade. A total real return of 5.5% is in line with the average returns of the past 50 years, but
below the average performance since 1925 of 6.7%. Treasuries and corporate bond returns
were also in the region of 5% over the past decade, so the gap between equity and bond
performance has closed substantially relative to prior decades. Equities best decades were in the
1950s and the 1980s. Bonds have enjoyed strong performance over the past three decades
relative to preceding decades, largely as a result of continued disinflation since the late 1970s.
The strong bond performance of 2014 has pulled the average returns for the past decade up
from 3.4% last year to 5.1%, comfortably higher than the long-run average of 2.6%.
Figure 3 ranks the relative performance of 2014 returns by deciles to get a clearer indication
of their historical significance. The US equity ranking has fallen from the second decile in
2013 to the sixth in 2014 as returns failed to match 2013s near-30% total return. Bonds
moved from the worst decile in 2013 to the best in 2014 as investors switched from fears of
Fed policy normalisation in 2013 to global deflationary concerns in 2014. Cash returns
remained weak, with negative real returns placing them in the seventh decile.
FIGURE 3
Comparison of 2014 real returns with historical performance ranked by decile
Decile
Equities
Government bonds
Cash
Note: Deciles ranking - 1 signifies the best 10% of the history, 10 the worst 10%. Source: CRSP, Barclays Research
Figures 4-6 plot the sample distributions over the past 89 years; 2014 is highlighted within
each distribution. These charts allow readers to appreciate the volatility of each asset class
while gaining an understanding of the distribution of the annual return observations. Clearly,
cash has exhibited the lowest volatility of each asset class, with bonds next and equities
having the highest dispersion of returns.
FIGURE 4
Distribution of real annual cash returns since 1925
FIGURE 5
Distribution of real annual bond returns since 1925
35
12
30
10
25
20
6
15
4
10
5
0
0
-50 -40 -30 -20 -10
24 February 2015
10
20
30
40
50
60
10
20
30
40
50
60
158
FIGURE 6
Distribution of real annual equity returns since 1925
FIGURE 7
Maximum and minimum real returns over different periods
7
Cash
Bonds
Equities
20 year
5
4
10 year
3
2
5 year
1
1 year
0
-50 -40 -30 -20 -10
10
20
30
40
50
60
-50%
-30%
-10%
10%
30%
50%
Figure 7 shows the extremes of the return distribution for various holding periods. The
volatility of equities over very short horizons is clearly demonstrated in the maximum and
minimum distributions of one-year returns. As we extend the holding period, the distribution
begins to narrow. Over the past 89 years, the worst average annualised 20-year return for
equities was 0.9%, while the best was 13%. However, this is not to say that it is impossible to
lose money by holding equities over a 20-year period, as the analysis is conducted on an expost basis. The figure simply highlights that such an occurrence seems unlikely, given equities
performance over the past 89 years.
In addition, we would expect the ex-ante equity risk premium to act as a cushion against
uncertainty in the long term. Bonds and cash have had negative returns on a 20-year
investment horizon, reflecting unexpected inflation surges at various times in the past century.
Figure 8 plots the US equity risk premium and shows that the 10-year annualised excess return
of equities over bonds has recovered from the lows of 2008 and remains in positive territory.
FIGURE 8
Equity-risk premium Excess return of equities relative to bonds (10y annualised)
20%
15%
10%
5%
0%
-5%
-10%
1935
1948
1961
1974
1987
2000
2013
24 February 2015
159
FIGURE 10
Barclays US total return indices in nominal terms with gross
income reinvested
1,000,000
1,000,000
10,000
1,000
100
Equities
Bonds
Consumer Prices
Equity
Bonds
2013
2005
1997
1989
1981
1973
1965
1957
1949
1941
1933
1925
2013
2005
1997
1989
1981
1973
1965
1957
1949
1941
1933
1925
Cash
FIGURE 11
Value of $100 invested at the end of 1925 without reinvesting income
Equities
Bonds
Nominal
Real
$14,328
$1,092
$140
$11
FIGURE 12
Value of $100 invested at the end of 1925 with income reinvested gross
Equities
Nominal
Real
$408,413
$31,134
Bonds
$13,327
$1,016
Cash
$2,043
$156
24 February 2015
160
CHAPTER 10
Barclays indices
Sreekala Kochugovindan
+44 (0)20 7773 2234
sreekala.kochugovindan@
barclays.com
We have calculated three indices showing: 1) changes in the capital value of each asset
class; 2) changes to income from these investments; and 3) a combined measure of the
overall return, on the assumption that all income is reinvested.
Additional series allow for the effects of inflation. The data for cash include building society
deposit rates and Treasury bills. The series on index-linked securities is based at December
1982 and the corporate bond index starts at the end of 1990.
24 February 2015
161
FIGURE 1
Equity Index constituents
Constituents at December 1899
Woolworth Ltd
Imperial Chemical Industries
Shell Transport & Trading Ltd
Courtaulds Ltd
Royal Insurance Co
Barclay & Company
Lloyds Bank
Prudential Assurance Co Ltd
Westminster Bank Ltd
Midland Bank Ltd
London & Lancashire Fire Ins. Co
North British & Mercantile In. Co Ltd
Reckitt & Sons Ltd
County of London Electric Supply Co
Unilever Ltd
Tate & Lyle Ltd
Alliance Assurance Company
Boots Pure Drug Co Ltd
Pearl Assurance Co
Marks & Spencer Ltd
Cory (WM.) & Son
National Bank Of Egypt
Consolidated Gold Fields Of South Africa
Bass, Ratcliff & Gretton Ltd
GeduldProp Mines Ltd
Sun Insurance Office
Bank Of Australasia
British South Africa Co
Chartered Bank Of India, Australia & China
North Eastern Elec Supply Co
The Equity Index is a weighted arithmetic average. In the Equity Index, the weights of the 30
constituent companies for each year are proportional to their market capitalisation at the
beginning of the year. Each year a fund was constructed. The number of shares in the fund
for each company was calculated so that its market value at the beginning of the year was
equal to the companys index weighting. The value of the fund was calculated annually at
the end of the year.
For 1899-1962, the Equity Income Index is based on the Barclays Equity Fund. The Income
Index relates to the dividend income actually received in the 12 months prior to the date of
the index. It is calculated by totalling the dividends paid on the shares in the fund. We
believe that it is the only published index based on actual income receipts.
From 1963 the Income Index is derived from the yield on the FTSE All-Share Index. Despite
a minimal discontinuity in the yield, in our view, this is the most representative method of
evaluating equity performance over the period. The dividend yield is quoted net from 1998,
with non-taxpayers no longer able to reclaim ACT.
24 February 2015
162
163
US asset returns
The US indices used in this study were provided by the Center for Research in Security
Prices (CRSP) at the Graduate School of Business of the University of Chicago. The valueweighted equity index covers all common stocks trading on the New York, Nasdaq, and
Arca Stock Exchanges, excluding ADRs. For the bond index, the CRSP has used software
which selects the bond that is closest to a 20-year bond in each month. The same
methodology has been employed for the 30-day T-Bill.
Total returns
In this study, we have shown the performance of representative investments in UK equities
and long gilts, with additional analysis of equivalent US returns in both monetary and real
(inflation adjusted) terms. The total returns to the investor, however, also include the income
on the investment. This is important throughout the study for comparability between asset
classes. For example, when constructing an index for a cash investment such as the UK
Treasury Bill Index, the 100 invested at the end of 1899 grew to approximately 104 by the
end of the following year. This full amount is reinvested and by the end of 1920 the value of
this investment had grown to about 190. In contrast, equity and bond market returns can be
split into two components: capital appreciation; and dividend income. The most commonly
quoted stock market indices usually include only the capital component of the return. In order
to calculate returns on a comparable basis, we need to include the returns obtained by
reinvesting this income. This is particularly important in looking at bonds where the scope for
capital appreciation is small, so almost all of the return will be from income. In this study, total
returns are calculated assuming income is reinvested at the end of the year.
Taxation
The total return to an investor depends crucially on the tax regime. The largest long-term
investors in the British equity and gilt markets are pension funds and similar institutions
that (until the abolition of the advance corporation tax (ACT) credit) have not suffered tax
on their income or capital; our main tables therefore make no allowance for tax until 1998,
which was the first full year that non-taxpayers were unable to reclaim the ACT credit. This
effectively reduced the dividend yield to non-taxpayers, and is reflected in our main tables
and gross total return series.
The personal investor must suffer tax. The net return to a building society account is
straightforward to compute. However, changes in the tax regime in recent years make the
net return to equity and gilt investment less straightforward to calculate on a consistent
basis. For example, the change to total return taxation for gilts means that it is
inappropriate to calculate a net total return on the basis of taxing income alone. Thus,
returns are quoted gross throughout, but for reference we also quote basic tax rates.
24 February 2015
164
For periods of one year, arithmetic and geometric averages will be the same. But over
longer periods, the geometric average is always less than the arithmetic average, except
when all the individual yearly returns are the same. For the mathematically minded, the
geometric return is approximately equal to the arithmetic return minus one-half the
variance of the arithmetic return.
Although geometric returns are appropriate to analyse the past, arithmetic returns should
be used to provide forecasts. Arithmetic averages provide the better unbiased estimator
of returns (for a statistical proof of this see Ian Coopers paper Arithmetic vs Geometric
Premium: setting discount rates for capital budgeting calculations, IFA Working Paper
174-93, April 1993).
FIGURE 2
Barclays price indices in nominal terms
FIGURE 3
Barclays price indices in real terms
300
100,000
250
10,000
200
1,000
150
100
100
10
50
1
1899 1913 1927 1941 1955 1969 1983 1997 2011
Equities
Source: Barclays Research
24 February 2015
Gilts
0
1899 1913 1927 1941 1955 1969 1983 1997 2011
Retail prices
Equities
Gilts
165
FIGURE 4
Barclays total return indices in nominal terms with gross
income reinvested
FIGURE 5
Barclays total return indices in real terms with gross income
reinvested
100,000
10,000,000
1,000,000
10,000
100,000
1,000
10,000
1,000
100
100
10
10
1
1899 1913 1927 1941 1955 1969 1983 1997 2011
Equities
Source: Barclays Research
24 February 2015
Gilts
T-Bills
1
1899 1913 1927 1941 1955 1969 1983 1997 2011
Equities
Gilts
T- Bills
166
FIGURE 6
Barclays UK Cost of Living Index
Change %
Year
1900
1901
1902
1903
1904
1905
1906
1907
1908
1909
1910
1911
1912
1913
1914
1915
1916
1917
1918
1919
1920
1921
1922
1923
1924
1925
1926
1927
1928
1929
1930
1931
1932
1933
1934
1935
1936
1937
1938
1939
1940
1941
1942
1943
1944
1945
1946
1947
1948
1949
1950
1951
1952
1953
1954
1955
1956
1957
24 February 2015
December
(1899=100)
103.3
103.3
106.7
106.7
106.7
106.7
100.0
110.0
113.3
113.3
113.3
116.7
120.0
120.0
120.0
148.3
175.8
212.5
244.7
250.3
299.2
221.4
200.2
196.9
201.3
196.9
199.1
188.0
186.9
185.8
172.4
164.6
159.1
159.1
160.2
163.5
168.0
178.0
173.5
192.4
216.9
223.6
222.5
221.4
223.6
225.8
226.9
234.2
245.7
254.3
262.4
294.0
312.7
316.0
328.5
347.7
358.3
374.9
In year
3.3
0.0
3.2
0.0
0.0
0.0
-6.2
10.0
3.0
0.0
0.0
2.9
2.9
0.0
0.0
23.6
18.5
20.9
15.2
2.3
19.6
-26.0
-9.5
-1.7
2.3
-2.2
1.1
-5.6
-0.6
-0.6
-7.2
-4.5
-3.4
0.0
0.7
2.1
2.7
6.0
-2.5
10.9
12.7
3.1
-0.5
-0.5
1.0
1.0
0.5
3.2
4.9
3.5
3.2
12.0
6.3
1.1
4.0
5.8
3.0
4.6
5y average
1.3
0.6
-0.7
0.6
1.2
1.2
1.2
3.1
1.8
1.1
1.1
5.5
8.6
12.1
15.3
15.8
15.1
4.7
-1.2
-4.3
-4.3
-8.0
-2.1
-1.3
-1.0
-1.6
-2.6
-3.7
-3.3
-3.2
-2.9
-1.1
0.4
2.3
1.8
3.7
5.8
5.9
4.6
5.0
3.0
0.8
0.3
1.0
2.1
2.6
3.0
5.3
6.0
5.2
5.3
5.8
4.0
3.7
Change %
Year
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
December
381.8
381.8
388.7
405.7
416.5
424.2
444.6
464.5
481.6
493.4
522.7
547.1
590.3
643.6
692.9
766.2
912.8
1140.0
1311.8
1471.1
1594.4
1869.3
2151.9
2411.2
2541.6
2676.7
2799.3
2958.5
3068.6
3182.0
3397.6
3659.5
4001.4
4180.0
4287.8
4369.3
4495.6
4640.3
4754.2
4926.6
5062.1
5151.4
5302.3
5339.2
5496.3
5650.2
5847.3
5976.6
6241.4
6493.9
6555.5
6712.5
7032.8
7371.5
7599.3
7802.6
7928.8
In year
1.8
0.0
1.8
4.4
2.6
1.9
4.8
4.5
3.7
2.5
5.9
4.7
7.9
9.0
7.7
10.6
19.1
24.9
15.1
12.1
8.4
17.2
15.1
12.0
5.4
5.3
4.6
5.7
3.7
3.7
6.8
7.7
9.3
4.5
2.6
1.9
2.9
3.2
2.5
3.6
2.8
1.8
2.9
0.7
2.9
2.8
3.5
2.2
4.4
4.0
0.9
2.4
4.8
4.8
3.1
2.7
1.6
5y average
3.9
3.1
2.3
2.5
2.1
2.1
3.1
3.6
3.5
3.4
4.3
4.2
4.9
6.0
7.0
7.9
10.8
14.1
15.3
16.3
15.8
15.4
13.5
12.9
11.6
10.9
8.4
6.6
4.9
4.6
4.9
5.5
6.2
6.4
6.1
5.2
4.2
3.0
2.6
2.8
3.0
2.8
2.7
2.3
2.2
2.2
2.6
2.4
3.2
3.4
3.0
2.8
3.3
3.4
3.2
3.5
3.4
167
FIGURE 7
Barclays UK Equity Index
Year
1899
1900
1901
1902
1903
1904
1905
1906
1907
1908
1909
1910
1911
1912
1913
1914
1915
1916
1917
1918
1919
1920
1921
1922
1923
1924
1925
1926
1927
1928
1929
1930
1931
1932
1933
1934
1935
1936
1937
1938
1939
1940
1941
1942
1943
1944
1945
1946
1947
1948
1949
1950
1951
1952
1953
1954
1955
1956
1957
24 February 2015
+8.3%
-7.9%
+1.3%
-2.7%
+8.0%
-0.7%
+6.1%
-4.7%
+1.3%
+6.3%
-2.1%
-2.9%
-1.4%
-7.1%
-4.4%
0.0%
-6.8%
+4.2%
+16.3%
+7.7%
-25.6%
-7.1%
+19.8%
-4.0%
+15.3%
+9.9%
+1.8%
+4.0%
+12.2%
-19.1%
-9.2%
-24.3%
+27.9%
+20.6%
+9.8%
+9.9%
+15.1%
-16.7%
-14.9%
-3.1%
-10.2%
+16.8%
+12.9%
+7.1%
+8.3%
+2.0%
+13.9%
-6.3%
-7.7%
-10.3%
+5.6%
+3.0%
-5.9%
+17.8%
+42.4%
+5.8%
-13.9%
-7.0%
Income
yield %
100
69
80
66
62
71
77
79
57
73
69
71
69
57
57
36
67
66
63
34
77
79
73
72
67
73
83
76
79
90
80
65
64
60
70
78
82
93
94
90
94
91
86
86
87
88
93
107
98
103
109
121
128
134
155
179
183
188
6.3
4.8
5.4
4.6
4.0
4.6
4.7
5.1
3.6
4.3
4.2
4.4
4.4
3.9
4.1
2.6
5.2
4.8
4.0
2.0
6.1
6.7
5.2
5.3
4.3
4.3
4.8
4.2
3.9
5.5
5.4
5.8
4.4
3.5
3.6
3.7
3.4
4.6
5.5
5.4
6.3
5.2
4.4
4.1
3.8
3.8
3.5
4.3
4.3
5.0
5.0
5.4
6.1
5.4
4.4
4.8
5.7
6.3
-30.6%
+15.6%
-17.3%
-6.1%
+13.7%
+8.5%
+2.9%
-27.4%
+26.5%
-4.5%
+2.1%
-3.2%
-16.5%
+0.1%
-37.8%
+88.2%
-2.2%
-3.6%
-47.0%
+128.9%
+2.7%
-7.9%
-0.8%
-7.5%
+10.3%
+12.5%
-8.2%
+3.9%
+14.9%
-11.0%
-18.7%
-2.4%
-5.6%
+15.7%
+11.5%
+5.8%
+12.7%
+1.8%
-4.8%
+4.8%
-3.6%
-4.5%
-0.2%
+0.4%
+2.0%
+4.9%
+15.1%
-7.7%
+4.4%
+5.6%
+11.2%
+6.3%
+4.3%
+16.0%
+15.4%
+2.2%
+2.8%
+4.8%
-7.9%
-1.9%
-2.7%
+8.0%
-0.7%
+13.2%
-13.3%
-1.7%
+6.3%
-2.1%
-5.7%
-4.2%
-7.1%
-4.4%
-19.1%
-21.4%
-13.8%
+1.0%
+5.3%
-37.8%
+25.5%
+32.5%
-2.4%
+12.8%
+12.4%
+0.7%
+10.1%
+12.9%
-18.6%
-2.1%
-20.8%
+32.4%
+20.6%
+9.0%
+7.7%
+12.1%
-21.4%
-12.7%
-12.6%
-20.3%
+13.3%
+13.4%
+7.7%
+7.3%
+1.0%
+13.3%
-9.2%
-12.1%
-13.3%
+2.3%
-8.1%
-11.5%
+16.6%
+36.9%
-0.0%
-16.5%
-11.1%
-30.6%
+11.9%
-17.3%
-6.1%
+13.7%
+15.7%
-6.4%
-29.5%
+26.5%
-4.5%
-0.8%
-5.8%
-16.5%
+0.1%
-49.7%
+58.8%
-19.1%
-16.3%
-48.2%
+91.4%
+38.8%
+1.8%
+0.9%
-9.5%
+12.7%
+11.2%
-2.8%
+4.5%
+15.6%
-4.2%
-14.8%
+1.0%
-5.6%
+14.9%
+9.2%
+3.0%
+6.4%
+4.4%
-14.2%
-7.1%
-6.5%
-4.0%
+0.3%
-0.6%
+1.0%
+4.4%
+11.6%
-12.1%
+0.8%
+2.3%
-0.7%
-0.0%
+3.2%
+11.6%
+9.1%
-0.8%
-1.7%
168
Year
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
24 February 2015
+41.1%
+49.5%
-2.6%
-3.0%
-4.4%
+15.2%
-10.0%
+5.9%
-9.3%
+28.7%
+43.5%
-15.2%
-7.5%
+41.9%
+12.8%
-31.4%
-55.3%
+136.3%
-3.9%
+41.2%
+2.7%
+4.3%
+27.1%
+7.2%
+22.1%
+23.1%
+26.0%
+15.2%
+22.3%
+4.2%
+6.5%
+30.0%
-14.3%
+15.1%
+14.8%
+23.3%
-9.6%
+18.5%
+11.7%
+19.7%
+10.9%
+21.2%
-8.0%
-15.4%
-25.0%
+16.6%
+9.2%
+18.1%
+13.2%
+2.0%
-32.8%
+25.0%
+10.9%
-6.7%
+8.2%
+16.7%
-2.1%
+7.5%
+12.1%
+21.7%
+3.5%
-0.4%
-6.5%
+13.7%
+7.7%
+0.5%
-2.5%
+6.1%
+0.8%
+5.5%
+5.1%
+9.3%
+3.9%
+9.6%
+10.4%
+12.8%
+16.1%
+12.6%
+23.8%
+12.8%
+3.5%
+9.0%
+8.1%
+20.6%
+12.8%
+14.1%
+11.4%
+16.1%
+17.0%
+10.5%
+5.6%
-0.5%
-4.4%
+7.9%
+12.0%
+9.9%
+3.4%
-14.2%
+2.8%
-3.2%
-0.2%
+1.3%
+1.8%
+7.5%
+14.2%
+9.7%
+7.7%
-0.1%
-10.9%
+0.2%
+13.6%
+9.8%
+7.2%
+0.6%
Income
yield %
4.8
3.6
4.5
4.8
5.0
4.1
5.1
5.2
5.8
4.4
3.2
3.9
4.4
3.3
3.2
4.8
11.7
5.5
6.4
5.3
5.8
6.9
6.1
5.9
5.3
4.6
4.4
4.3
4.0
4.3
4.7
4.2
5.5
5.0
4.4
3.4
4.0
3.8
3.7
3.2
2.5
2.1
2.2
2.6
3.6
3.1
3.1
3.0
2.9
3.0
4.5
3.2
2.9
3.5
3.6
3.3
3.4
+38.5%
+49.5%
-4.4%
-7.0%
-6.9%
+13.1%
-14.2%
+1.3%
-12.5%
+25.6%
+35.4%
-19.0%
-14.3%
+30.2%
+4.8%
-37.9%
-62.5%
+89.2%
-16.5%
+25.9%
-5.3%
-11.0%
+10.4%
-4.3%
+15.8%
+16.9%
+20.5%
+9.0%
+17.9%
+0.4%
-0.3%
+20.7%
-21.6%
+10.1%
+11.9%
+21.0%
-12.1%
+14.8%
+9.0%
+15.5%
+7.9%
+19.1%
-10.6%
-16.0%
-27.1%
+13.4%
+5.5%
+15.5%
+8.3%
-1.9%
-33.4%
+22.0%
+5.9%
-11.0%
+5.0%
+13.6%
-3.7%
+5.5%
+12.1%
+19.5%
-0.8%
-3.0%
-8.2%
+8.5%
+3.1%
-3.1%
-4.8%
+0.2%
-3.7%
-2.3%
-3.6%
+1.6%
-6.0%
-8.0%
-11.6%
-2.0%
+3.5%
+3.9%
+5.6%
-2.0%
-7.6%
+3.4%
+2.7%
+15.3%
+6.8%
+10.0%
+7.4%
+8.7%
+8.7%
+1.1%
+1.1%
-3.0%
-6.2%
+4.9%
+8.5%
+7.3%
-0.2%
-16.5%
+1.0%
-5.9%
-0.9%
-1.6%
-1.0%
+3.8%
+11.8%
+5.0%
+3.5%
-1.0%
-13.0%
-4.4%
+8.4%
+6.5%
+4.4%
-1.0%
169
FIGURE 8
Barclays UK Gilt Index
Gilt Price Index
December
Year
Yield %
1899
100.0
1900
98.4
-1.6%
2.8
95.2
-4.8%
1901
94.6
-3.8%
2.9
91.5
-3.8%
1902
93.7
-0.9%
3.0
87.8
-4.0%
1903
88.3
-5.8%
2.9
82.8
-5.8%
1904
89.4
+1.2%
2.8
83.8
+1.2%
1905
90.1
+0.8%
2.8
84.4
+0.8%
1906
86.6
-3.8%
2.9
86.6
+2.6%
1907
84.1
-2.9%
3.0
76.5
-11.7%
1908
84.6
+0.6%
3.0
74.7
-2.4%
1909
83.6
-1.3%
3.0
73.7
-1.3%
1910
80.0
-4.3%
3.1
70.6
-4.3%
1911
77.7
-2.8%
3.2
66.6
-5.6%
1912
75.8
-2.4%
3.3
63.2
-5.1%
1913
72.3
-4.7%
3.5
60.2
-4.7%
1914
73.0
+1.0%
3.4
60.9
+1.0%
1915
73.0
0.0
3.4
49.2
-19.1%
1916
55.7
-23.8%
4.5
31.7
-35.7%
1917
54.9
-1.4%
4.6
25.8
-18.4%
1918
59.4
+8.3%
4.2
24.3
-6.0%
1919
51.9
-12.7%
4.8
20.7
-14.6%
100.0
1920
45.6
-12.1%
5.5
15.2
-26.5%
1921
50.6
+11.1%
4.9
22.9
+50.2%
1922
56.2
+10.9%
4.4
28.1
+22.6%
1923
56.1
-0.2%
4.5
28.5
+1.5%
1924
57.7
+2.9%
4.3
28.6
+0.6%
1925
55.4
-3.9%
4.5
28.1
-1.7%
1926
54.5
-1.6%
4.6
27.4
-2.7%
1927
55.9
+2.6%
4.5
29.8
+8.7%
1928
56.7
+1.3%
4.4
30.3
+1.9%
1929
53.3
-6.0%
4.7
28.7
-5.4%
1930
57.8
+8.5%
4.3
33.5
+16.9%
1931
55.0
-4.7%
4.5
33.4
-0.2%
1932
74.7
+35.6%
3.3
46.9
+40.4%
1933
74.6
-0.1%
3.3
46.9
-0.1%
1934
92.8
+24.4%
2.7
57.9
+23.5%
1935
87.4
-5.8%
2.9
53.4
-7.8%
1936
85.1
-2.6%
2.9
50.7
-5.2%
1937
74.8
-12.2%
3.3
42.0
-17.1%
1938
70.7
-5.4%
3.5
40.8
-3.0%
1939
68.9
-2.6%
3.6
35.8
-12.2%
1940
77.4
+12.3%
3.2
35.7
-0.3%
1941
83.1
+7.4%
3.0
37.2
+4.2%
1942
82.9
-0.3%
3.0
37.2
+0.2%
1943
80.0
-3.4%
3.1
36.1
-3.0%
1944
82.1
+2.6%
3.0
36.7
+1.6%
1945
91.8
+11.8%
2.7
40.6
+10.7%
1946
99.2
+8.0%
2.5
43.7
+7.5%
1947
82.5
-16.8%
3.0
35.2
-19.4%
1948
80.6
-2.3%
3.1
32.8
-6.9%
1949
70.9
-12.0%
3.5
27.9
-15.0%
1950
71.3
+0.5%
3.5
27.2
-2.6%
24 February 2015
170
Year
Yield %
1951
61.9
-13.1%
4.0
21.1
-22.4%
1952
59.0
-4.8%
4.2
18.9
-10.5%
1953
64.7
+9.7%
3.9
20.5
+8.5%
1954
66.1
+2.2%
3.8
20.1
-1.7%
1955
56.9
-13.8%
4.4
16.4
-18.6%
1956
52.7
-7.5%
4.7
14.7
-10.2%
1957
46.9
-10.9%
5.3
12.5
-14.9%
1958
52.4
+11.7%
4.8
13.7
+9.6%
1959
50.4
-3.9%
5.0
13.2
-3.9%
1960
44.3
-11.9%
5.6
11.4
-13.5%
1961
38.3
-13.7%
6.5
9.4
-17.3%
1962
45.3
+18.3%
5.4
10.9
+15.3%
1963
44.5
-1.7%
5.5
10.5
-3.5%
1964
41.0
-7.9%
6.1
9.2
-12.1%
1965
40.3
-1.7%
6.2
8.7
-6.0%
1966
39.5
-2.1%
6.4
8.2
-5.5%
1967
37.9
-4.1%
6.9
7.7
-6.4%
1968
34.4
-9.3%
7.6
6.6
-14.4%
1969
31.7
-7.6%
8.5
5.8
-11.7%
1970
30.1
-5.2%
9.3
5.1
-12.2%
1971
35.4
+17.6%
8.3
5.5
+7.8%
1972
31.0
-12.3%
9.6
4.5
-18.5%
1973
25.3
-18.6%
11.9
3.3
-26.4%
1974
18.3
-27.5%
17.0
2.0
-39.2%
1975
21.8
+19.2%
14.8
1.9
-4.6%
1976
21.6
-1.1%
15.0
1.6
-14.0%
1977
28.2
+30.6%
10.9
1.9
+16.4%
1978
24.4
-13.3%
13.2
1.5
-20.0%
1979
22.2
-9.2%
14.7
1.2
-22.6%
1980
23.5
+6.2%
13.9
1.1
-7.8%
1981
20.7
-12.1%
15.8
0.9
-21.6%
1982
28.2
+36.2%
11.1
1.1
+29.2%
1983
29.5
+4.9%
10.5
1.1
-0.4%
1984
28.5
-3.4%
10.6
1.0
-7.7%
1985
28.7
+0.4%
10.5
1.0
-5.0%
1986
28.8
+0.4%
10.5
0.9
-3.2%
1987
30.6
+6.2%
9.5
1.0
+2.4%
1988
30.6
+0.0%
9.3
0.9
-6.3%
1989
29.4
-3.7%
10.0
0.8
-10.6%
1990
28.1
-4.5%
10.6
0.7
-12.7%
1991
30.4
+8.0%
9.8
0.7
+3.4%
1992
33.0
+8.7%
8.7
0.8
+6.0%
1993
39.4
+19.3%
6.4
0.9
+17.1%
1994
32.2
-18.1%
8.6
0.7
-20.4%
1995
35.5
+10.3%
7.6
0.8
+6.8%
1996
35.7
+0.6%
7.6
0.8
-1.8%
1997
40.0
+11.8%
6.3
0.8
+7.9%
1998
47.4
+18.6%
4.4
0.9
+15.4%
1999
43.4
-8.4%
5.3
0.8
-10.0%
2000
45.2
+4.0%
4.7
0.9
+1.0%
2001
43.4
-3.8%
5.0
0.8
-4.5%
2002
45.5
+4.8%
4.4
0.8
+1.8%
-5.8%
2003
44.1
-3.2%
4.7
0.8
2004
45.2
+2.5%
4.5
0.8
-1.0%
2005
47.0
+3.9%
4.1
0.8
+1.7%
24 February 2015
171
Year
Yield %
2006
44.8
-4.6%
4.7
0.7
2007
45.1
+0.6%
4.5
0.7
-3.3%
2008
48.8
+8.3%
3.4
0.7
+7.3%
2009
46.4
-5.0%
4.2
0.7
-7.3%
2010
48.7
+5.0%
3.6
0.7
+0.3%
2011
57.2
+17.4%
2.4
0.8
+12.0%
2012
57.9
+1.3%
2.2
0.8
-1.7%
2013
51.8
-10.6%
3.3
0.7
-12.9%
2014
59.3
+14.4%
2.1
0.7
+12.6%
24 February 2015
-8.6%
172
FIGURE 9
Barclays UK Treasury Bill Index
Treasury Bill Index
December
Year
1899
1900
1901
1902
1903
1904
1905
1906
1907
1908
1909
1910
1911
1912
1913
1914
1915
1916
1917
1918
1919
1920
1921
1922
1923
1924
1925
1926
1927
1928
1929
1930
1931
1932
1933
1934
1935
1936
1937
1938
1939
1940
1941
1942
1943
1944
1945
1946
1947
1948
1949
1950
1951
1952
1953
1954
1955
24 February 2015
100
104
107
110
114
117
119
123
128
130
133
137
141
144
148
153
158
162
167
172
179
190
199
204
210
217
226
237
247
257
271
278
289
293
295
297
298
300
302
304
308
311
314
317
320
324
327
328
330
332
333
335
337
344
352
359
371
+4.0%
+2.5%
+3.0%
+3.4%
+2.9%
+2.2%
+3.0%
+3.8%
+2.2%
+2.1%
+3.1%
+2.8%
+2.0%
+3.0%
+3.0%
+3.0%
+3.0%
+3.0%
+3.0%
+3.6%
+6.5%
+4.7%
+2.6%
+2.7%
+3.5%
+4.2%
+4.6%
+4.4%
+4.3%
+5.4%
+2.5%
+3.7%
+1.5%
+0.6%
+0.7%
+0.5%
+0.6%
+0.6%
+0.6%
+1.3%
+1.0%
+1.0%
+2.0%
+1.0%
+1.0%
+0.9%
+0.5%
+0.5%
+0.5%
+0.5%
+0.5%
+0.5%
+2.1%
+2.4%
+1.9%
+3.5%
+0.6%
+2.5%
-0.3%
+3.4%
+2.9%
+2.2%
+9.9%
-5.7%
-0.8%
+2.1%
+3.1%
-0.1%
-0.8%
+3.0%
+3.0%
-16.6%
-13.1%
-14.7%
-10.5%
+1.3%
-11.0%
+41.5%
+13.4%
+4.4%
+1.2%
+6.6%
+3.5%
+10.5%
+4.9%
+6.1%
+10.5%
+8.6%
+5.0%
+0.6%
+0.0%
-1.5%
-2.1%
-5.1%
+3.2%
-8.6%
-10.4%
-2.0%
+1.5%
+1.5%
+0.0%
-0.1%
+0.0%
-2.6%
-4.2%
-2.9%
-2.6%
-10.3%
-4.0%
+1.3%
-2.0%
-2.2%
173
Year
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
24 February 2015
+5.0%
+5.0%
+5.1%
+3.4%
+5.0%
+5.1%
+4.5%
+3.8%
+4.4%
+6.3%
+6.1%
+5.9%
+7.4%
+7.9%
+7.5%
+6.2%
+5.4%
+9.0%
+12.6%
+10.8%
+11.3%
+9.4%
+8.1%
+13.5%
+17.2%
+13.8%
+12.4%
+10.1%
+9.5%
+11.9%
+10.9%
+9.6%
+11.0%
+14.6%
+15.9%
+11.6%
+9.5%
+5.9%
+5.4%
+6.7%
+6.2%
+6.9%
+7.9%
+5.5%
+6.2%
+5.5%
+4.1%
+3.8%
+4.6%
+5.0%
+4.9%
+5.9%
+5.2%
+0.7%
+0.5%
+0.5%
+0.3%
+0.3%
+0.4%
+1.9%
+0.4%
+3.2%
+3.4%
+3.2%
+0.7%
+1.8%
+1.9%
-0.4%
+1.7%
+2.4%
+3.4%
+1.4%
+3.1%
-0.4%
-2.6%
-2.1%
-1.4%
-5.5%
-11.3%
-3.2%
-2.4%
-0.3%
-3.2%
+1.8%
+1.5%
+6.6%
+4.6%
+4.8%
+5.8%
+7.0%
+5.7%
+4.0%
+6.4%
+6.0%
+6.8%
+6.7%
+3.9%
+2.4%
+3.4%
+3.6%
+3.1%
+5.0%
+3.7%
+3.2%
+4.8%
+1.1%
+0.9%
+1.1%
+2.7%
+0.4%
+1.8%
+4.2%
-1.7%
-4.1%
-4.1%
-2.7%
-2.3%
-1.2%
174
FIGURE 10
Barclays UK Index-linked Gilt Index
Year
Real
yield %
Money
yield %
2.7
8.3
100
1982
100
1983
98.1
-1.9%
3.2
8.7
93.2
-6.8%
1984
101.6
+3.6%
3.3
8.1
92.3
-1.0%
1985
98.5
-3.1%
3.9
9.8
84.6
-8.3%
1986
101.4
+3.0%
4.1
7.9
84.0
-0.7%
1987
105.1
+3.6%
4.0
7.9
84.0
-0.1%
1988
116.0
+10.4%
3.8
10.8
86.8
+3.3%
1989
129.1
+11.3%
3.5
11.5
89.7
+3.3%
1990
130.8
+1.3%
4.0
13.8
83.1
-7.4%
1991
133.2
+1.8%
4.5
9.2
81.0
-2.5%
1992
151.1
+13.4%
3.9
6.6
89.6
+10.6%
1993
177.1
+17.2%
2.9
4.9
103.0
+15.0%
1994
158.3
-10.6%
4.0
7.0
89.5
-13.1%
1995
171.1
+8.1%
3.6
6.9
93.7
+4.7%
1996
176.2
+3.0%
3.6
6.1
94.2
+0.5%
1997
193.4
+9.8%
3.1
6.9
99.8
+5.9%
1998
227.4
+17.6%
2.0
4.8
114.2
+14.4%
1999
233.7
+2.8%
2.2
4.0
115.3
+1.0%
2000
235.4
+0.8%
2.3
5.3
112.9
-2.1%
2001
227.7
-3.3%
2.7
3.4
108.4
-4.0%
2002
240.7
+5.7%
2.1
5.1
111.3
+2.7%
2003
251.9
+4.7%
1.7
4.5
113.3
+1.8%
2004
267.6
+6.3%
1.7
5.3
116.3
+2.7%
2005
286.7
+7.1%
1.5
3.8
121.9
+4.8%
2006
287.0
+0.1%
1.6
6.0
116.9
-4.1%
2007
297.9
+3.8%
1.4
5.5
116.6
-0.3%
2008
290.3
-2.5%
1.4
2.3
112.5
-3.5%
2009
302.5
+4.2%
0.8
3.2
114.5
+1.8%
2010
328.3
+8.5%
0.4
5.2
118.6
+3.6%
2011
369.5
+12.5%
-0.5
4.2
127.4
+7.4%
2012
363.6
-1.6%
-0.5
2.6
121.6
-4.5%
2013
355.7
-2.2%
-0.2
2.5
115.9
-4.7%
2014
409.6
+15.2%
-0.8
0.8
131.3
+13.3%
24 February 2015
175
FIGURE 11
Barclays UK Equity, Gilt and Treasury Bill Funds
Equities
Year
Value of Fund
December
Gilts
Value of Fund
December
Treasury Bills
Value of Fund
December
1945
100
1946
118
+17.9%
117
+17.3%
111
+10.7%
110
+10.2%
101
+0.5%
100
+0.0%
1947
115
-2.3%
111
-5.3%
95
-14.3%
92
-16.9%
101
+0.5%
97
-2.6%
1948
111
-3.8%
102
-8.3%
96
+0.7%
88
-4.0%
102
+0.5%
93
-4.2%
1949
104
-5.8%
93
-8.9%
87
-8.9%
77
-12.0%
102
+0.5%
91
-2.9%
1950
116
+10.9%
100
+7.4%
91
+4.0%
78
+0.8%
103
+0.5%
88
-2.6%
1951
126
+8.5%
97
-3.1%
82
-9.6%
63
-19.3%
103
+0.5%
79
-10.3%
1952
126
-0.1%
91
-6.1%
81
-0.8%
59
-6.7%
105
+2.1%
76
-4.0%
1953
156
+24.2%
111
+22.9%
93
+14.0%
66
+12.8%
108
+2.4%
77
+1.3%
1954
232
+48.6%
159
+42.9%
98
+6.1%
67
+2.0%
110
+1.9%
75
-2.0%
1955
257
+10.9%
167
+4.8%
88
-10.1%
57
-15.0%
114
+3.5%
74
-2.2%
1956
234
-9.0%
147
-11.7%
85
-3.2%
54
-6.0%
119
+5.0%
75
+1.9%
1957
231
-1.1%
139
-5.5%
80
-6.2%
48
-10.4%
125
+5.0%
75
+0.4%
1958
342
+47.9%
202
+45.2%
94
+17.0%
55
+14.9%
132
+5.1%
78
+3.2%
1959
529
+54.8%
313
+54.8%
95
+0.9%
56
+0.9%
136
+3.4%
81
+3.4%
1960
539
+1.8%
313
-0.1%
88
-7.0%
51
-8.7%
143
+5.0%
83
+3.2%
1961
548
+1.7%
305
-2.5%
81
-8.1%
45
-11.9%
150
+5.1%
84
+0.7%
100
100
100
100
100
1962
550
+0.4%
298
-2.2%
101
+24.7%
55
+21.5%
157
+4.5%
85
+1.8%
1963
659
+19.9%
351
+17.7%
105
+3.7%
56
+1.8%
163
+3.8%
87
+1.9%
1964
623
-5.4%
317
-9.8%
102
-2.3%
52
-6.7%
170
+4.4%
87
-0.4%
1965
694
+11.4%
337
+6.6%
107
+4.4%
52
-0.1%
181
+6.3%
88
+1.7%
1966
666
-4.0%
312
-7.4%
111
+4.2%
52
+0.5%
192
+6.1%
90
+2.4%
1967
895
+34.3%
410
+31.1%
114
+2.6%
52
+0.1%
203
+5.9%
93
+3.4%
1968
1326
+48.1%
573
+39.8%
111
-2.4%
48
-7.8%
219
+7.4%
94
+1.4%
1969
1168
-11.9%
482
-15.9%
112
+0.2%
46
-4.2%
236
+7.9%
97
+3.1%
1970
1127
-3.5%
431
-10.5%
116
+3.6%
44
-4.0%
253
+7.5%
97
-0.4%
1971
1652
+46.5%
579
+34.4%
147
+27.3%
52
+16.8%
269
+6.2%
94
-2.6%
1972
1922
+16.4%
626
+8.1%
142
-3.8%
46
-10.7%
284
+5.4%
92
-2.1%
1973
1382
-28.1%
407
-35.0%
129
-8.9%
38
-17.6%
309
+9.0%
91
-1.4%
1974
690
-50.1%
171
-58.1%
109
-15.2%
27
-28.8%
348
+12.6%
86
-5.5%
1975
1719
+149.3%
341
+99.6%
150
+36.8%
30
+9.5%
386
+10.8%
76
-11.3%
1976
1759
+2.3%
303
-11.1%
170
+13.7%
29
-1.1%
429
+11.3%
74
-3.2%
1977
2614
+48.6%
401
+32.5%
247
+44.8%
38
+29.1%
470
+9.4%
72
-2.4%
1978
2839
+8.6%
402
+0.2%
242
-1.8%
34
-9.4%
508
+8.1%
72
-0.3%
1979
3165
+11.5%
382
-4.9%
252
+4.1%
30
-11.2%
576
+13.5%
70
-3.2%
1980
4268
+34.8%
448
+17.1%
305
+20.9%
32
+5.0%
675
+17.2%
71
+1.8%
1981
4846
+13.6%
454
+1.3%
310
+1.8%
29
-9.2%
768
+13.8%
72
+1.5%
1982
6227
+28.5%
553
+21.9%
469
+51.3%
42
+43.6%
863
+12.4%
77
+6.6%
1983
8019
+28.8%
676
+22.3%
544
+15.9%
46
+10.0%
950
+10.1%
80
+4.6%
1984
10552
+31.6%
851
+25.8%
581
+6.8%
47
+2.1%
1041
+9.6%
84
+4.8%
1985
12680
+20.2%
968
+13.7%
644
+11.0%
49
+5.0%
1165
+11.9%
89
+5.8%
1986
16139
+27.3%
1188
+22.7%
715
+11.0%
53
+7.0%
1292
+10.9%
95
+7.0%
1987
17536
+8.7%
1244
+4.8%
831
+16.3%
59
+12.1%
1416
+9.6%
100
+5.7%
1988
19552
+11.5%
1299
+4.4%
909
+9.4%
60
+2.4%
1572
+11.0%
104
+4.0%
1989
26498
+35.5%
1635
+25.8%
963
+5.9%
59
-1.7%
1801
+14.6%
111
+6.4%
24 February 2015
176
Equities
Year
Value of Fund
December
Gilts
Value of Fund
December
Treasury Bills
Value of Fund
December
1990
23947
-9.6%
1351
-17.4%
1017
+5.6%
57
-3.4%
2086
+15.9%
118
+6.0%
1991
28936
+20.8%
1563
+15.7%
1209
+18.9%
65
+13.8%
2328
+11.6%
126
+6.8%
1992
34672
+19.8%
1826
+16.8%
1432
+18.4%
75
+15.4%
2549
+9.5%
134
+6.7%
1993
44207
+27.5%
2285
+25.1%
1844
+28.8%
95
+26.4%
2698
+5.9%
139
+3.9%
1994
41590
-5.9%
2089
-8.6%
1635
-11.3%
82
-13.8%
2844
+5.4%
143
+2.4%
1995
51163
+23.0%
2490
+19.2%
1945
+19.0%
95
+15.3%
3035
+6.7%
148
+3.4%
1996
59275
+15.9%
2815
+13.1%
2095
+7.7%
100
+5.1%
3222
+6.2%
153
+3.6%
1997
73263
+23.6%
3358
+19.3%
2503
+19.4%
115
+15.3%
3444
+6.9%
158
+3.1%
1998
83284
+13.7%
3715
+10.6%
3129
+25.0%
140
+21.7%
3717
+7.9%
166
+5.0%
1999
103120
+23.8%
4520
+21.7%
3018
-3.5%
132
-5.2%
3921
+5.5%
172
+3.7%
2000
97023
-5.9%
4132
-8.6%
3296
+9.2%
140
+6.1%
4165
+6.2%
177
+3.2%
2001
84226
-13.2%
3562
-13.8%
3340
+1.3%
141
+0.6%
4394
+5.5%
186
+4.8%
2002
65440
-22.3%
2689
-24.5%
3668
+9.8%
151
+6.7%
4575
+4.1%
188
+1.1%
2003
78643
+20.2%
3143
+16.9%
3725
+1.6%
149
-1.2%
4747
+3.8%
190
+0.9%
2004
88508
+12.5%
3418
+8.8%
3994
+7.2%
154
+3.6%
4964
+4.6%
192
+1.1%
2005
107609
+21.6%
4066
+18.9%
4329
+8.4%
164
+6.0%
5213
+5.0%
197
+2.7%
2006
125243
+16.4%
4531
+11.4%
4323
-0.1%
156
-4.4%
5468
+4.9%
198
+0.4%
2007
131639
+5.1%
4577
+1.0%
4550
+5.2%
158
+1.2%
5789
+5.9%
201
+1.8%
2008
92460
-29.8%
3185
-30.4%
5135
+12.9%
177
+11.8%
6091
+5.2%
210
+4.2%
2009
119238
+29.0%
4011
+25.9%
5087
-1.0%
171
-3.3%
6133
+0.7%
206
-1.7%
2010
136107
+14.1%
4370
+8.9%
5565
+9.4%
179
+4.4%
6163
+0.5%
198
-4.1%
2011
131469
-3.4%
4027
-7.8%
6755
+21.4%
207
+15.8%
6195
+0.5%
190
-4.1%
2012
147384
+12.1%
4379
+8.7%
7078
+4.8%
210
+1.6%
6215
+0.3%
185
-2.7%
2013
177620
+20.5%
5140
+17.4%
6569
-7.2%
190
-9.6%
6236
+0.3%
180
-2.3%
2014
179695
+1.2%
5118
-0.4%
7773
+18.3%
221
+16.4%
6261
+0.4%
178
-1.2%
24 February 2015
177
FIGURE 12
Barclays UK Treasury Bills and Building Society Accounts
Year
Treasury Bills
Annual
Return %
1946
0.51
6.51
46.25
1947
0.51
6.36
45.00
Year
1948
0.51
6.36
45.00
1949
0.52
6.36
45.00
1950
0.52
6.36
45.00
1990
15.86
12.04
25.00
1951
0.52
4.82
46.88
1991
11.59
9.32
25.00
1952
2.09
4.65
47.50
1992
9.47
9.59
24.68
1953
2.36
4.60
45.62
1993
5.86
4.12
24.50
1954
1.89
4.55
45.00
1994
5.40
3.69
20.00
1955
3.50
4.69
43.12
1995
6.74
3.93
20.00
1956
5.02
5.44
42.50
1996
6.16
2.61
20.00
1957
5.01
6.09
42.50
1997
6.88
3.06
20.00
1958
5.11
6.09
42.50
1998
7.92
7.06
20.00
1959
3.42
5.59
39.69
1999
5.51
5.11
23.00
1960
5.04
5.52
38.75
2000
6.22
5.50
22.00
1961
5.14
5.81
38.75
2001
5.50
4.70
22.00
1962
4.46
6.12
38.75
2002
4.12
3.40
22.00
1963
3.80
5.81
38.75
2003
3.75
3.33
22.00
1964
4.40
5.71
38.75
2004
4.59
4.21
22.00
1965
6.29
6.50
40.62
2005
5.00
3.95
22.00
1966
6.12
6.81
41.25
2006
4.90
4.36
22.00
1967
5.90
7.23
41.25
2007
5.87
4.77
22.00
1968
7.43
7.52
41.25
2008
5.23
0.85
20.00
1969
7.93
8.29
41.25
2009
0.68
0.25
20.00
1970
7.45
8.51
41.25
2010
0.50
0.20
20.00
1971
6.18
8.25
39.38
2011
0.51
0.20
20.00
1972
5.42
8.16
38.75
2012
0.32
0.20
20.00
1973
9.01
9.70
32.19
2013
0.34
0.20
20.00
1974
12.56
11.07
32.25
2014
0.39
0.25
20.00
1975
10.75
11.01
34.50
1976
11.34
10.65
35.00
1977
9.44
10.65
34.25
1978
8.06
9.42
33.25
1979
13.45
12.22
30.75
1980
17.17
15.00
30.00
1981
13.76
12.94
30.00
1982
12.38
12.19
30.00
1983
10.14
9.64
30.00
1984
9.55
9.99
30.00
1985
11.87
10.81
30.00
1986
10.95
10.55
29.26
1987
9.58
9.66
27.50
1988
11.01
8.26
25.50
1989
14.55
10.71
25.00
Note:
1. Annual returns on Treasury bills are based on four consecutive investments in 91-day bills.
2. The building society rate of interest above is gross of tax.
24 February 2015
178
FIGURE 13
Barclays Index-linked Funds
Index Linked gilts
Value of Fund December
1982
100
1983
101
+0.8%
96
-4.3%
1984
107
+6.6%
98
+1.9%
1985
107
-0.2%
92
-5.5%
1986
114
+6.1%
94
+2.3%
1987
122
+6.9%
97
+3.1%
1988
138
+13.7%
103
+6.5%
1989
158
+14.5%
110
+6.3%
1990
165
+4.4%
105
-4.5%
1991
174
+5.2%
106
+0.7%
1992
204
+17.1%
121
+14.1%
1993
247
+21.1%
144
+18.9%
1994
227
-7.9%
128
-10.5%
1995
254
+12.0%
139
+8.5%
1996
271
+6.5%
145
+4.0%
1997
307
+13.4%
158
+9.4%
1998
369
+20.3%
186
+17.1%
1999
388
+5.0%
191
+3.2%
2000
400
+3.1%
192
+0.1%
2001
396
-0.9%
189
-1.6%
2002
428
+8.2%
198
+5.1%
2003
457
+6.8%
206
+3.9%
2004
497
+8.6%
216
+4.9%
2005
542
+9.1%
231
+6.7%
2006
554
+2.3%
226
-2.1%
2007
585
+5.5%
229
+1.4%
2008
578
-1.2%
224
-2.1%
2009
610
+5.6%
231
+3.1%
2010
673
+10.3%
243
+5.3%
2011
808
+19.9%
278
+14.4%
2012
834
+3.3%
279
+0.2%
2013
824
-1.3%
268
-3.9%
2014
954
+15.9%
306
+14.0%
24 February 2015
179
FIGURE 14
Barclays US Equity Index
Year
Income
Yield %
1925
100
1926
104
+4.3%
100
5.3
105
+5.5%
100
1927
132
+26.6%
119
+19.0%
5.0
137
+29.6%
121.7074
+21.7%
1928
177
+33.7%
132
+11.3%
4.2
185
+35.3%
137.1045
+12.7%
1929
144
-18.2%
98
-26.3%
3.8
150
-18.7%
100.5217
-26.7%
1930
98
-32.1%
80
-17.7%
4.6
109
-27.5%
88.3734
-12.1%
1931
51
-47.7%
54
-32.6%
5.9
63
-42.3%
65.70176
-25.7%
1932
44
-14.1%
55
+1.7%
7.0
60
-4.2%
74.44178
+13.3%
1933
66
+50.9%
53
-4.4%
4.4
90
+49.8%
70.63025
-5.1%
1934
66
-1.0%
50
-5.7%
4.2
88
-2.4%
65.63299
-7.1%
1935
92
+39.6%
71
+42.2%
4.3
119
+35.6%
90.62134
+38.1%
1936
116
+26.7%
95
+34.1%
4.5
149
+24.9%
119.7611
+32.2%
1937
72
-38.1%
69
-27.4%
5.3
90
-39.8%
84.57095
-29.4%
1938
89
+23.0%
70
+1.6%
4.4
113
+26.5%
88.37328
+4.5%
1939
86
-2.9%
75
+7.1%
4.8
110
-2.9%
94.62172
+7.1%
1940
75
-12.8%
79
+5.7%
5.9
95
-13.4%
99.3167
+5.0%
100
1941
63
-16.1%
81
+1.9%
7.1
73
-23.7%
92.04035
-7.3%
1942
69
+9.1%
87
+8.3%
7.1
73
+0.0%
91.41445
-0.7%
1943
84
+21.6%
80
-8.6%
5.3
86
+18.1%
81.17977
-11.2%
1944
96
+15.5%
90
+12.7%
5.2
97
+12.9%
89.42189
+10.2%
1945
129
+33.5%
98
+9.0%
4.2
127
+30.6%
95.32691
+6.6%
1946
116
-10.2%
86
-12.6%
4.1
96
-24.0%
70.50384
-26.0%
1947
113
-2.3%
115
+34.5%
5.7
87
-10.2%
87.15088
+23.6%
1948
108
-4.1%
125
+8.1%
6.4
81
-6.9%
91.50819
+5.0%
1949
122
+12.1%
156
+25.6%
7.2
92
+14.5%
117.3343
+28.2%
1950
148
+21.7%
194
+24.3%
7.3
106
+14.9%
137.6241
+17.3%
1951
169
+14.3%
178
-8.3%
5.9
114
+7.8%
119.0512
-13.5%
1952
182
+7.4%
182
+2.2%
5.6
122
+6.6%
120.734
+1.4%
1953
173
-5.0%
175
-3.8%
5.7
115
-5.7%
115.3204
-4.5%
1954
247
+43.4%
225
+28.5%
5.1
166
+44.4%
149.2457
+29.4%
1955
298
+20.4%
228
+1.1%
4.3
199
+20.0%
150.3591
+0.7%
1956
311
+4.4%
225
-1.4%
4.0
202
+1.3%
144.0197
-4.2%
1957
267
-14.1%
205
-8.6%
4.3
168
-16.5%
127.8885
-11.2%
1958
372
+39.3%
270
+31.6%
4.0
231
+36.9%
165.3901
+29.3%
1959
406
+9.1%
240
-11.1%
3.3
247
+7.2%
144.5078
-12.6%
1960
397
-2.2%
251
+4.5%
3.5
238
-3.5%
148.9918
+3.1%
1961
490
+23.3%
266
+5.9%
3.0
292
+22.5%
156.7545
+5.2%
1962
425
-13.3%
262
-1.3%
3.4
250
-14.4%
152.7379
-2.6%
1963
497
+17.1%
291
+11.0%
3.3
288
+15.2%
166.7967
+9.2%
1964
561
+12.8%
310
+6.6%
3.1
322
+11.8%
176.0222
+5.5%
1965
623
+11.0%
343
+10.6%
3.1
350
+8.9%
190.9574
+8.5%
1966
550
-11.7%
327
-4.7%
3.3
299
-14.6%
175.8412
-7.9%
1967
686
+24.7%
381
+16.5%
3.1
362
+21.0%
198.7692
+13.0%
1968
761
+10.9%
404
+6.1%
3.0
384
+5.9%
201.3076
+1.3%
1969
658
-13.5%
361
-10.5%
3.1
312
-18.6%
169.5865
-15.8%
1970
636
-3.4%
413
+14.4%
3.6
286
-8.5%
183.7472
+8.4%
24 February 2015
180
Year
1971
717
+12.8%
389
1972
819
+14.3%
1973
646
-21.2%
1974
445
-31.1%
1975
587
+31.8%
1976
715
1977
663
1978
1979
Income
Yield %
-5.9%
3.0
312
+9.2%
167.4684
-8.9%
405
+4.0%
2.8
345
+10.5%
168.4823
+0.6%
344
-15.0%
3.0
250
-27.5%
131.7533
-21.8%
348
+1.1%
4.4
154
-38.6%
118.5986
-10.0%
453
+30.3%
4.3
189
+23.3%
144.5524
+21.9%
+21.9%
515
+13.7%
4.0
220
+16.3%
156.752
+8.4%
-7.3%
553
+7.3%
4.6
191
-13.1%
157.6118
+0.5%
685
+3.3%
629
+13.8%
5.1
181
-5.3%
164.4744
+4.4%
810
+18.3%
764
+21.4%
5.2
189
+4.4%
176.2464
+7.2%
1980
1030
+27.1%
910
+19.2%
4.9
214
+13.0%
186.6452
+5.9%
1981
944
-8.4%
804
-11.7%
4.7
180
-15.9%
151.3247
-18.9%
1982
1078
+14.2%
1059
+31.7%
5.5
198
+10.0%
191.9924
+26.9%
1983
1271
+17.9%
936
-11.6%
4.1
225
+13.6%
163.4811
-14.9%
1984
1257
-1.1%
985
+5.3%
4.4
214
-4.9%
165.5295
+1.3%
1985
1589
+26.5%
1141
+15.8%
4.0
260
+21.8%
184.7011
+11.6%
1986
1777
+11.8%
1096
-3.9%
3.4
288
+10.6%
175.5055
-5.0%
1987
1753
-1.4%
1012
-7.6%
3.2
272
-5.5%
155.2007
-11.6%
1988
1980
+13.0%
1452
+43.5%
4.1
294
+8.2%
213.2228
+37.4%
1989
2456
+24.0%
1594
+9.8%
3.6
349
+18.5%
223.712
+4.9%
1990
2225
-9.4%
1454
-8.8%
3.6
298
-14.6%
192.3235
-14.0%
1991
2885
+29.6%
1640
+12.8%
3.2
374
+25.8%
210.4701
+9.4%
1992
3061
+6.1%
1533
-6.5%
2.8
386
+3.1%
191.1616
-9.2%
1993
3330
+8.8%
1547
+0.9%
2.6
409
+5.9%
187.7915
-1.8%
1994
3221
-3.3%
1502
-2.9%
2.6
385
-5.8%
177.6459
-5.4%
1995
4268
+32.5%
1876
+24.9%
2.4
498
+29.2%
216.3456
+21.8%
1996
5069
+18.8%
1876
+0.0%
2.1
572
+15.0%
209.4134
-3.2%
1997
6498
+28.2%
2011
+7.2%
1.7
721
+26.0%
220.6758
+5.4%
1998
7831
+20.5%
2082
+3.5%
1.5
855
+18.6%
224.8463
+1.9%
1999
9682
+23.6%
2308
+10.9%
1.3
1030
+20.4%
242.7806
+8.0%
2000
8507
-12.1%
1688
-26.9%
1.1
875
-15.0%
171.6951
-29.3%
2001
7448
-12.4%
1779
+5.4%
1.3
754
-13.8%
178.1937
+3.8%
2002
5801
-22.1%
1660
-6.7%
1.6
574
-23.9%
162.4155
-8.9%
2003
7587
+30.8%
2511
+51.3%
1.8
737
+28.4%
241.1224
+48.5%
2004
8409
+10.8%
2970
+18.3%
2.0
791
+7.3%
276.2662
+14.6%
2005
8862
+5.4%
2929
-1.4%
1.8
806
+1.9%
263.4717
-4.6%
2006
10106
+14.0%
3474
+18.6%
1.9
896
+11.2%
304.7062
+15.7%
2007
10638
+5.3%
3674
+5.8%
1.9
907
+1.1%
309.6048
+1.6%
2008
6420
-39.65%
2639
-28.18%
2.3
547
-39.71%
222.1679
-28.24%
2009
8223
+28.08%
3767
+42.76%
2.6
682
+24.69%
308.76
+38.98%
2010
9476
+15.23%
3692
-2.00%
2.2
774
+13.54%
298.1215
-3.45%
2011
9181
-3.11%
3438
-6.88%
2.1
728
-5.89%
269.6247
-9.56%
2012
10368
+12.92%
4719
+37.29%
2.5
808
+10.99%
363.8265
+34.94%
2013
13238
+27.68%
5233
+10.89%
2.2
1017
+25.79%
397.4745
+9.25%
2014
14328
+8.23%
5444
+4.02%
2.1
1092
+7.42%
410.3615
+3.24%
24 February 2015
181
FIGURE 15
Barclays US Bond Index
Bond Price Index
December
Year
Yield %
1925
100
1926
104
+3.9%
3.5
105
+5.1%
1927
110
+5.4%
3.2
113
+7.8%
100
1928
106
-3.1%
3.4
111
-2.0%
1929
106
-0.2%
3.4
110
-0.8%
1930
107
+1.3%
3.3
119
+8.2%
1931
98
-8.5%
4.1
120
+0.9%
1932
111
+12.9%
3.2
151
+25.8%
1933
107
-3.1%
3.4
146
-3.9%
1934
115
+6.8%
2.9
153
+5.2%
1935
117
+2.1%
2.8
152
-0.8%
1936
122
+4.6%
2.6
157
+3.1%
1937
119
-2.5%
2.7
148
-5.2%
1938
123
+2.8%
2.5
157
+5.8%
1939
127
+3.5%
2.3
163
+3.5%
1940
132
+3.8%
1.9
167
+3.0%
1941
131
-1.0%
2.0
151
-10.0%
+0.7%
2.4
139
-7.6%
1942
131
1943
131
-0.4%
2.5
135
-3.3%
1944
131
+0.3%
2.4
132
-1.9%
1945
142
+8.1%
2.0
140
+5.8%
1946
139
-2.4%
2.1
115
-17.4%
1947
132
-4.9%
2.4
101
-12.6%
1948
133
+0.9%
2.4
99
-2.0%
1949
138
+4.0%
2.1
105
+6.2%
1950
135
-2.3%
2.2
97
-7.8%
1951
127
-6.3%
2.7
86
-11.6%
1952
125
-1.4%
2.8
84
-2.1%
1953
126
+0.9%
2.7
84
+0.2%
1954
131
+4.1%
2.6
88
+4.9%
1955
126
-3.6%
3.0
84
-4.0%
1956
115
-9.1%
3.4
75
-11.7%
1957
120
+4.7%
3.2
76
+1.8%
1958
110
-8.4%
3.8
68
-10.0%
1959
103
-6.4%
4.4
63
-8.0%
1960
112
+9.0%
3.8
68
+7.5%
1961
109
-3.4%
4.0
65
-4.0%
1962
113
+4.0%
3.8
67
+2.6%
1963
108
-4.3%
4.1
63
-5.8%
1964
109
+0.4%
4.1
62
-0.6%
1965
104
-3.9%
4.4
59
-5.7%
1966
104
+0.0%
4.5
57
-3.3%
1967
94
-9.9%
5.2
50
-12.6%
1968
89
-14.9%
5.7
45
-21.1%
1969
79
-11.1%
6.6
37
-16.3%
1970
85
+7.0%
6.2
38
+1.4%
24 February 2015
182
Year
Yield %
1971
95
+12.2%
4.5
41
+8.6%
1972
96
+1.3%
4.5
40
-2.1%
1973
88
-8.8%
7.1
34
-16.1%
1974
84
-3.8%
7.7
29
-14.4%
1975
83
-1.7%
7.7
27
-8.0%
1976
91
+9.8%
6.9
28
+4.7%
1977
86
-6.0%
7.5
25
-11.9%
1978
77
-10.3%
8.8
20
-17.7%
1979
69
-10.0%
9.9
16
-20.5%
1980
60
-13.3%
11.6
12
-22.9%
1981
53
-11.5%
13.7
10
-18.7%
1982
65
+23.3%
10.5
12
+18.8%
1983
59
-9.4%
11.6
10
-12.7%
1984
61
+2.5%
11.3
10
-1.4%
1985
72
+18.7%
9.3
12
+14.3%
1986
84
+16.1%
7.6
14
+14.8%
1987
75
-11.0%
8.8
12
-14.8%
1988
74
-0.6%
8.8
11
-4.8%
1989
81
+9.5%
7.9
12
+4.6%
1990
79
-2.8%
8.2
11
-8.4%
1991
86
+9.1%
7.3
11
+5.9%
1992
86
-0.3%
7.3
11
-3.1%
1993
93
+8.8%
6.4
11
+5.9%
1994
80
-14.3%
7.9
10
-16.5%
1995
97
+21.1%
5.9
11
+18.1%
1996
90
-7.0%
6.6
10
-10.0%
1997
97
+7.7%
5.9
11
+5.9%
1998
103
+6.1%
5.3
11
+4.4%
1999
88
-14.5%
6.7
-16.8%
2000
100
+13.3%
5.5
10
+9.6%
2001
98
-2.1%
5.7
10
-3.6%
2002
108
+10.5%
4.8
11
+7.9%
2003
105
-2.9%
5.0
10
-4.7%
2004
107
+2.4%
4.8
10
-0.8%
2005
110
+2.2%
4.6
10
-1.2%
2006
105
-4.1%
4.8
-6.5%
2007
109
+4.1%
4.5
-0.0%
2008
131
+19.8%
3.1
11
+19.7%
2009
107
-17.9%
4.5
-20.1%
2010
113
+4.8%
4.1
+3.3%
2011
137
+21.7%
2.5
11
+18.2%
2012
138
+0.4%
2.7
11
-1.3%
2013
116
-15.4%
3.7
-16.7%
2014
140
+20.2%
2.4
11
+19.3%
24 February 2015
183
FIGURE 16
Barclays US Treasury Bill Index
Year
1925
100
1926
103
+3.2%
104
+4.4%
1927
106
+3.1%
110
+5.5%
1928
110
+3.8%
116
+5.0%
100
1929
116
+4.7%
120
+4.1%
1930
118
+2.3%
132
+9.3%
1931
120
+1.0%
147
+11.4%
1932
121
+0.8%
165
+12.3%
1933
121
+0.3%
164
-0.5%
1934
121
+0.2%
162
-1.3%
1935
121
+0.2%
157
-2.7%
1936
122
+0.2%
155
-1.3%
1937
122
+0.3%
152
-2.5%
1938
122
+0.0%
156
+2.9%
1939
122
+0.0%
156
+0.0%
1940
122
-0.1%
155
-0.8%
1941
122
+0.0%
141
-9.0%
1942
122
+0.3%
130
-8.0%
1943
123
+0.3%
126
-2.5%
1944
123
+0.3%
124
-1.9%
1945
124
+0.3%
121
-1.9%
1946
124
+0.4%
103
-15.1%
1947
125
+0.5%
95
-7.7%
1948
126
+1.0%
93
-2.0%
1949
127
+1.1%
96
+3.2%
1950
129
+1.2%
92
-4.5%
1951
131
+1.5%
88
-4.3%
1952
133
+1.6%
89
+0.9%
1953
135
+1.8%
90
+1.0%
1954
136
+0.9%
91
+1.6%
1955
138
+1.6%
92
+1.2%
1956
142
+2.4%
92
-0.5%
1957
146
+3.1%
92
+0.2%
1958
148
+1.4%
92
-0.3%
1959
152
+2.8%
93
+1.1%
1960
156
+2.6%
94
+1.2%
1961
160
+2.2%
95
+1.5%
1962
164
+2.7%
97
+1.4%
1963
169
+3.2%
98
+1.5%
1964
175
+3.5%
101
+2.5%
1965
182
+4.0%
103
+2.0%
1966
191
+4.7%
104
+1.2%
1967
199
+4.1%
105
+1.1%
1968
209
+9.7%
105
+0.5%
1969
223
+6.6%
106
+0.4%
1970
237
+6.4%
107
+0.8%
1971
247
+4.3%
108
+1.0%
24 February 2015
184
Year
1972
257
+3.9%
108
+0.5%
1973
275
+7.1%
107
-1.5%
1974
297
+8.1%
103
-3.8%
1975
315
+5.8%
102
-1.0%
1976
331
+5.2%
102
+0.3%
1977
348
+5.2%
100
-1.5%
1978
373
+7.3%
99
-1.6%
1979
413
+10.7%
96
-2.3%
1980
461
+11.5%
96
-0.9%
1981
529
+14.9%
101
+5.4%
1982
586
+10.7%
107
+6.6%
1983
638
+8.8%
113
+4.9%
1984
701
+10.0%
119
+5.8%
1985
755
+7.7%
124
+3.7%
1986
801
+6.1%
130
+4.9%
1987
844
+5.4%
131
+0.9%
1988
897
+6.3%
133
+1.8%
1989
971
+8.2%
138
+3.4%
1990
1046
+7.7%
140
+1.5%
1991
1103
+5.5%
143
+2.4%
1992
1141
+3.4%
144
+0.5%
1993
1174
+2.9%
144
+0.1%
1994
1219
+3.9%
146
+1.2%
1995
1287
+5.5%
150
+2.9%
1996
1353
+5.1%
153
+1.8%
1997
1422
+5.1%
158
+3.3%
1998
1490
+4.8%
163
+3.1%
1999
1558
+4.6%
166
+1.8%
2000
1647
+5.8%
169
+2.3%
2001
1710
+3.8%
173
+2.2%
2002
1738
+1.6%
172
-0.7%
2003
1755
+1.0%
170
-0.8%
2004
1776
+1.2%
167
-2.0%
2005
1829
+3.0%
166
-0.4%
2006
1916
+4.8%
170
+2.2%
2007
2006
+4.7%
171
+0.6%
2008
2036
+1.5%
173
+1.4%
2009
2038
+0.1%
169
-2.6%
2010
2040
+0.1%
167
-1.4%
2011
2041
+0.04%
162
-2.8%
2012
2042
+0.06%
159
-1.7%
2013
2043
+0.03%
157
-1.5%
2014
2043
+0.02%
156
-0.7%
24 February 2015
185
CHAPTER 11
Our final chapter presents a series of tables showing the performance of equity and fixedinterest investments over any period since December 1899.
The first section reviews the performance of each asset class, taking inflation into account,
since December 1960. On each page, we provide two tables illustrating the same
information in alternative forms. The first table shows the average annual real rate of return;
the second shows the real value of a portfolio at the end of each year, which includes
reinvested income. This section provides data on equities and gilts, with dividend income
reinvested gross. Finally, we provide figures for Treasury bills and building society shares.
The final pullout section provides the annual real rate of return on UK and US equities and
bonds (with reinvestment of income for each year since 1899 for the UK, and since 1925 for
the US). There is also a table showing the real capital value of equities for the UK. The
sources for all data in this chapter are the Barclays indices, as outlined in Chapter 8.
1960-2014
UK: 1899-2014
US: 1925-2014
24 February 2015
186
3.9
7.3
17.7
1964
0.3
1.3
3.0
1965
1.5
2.6
4.2
(1.9)
1966 (0.0)
0.5
1.2
1967
3.9
5.0
6.6
1968
7.8
9.4
1969
4.9
5.9
7.1
5.4
8.8
9.3
1970
3.3
3.9
4.7
3.0
5.3
1971
5.8
6.6
7.7
6.5
9.0
1972
6.0
6.8
7.7
6.7
1973
2.1
2.4
2.9
1.5
(9.8)
3.9
6.6
9.0
10.2 31.1
15.5
8.5 (15.9)
5.0
8.4
9.4
13.2
9.1
0.4
9.7
34.4
8.9
9.2
12.3
8.9
2.3
9.1
20.5
2.8
2.4
3.9
8.1
1974 (4.2) (4.4) (4.5) (6.3) (6.0) (7.3) (7.3) (11.8) (18.3) (18.7) (20.7) (33.5) (47.8) (58.1)
1975
1.0
(0.2)
0.6
0.8
0.1
(1.1) (0.4) (1.0) (0.3) (3.3) (7.7) (6.4) (5.7) (12.2) (16.6) (9.4) 33.2 (11.1)
0.7
0.1
1.0
1977
1.5
1.7
2.0
1.0
1.8
1.5
2.3
8.5
32.5
1978
1.4
1.6
1.9
0.9
1.7
1.4
2.1
5.7
15.2
1979
1.1
1.3
1.5
0.5
1.3
0.9
1.6
2.9
8.1
1980
1.8
2.0
2.3
1.4
2.2
1.9
2.6
0.7
(2.0) (0.7)
0.4
(2.8) (4.1)
1.4
17.4
5.6
10.3
3.7
5.5
1981
1.8
2.0
2.2
1.4
2.1
1.9
2.5
0.7
(1.8) (0.5)
0.5
(2.4) (3.5)
1.4
15.0
4.9
8.4
3.1
4.1
8.9
1982
2.6
2.9
3.1
2.4
3.2
3.0
3.6
2.0
(0.2)
1.1
2.1
(0.4) (1.2)
3.5
15.8
7.2
10.6
6.6
8.3
1983
3.4
3.7
4.0
3.3
4.1
3.9
4.6
3.2
1.1
2.5
3.5
1.3
0.7
5.2
16.5
9.0
12.2
9.1
1984
4.3
4.6
4.9
4.3
5.1
5.0
5.7
4.4
2.5
3.9
5.0
3.0
2.6
6.9
17.4 10.7 13.8 11.3 13.3 17.4 17.4 23.3 24.0 25.8
0.2
(2.4) (4.9)
17.1
1.3
1985
4.6
4.9
5.3
4.7
5.5
5.4
6.1
4.9
3.1
4.5
5.5
3.7
3.4
7.5
17.1 11.0 13.8 11.6 13.4 16.7 16.7 20.8 20.5 19.6 13.7
1986
5.3
5.6
5.9
5.4
6.2
6.2
6.9
5.8
4.1
5.4
6.5
4.9
4.7
8.6
17.5 12.0 14.6 12.8 14.5 17.6 17.7 21.2 21.0 20.6 18.1 22.7
1987
5.2
5.6
5.9
5.4
6.1
6.1
6.8
5.7
4.2
5.4
6.4
4.9
4.7
8.3
16.5 11.4 13.7 12.0 13.4 15.9 15.7 18.3 17.6 16.5 13.5 13.4
4.8
1988
5.2
5.5
5.8
5.4
6.1
6.0
6.7
5.7
4.2
5.4
6.3
4.9
4.7
8.0
15.6 10.9 12.9 11.3 12.4 14.6 14.2 16.2 15.3 13.9 11.2 10.3
4.6
1989
5.9
6.2
6.5
6.1
6.8
6.8
7.5
6.5
5.1
6.3
7.3
5.9
5.8
9.1
16.3 11.9 13.9 12.4 13.6 15.6 15.5 17.4 16.7 15.8 13.9 14.0 11.2 14.6 25.8
1990
5.0
5.3
5.5
5.1
5.7
5.7
6.3
5.3
4.0
5.0
5.9
4.6
4.4
7.3
13.8
8.0
6.9
3.3
2.8
2.0 (17.4)
1991
5.3
5.6
5.9
5.5
6.1
6.1
6.7
5.7
4.5
5.5
6.3
5.1
4.9
7.8
13.9 10.0 11.6 10.2 11.0 12.4 12.0 13.2 12.2 11.0
9.1
8.3
5.6
5.9
6.4
1992
5.7
5.9
6.2
5.9
6.5
6.5
7.0
6.2
5.0
6.0
6.8
5.6
5.5
8.2
14.1 10.4 11.9 10.6 11.4 12.8 12.4 13.5 12.7 11.7 10.0
9.5
7.4
8.0
8.9
3.7
16.2 16.8
1993
6.2
6.5
6.8
6.4
7.1
7.1
7.6
6.8
5.7
6.7
7.5
6.4
6.4
9.0
14.6 11.2 12.6 11.5 12.3 13.6 13.4 14.4 13.8 12.9 11.6 11.3
9.8
10.7 11.9
8.7
1994
5.7
6.0
6.3
5.9
6.5
6.5
7.0
6.2
5.1
6.0
6.8
5.7
5.6
8.1
13.3 10.0 11.3 10.2 10.8 12.0 11.6 12.5 11.7 10.8
9.4
8.9
7.3
7.7
8.2
5.0
11.5 10.1
1995
6.1
6.4
6.6
6.3
6.9
6.9
7.4
6.7
5.6
6.5
7.3
6.3
6.2
8.6
13.6 10.5 11.7 10.7 11.3 12.4 12.1 12.9 12.3 11.5 10.2
9.9
8.6
9.1
9.7
7.3
4.4
19.2
1996
6.3
6.6
6.8
6.5
7.1
7.1
7.6
6.9
5.9
6.8
7.5
6.5
6.5
8.8
13.6 10.6 11.8 10.8 11.4 12.5 12.2 12.9 12.3 11.6 10.5 10.2
9.0
9.5
10.1
8.1
7.2
16.1 13.1
1997
6.6
6.9
7.2
6.9
7.4
7.4
8.0
7.3
6.3
7.2
7.9
7.0
6.9
9.2
13.8 11.0 12.1 11.2 11.8 12.8 12.6 13.3 12.8 12.1 11.1 10.9
9.9
10.4 11.1
9.4
1998
6.7
7.0
7.3
7.0
7.5
7.5
8.0
7.4
6.4
7.3
8.0
7.1
7.1
9.2
13.7 10.9 12.1 11.2 11.8 12.7 12.5 13.2 12.6 12.0 11.1 10.9 10.0 10.5 11.1
9.5
1999
7.1
7.4
7.6
7.4
7.9
7.9
8.4
7.8
6.9
7.7
8.4
7.6
7.6
9.7
14.0 11.4 12.5 11.6 12.2 13.1 12.9 13.6 13.2 12.6 11.8 11.6 10.8 11.3 12.0 10.7 14.4 14.2 13.8 12.0 16.7 16.1 17.1 16.0 21.7
2000
6.7
6.9
7.2
6.9
7.4
7.4
7.9
7.3
6.4
7.2
7.8
7.0
7.0
9.0
13.0 10.5 11.5 10.7 11.2 12.0 11.8 12.3 11.8 11.2 10.4 10.2
9.3
9.7
10.1
8.8
8.8
7.2
2001
6.1
6.3
6.6
6.3
6.8
6.8
7.2
6.6
5.7
6.5
7.0
6.2
6.2
8.1
11.9
9.4
10.4
9.5
9.9
9.7
8.8
8.5
7.6
7.8
8.1
6.7
9.2
8.6
7.7
5.7
7.9
6.2
1.5
2002
5.3
5.5
5.7
5.4
5.8
5.8
6.2
5.5
4.7
5.3
5.9
5.1
5.0
6.7
10.3
8.0
8.8
7.9
8.2
8.8
8.5
8.8
8.2
7.5
6.6
6.2
5.2
5.3
5.3
3.9
5.9
5.1
3.9
1.8
3.2
1.1
2003
5.5
5.7
5.9
5.6
6.1
6.0
6.4
5.8
5.0
5.7
6.2
5.4
5.3
7.0
10.6
8.3
9.1
8.2
8.6
9.2
8.8
9.2
8.6
8.0
7.1
6.8
5.9
6.0
6.1
4.8
6.7
6.0
5.1
3.2
4.6
3.0
1.6
2004
5.6
5.8
6.0
5.7
6.1
6.1
6.5
5.9
5.1
5.8
6.3
5.5
5.4
7.1
10.5
8.3
9.0
8.3
8.6
9.2
8.8
9.2
8.6
8.0
7.2
6.9
6.0
6.1
6.2
5.0
6.9
6.2
5.4
3.7
5.0
3.6
2.5
0.3
2005
5.9
6.1
6.3
6.0
6.4
6.4
6.8
6.2
5.4
6.1
6.6
5.9
5.8
7.5
10.8
8.6
9.4
8.6
8.9
9.5
9.2
9.6
9.1
8.5
7.7
7.4
6.7
6.8
6.9
5.9
7.6
7.1
6.4
4.9
6.2
5.0
4.2
2.4
1.3
2006
6.0
6.2
6.4
6.1
6.5
6.5
6.9
6.4
5.6
6.2
6.8
6.1
6.0
7.6
10.8
8.7
9.4
8.7
9.0
9.6
9.3
9.6
9.2
8.6
7.9
7.6
6.9
7.0
7.2
6.2
7.9
7.4
6.7
5.4
6.7
5.6
4.9
3.4
2.5
0.0
2007
5.9
6.1
6.3
6.0
6.4
6.4
6.8
6.2
5.5
6.1
6.6
5.9
5.8
7.4
10.5
8.5
9.2
8.5
8.7
9.3
9.0
9.3
8.8
8.3
7.6
7.3
6.6
6.7
6.9
5.9
7.4
6.9
6.3
5.1
6.2
5.2
4.5
3.1
2.3
0.2
2008
5.0
5.1
5.3
5.0
5.4
5.4
5.7
5.1
4.4
5.0
5.4
4.7
4.6
6.1
9.0
7.0
7.6
6.9
7.1
7.6
7.3
7.5
7.0
6.4
5.7
5.3
4.6
4.6
4.6
3.6
4.9
4.3
3.5
2.2
3.1
1.9
1.0
2009
5.3
5.5
5.7
5.4
5.8
5.8
6.1
5.6
4.9
5.4
5.9
5.2
5.1
6.6
9.4
7.5
8.1
7.5
7.7
8.1
7.9
8.1
7.6
7.1
6.4
6.1
5.4
5.5
5.5
4.6
5.9
5.4
4.7
3.6
4.4
3.5
2.8
1.5
0.7
(1.2) (0.3)
2010
5.4
5.6
5.8
5.5
5.9
5.9
6.2
5.7
5.0
5.5
6.0
5.3
5.2
6.6
9.4
7.6
8.2
7.5
7.7
8.2
7.9
8.1
7.7
7.2
6.5
6.2
5.6
5.6
5.7
4.8
6.0
5.6
5.0
3.9
4.7
3.8
3.2
2.0
1.4
(0.3)
2011
5.1
5.3
5.5
5.2
5.6
5.5
5.8
5.3
4.6
5.2
5.6
5.0
4.9
6.2
8.9
7.1
7.7
7.0
7.2
7.6
7.3
7.5
7.1
6.6
5.9
5.6
5.0
5.0
5.0
4.2
5.3
4.8
4.3
3.2
3.9
3.1
2.4
1.3
0.6
2012
5.2
5.4
5.5
5.3
5.6
5.6
5.9
5.4
4.7
5.3
5.7
5.1
5.0
6.3
8.9
7.1
7.7
7.1
7.3
7.7
7.4
7.6
7.1
6.7
6.0
5.8
5.1
5.2
5.2
4.4
5.5
5.0
4.5
3.5
4.2
3.4
2.8
1.8
2013
5.4
5.6
5.7
5.5
5.9
5.8
6.1
5.7
5.0
5.5
5.9
5.3
5.3
6.5
9.1
7.4
8.0
7.3
7.6
7.9
7.7
7.9
7.5
7.0
6.4
6.1
5.6
5.6
5.7
4.9
6.0
5.6
5.1
4.1
4.9
4.1
3.6
2014
5.3
5.5
5.6
5.4
5.7
5.7
6.0
5.5
4.9
5.4
5.8
5.2
5.1
6.4
8.9
7.2
7.7
7.1
7.3
7.7
7.4
7.6
7.2
6.7
6.2
5.9
5.4
5.4
5.4
4.7
5.7
5.3
4.8
3.9
4.6
3.9
3.4
24 February 2015
9.6
11.3
9.8
4.4
(2.2) 15.7
7.0
(8.6)
4.8
5.5
(8.6)
8.8
3.4
1.5
4.9
1.5
4.3
11.2
9.9
10.2
2.9
0.3
1.5
5.9
4.1
3.3
2.3
6.3
4.8
4.2
(1.0) (0.2)
1.2
4.6
3.1
2.4
1.2
(0.2)
0.5
1.9
5.0
3.8
3.1
1.1
2.7
2.2
0.9
1.7
3.1
6.1
5.0
4.6
3.0
1.8
2.5
2.0
0.8
1.5
2.8
5.5
4.5
4.1
2.6
1.5
6.1
1.0
8.9
8.1
0.2
(7.8)
8.3
3.0
0.1
8.7
2.0
10.0
6.4
5.6
13.0 17.4
1.6
8.2
5.0
4.0
8.3
(0.6) (0.9)
8.1 (0.4)
187
1961
97
1962
95
98
1963
112
115
118
1964
101
104
106
90
1965
108
111
113
96
1966
100
102
105
89
99
93
1967
131
134
137
117
129
121
131
1968
183
188
192
163
181
170
183
140
1969
154
158
162
137
152
143
154
118
84
1970
138
141
145
123
136
128
138
105
75
89
1971
185
190
194
165
183
172
185
141
101
120
134
1972
200
205
210
179
198
186
201
153
109
130
145
108
1973
130
134
137
116
129
121
130
99
71
85
94
70
65
1974
55
56
57
49
54
51
55
42
30
35
40
29
27
42
1975
109
112
114
97
108
101
109
83
59
71
79
59
54
84
200
1976
97
99
102
86
96
90
97
74
53
63
70
52
48
74
177
89
1977
128
132
135
114
127
119
128
98
70
83
93
69
64
98
235
118
133
1978
128
132
135
115
127
119
129
98
70
83
93
69
64
99
236
118
133
1979
122
125
128
109
121
113
122
93
67
79
89
66
61
94
224
112
126
95
95
1980
143
147
150
128
141
133
143
109
78
93
104
77
71
110
262
132
148
112
111
117
1981
145
149
152
129
143
134
145
111
79
94
105
78
72
111
266
133
150
113
113
119
1982
177
181
186
158
175
164
177
135
97
115
128
95
88
136
324
162
183
138
138
145
124
122
1983
216
222
227
193
214
200
217
165
118
140
157
117
108
166
396
199
223
169
168
177
151
149
122
1984
272
279
286
243
269
252
272
208
149
177
197
147
136
209
499
250
281
212
212
223
190
188
154
126
1985
309
317
325
276
306
287
310
236
169
201
224
167
154
238
567
284
320
241
241
253
216
213
175
143
114
1986
380
390
398
339
375
352
380
290
207
246
275
205
190
292
696
349
392
296
295
311
265
262
215
176
140
123
1987
398
408
417
355
393
369
398
304
217
258
289
215
199
305
729
365
411
310
310
325
278
274
225
184
146
129
105
1988
415
426
436
370
411
385
416
317
227
270
301
224
207
319
762
382
429
324
323
340
290
286
235
192
153
134
109
1989
523
536
549
466
517
485
523
399
286
339
379
282
261
401
958
480
540
408
407
428
365
360
296
242
192
169
138
131
126
1990
432
443
453
385
427
400
433
330
236
280
313
233
216
332
792
397
446
337
336
353
302
298
244
200
159
140
114
109
104
1991
500
513
524
446
494
463
500
382
273
324
363
270
250
384
916
459
516
390
389
409
349
344
283
231
184
162
132
126
120
96
116
1992
584
599
613
521
577
541
584
446
319
379
423
315
291
603
455
454
478
408
402
330
270
215
189
154
147
141
112
135
117
1993
730
749
766
651
722
677
731
558
399
474
530
394
365
755
569
568
597
510
503
413
338
268
236
192
184
176
140
169
146
125
1994
668
685
701
596
660
619
669
510
365
434
485
361
333
690
521
520
546
467
460
378
309
245
216
176
168
161
128
155
134
114
91
1995
796
817
835
710
787
738
797
608
435
517
577
430
397
822
621
619
651
556
549
450
368
292
257
210
200
192
152
184
159
136
109
119
1996
900
923
944
803
889
834
901
687
492
584
653
486
449
930
702
700
736
629
620
509
416
331
291
237
226
217
172
208
180
154
123
135
113
586
697
779
579
536
835
878
750
740
607
496
394
347
283
270
258
205
248
215
184
147
161
135
119
649
771
862
641
593
924
972
830
819
672
549
436
384
313
299
286
227
275
238
203
163
178
149
107
100
101
104
83
132
111
1999 1445 1482 1516 1289 1428 1340 1447 1104 789
722 1110 2649 1327 1493 1127 1124 1182 1009 996
817
668
531
467
381
363
348
276
334
289
248
198
216
182
161
135
122
2000 1321 1355 1386 1178 1305 1224 1323 1009 722
858
958
713
910
747
611
485
427
348
332
318
253
306
264
226
181
198
166
147
123
111
91
622
739
826
615
569
886
932
795
785
644
527
418
368
300
286
274
218
264
228
195
156
171
143
127
106
96
79
2002
859
882
902
861
888
86
766
849
797
656
470
558
624
464
429
670
669
703
600
592
486
397
316
278
226
216
207
164
199
172
147
118
129
108
95
80
72
59
65
75
993
549
652
729
542
502
782
822
702
693
568
465
369
325
265
253
242
192
233
201
172
138
150
126
112
94
85
70
76
88
597
709
793
590
546
850
894
763
753
618
505
402
353
288
275
263
209
253
219
187
150
164
137
121
102
92
76
83
96
127
109
710
844
943
702
649
896
735
601
478
420
342
327
313
249
301
260
223
178
195
163
144
121
109
90
98
114
151
129
119
117
2006 1448 1486 1520 1292 1432 1343 1450 1106 791
723 1112 2656 1331 1497 1129 1127 1185 1012 998
819
670
532
468
382
364
349
277
335
290
248
198
217
182
161
135
122
100
110
127
169
144
133
111
2007 1463 1501 1536 1305 1446 1356 1465 1118 799
731 1124 2683 1344 1512 1141 1139 1197 1022 1009 827
677
538
473
385
368
352
280
339
293
251
200
219
184
163
136
123
101
111
129
170
146
134
113
101
556
661
739
550
508
711
702
576
471
374
329
268
256
245
195
236
204
174
139
152
128
113
95
86
70
77
89
118
101
93
78
70
700
832
930
692
640
884
725
593
471
414
338
322
309
245
297
257
220
176
192
161
142
119
108
89
97
113
149
128
117
99
89
792
833
70
88
126
2010 1397 1433 1466 1246 1381 1295 1399 1067 763
963
790
646
513
452
368
351
336
267
323
280
239
191
209
176
155
130
118
97
106
123
163
139
128
107
96
95
137
109
836
643
887
728
595
473
416
339
324
310
246
298
258
221
176
193
162
143
120
108
89
97
113
150
128
118
99
89
88
126
100
92
965
792
647
514
453
369
352
337
268
324
280
240
192
210
176
156
130
118
97
106
123
163
139
128
108
97
96
138
109
100
109
2013 1643 1686 1724 1465 1624 1523 1645 1255 898 1067 1192 887
821 1262 3013 1510 1698 1281 1279 1344 1148 1133 929
760
604
531
433
413
396
314
380
329
282
225
246
206
183
153
138
114
124
144
191
164
150
126
113
112
161
128
118
128
117
2014 1636 1678 1717 1459 1617 1517 1638 1250 894 1062 1187 883
817 1256 2999 1503 1690 1275 1273 1338 1143 1128 925
757
601
529
431
411
394
313
379
327
280
224
245
206
182
152
138
113
124
144
190
163
150
126
113
112
161
128
117
127
117
703
2012 1400 1436 1469 1249 1384 1298 1402 1069 765
24 February 2015
934
695
100
188
3.5
21.5
1963
2.9
11.2
1964
0.4
1965
0.3
1966
0.3
0.5
1967
0.3
0.3
1.8
0.1
0.1
5.9
1972 (0.9) 0.2 (1.7) (2.1) (1.5) (1.7) (2.0) (2.4) (1.0) 0.0
16.8
2.1 (10.7)
1973 (2.3) (1.4) (3.3) (3.7) (3.4) (3.8) (4.4) (5.1) (4.6) (4.7) (4.9) (14.2)(17.6)
1974 (4.4) (3.8) (5.7) (6.4) (6.3) (7.0) (7.9) (9.0) (9.1) (10.1)(11.6) (19.4)(23.4)(28.8)
1975 (3.6) (2.9) (4.6) (5.1) (5.0) (5.4) (6.1) (6.8) (6.7) (7.1) (7.7) (13.0)(13.7)(11.7) 9.5
1976 (3.4) (2.8) (4.4) (4.8) (4.7) (5.1) (5.6) (6.2) (6.0) (6.3) (6.6) (10.7)(10.7) (8.3)
4.1 (1.1)
1977 (1.8) (1.1) (2.4) (2.7) (2.4) (2.6) (2.9) (3.2) (2.6) (2.4) (2.2) (5.1) (3.9) (0.1) 11.8 13.0 29.1
1978 (2.2) (1.6) (2.9) (3.2) (2.9) (3.1) (3.4) (3.8) (3.3) (3.2) (3.1) (5.7) (4.8) (2.1)
6.1
5.0
8.1 (9.4)
1979 (2.7) (2.2) (3.4) (3.7) (3.5) (3.7) (4.1) (4.4) (4.1) (4.1) (4.1) (6.4) (5.8) (3.6)
2.4
0.7
1980 (2.3) (1.8) (2.9) (3.2) (3.0) (3.2) (3.4) (3.7) (3.4) (3.3) (3.2) (5.2) (4.5) (2.5)
2.8
1.5
1981 (2.7) (2.2) (3.3) (3.6) (3.4) (3.6) (3.8) (4.1) (3.8) (3.8) (3.8) (5.6) (5.0) (3.3)
1982 (0.9) (0.4) (1.4) (1.5) (1.2) (1.3) (1.4) (1.5) (1.0) (0.8) (0.5) (1.9) (1.0) 1.0
5.5
5.0
6.0
1.9
5.0
1983 (0.5) 0.1 (0.8) (1.0) (0.7) (0.7) (0.8) (0.8) (0.3) (0.0) 0.3 (1.0) (0.1) 1.9
6.0
5.6
6.6
3.2
6.0
1984 (0.4) 0.2 (0.7) (0.8) (0.5) (0.5) (0.6) (0.6) (0.2) 0.1
1.9
5.6
5.2
6.0
3.1
5.3
9.0
10.0 17.3
6.0
2.1
1985 (0.2) 0.4 (0.5) (0.6) (0.3) (0.3) (0.3) (0.3) 0.1
0.4
2.2
5.6
5.2
5.9
3.3
5.3
8.3
9.0
14.1
5.7
3.6
5.0
1986
0.1
0.1
0.0
0.0
0.5
0.8
1.1
0.1
0.9
2.5
5.7
5.3
6.0
3.7
5.5
8.1
8.7
12.6
6.0
4.7
6.0
7.0
1987
0.5
1.0
0.3
0.2
0.6
0.6
0.6
0.6
1.1
1.4
1.7
0.8
1.6
3.2
6.2
5.9
6.6
4.5
6.2
8.6
9.1
12.5
7.2
6.5
8.0
9.5
1988
0.6
1.1
0.4
0.3
0.6
0.7
0.7
0.7
1.1
1.4
1.7
0.9
1.7
3.1
5.9
5.6
6.2
4.3
5.8
7.9
8.3
11.0
6.4
5.7
6.6
7.1
7.2
2.4
1989
0.5
1.0
0.3
0.2
0.5
0.6
0.6
0.6
1.0
1.3
1.6
0.8
1.5
2.8
5.4
5.1
5.6
3.8
5.1
6.9
7.1
9.4
5.2
4.4
4.9
4.8
4.2
0.4 (1.7)
1990
0.4
0.8
0.2
0.1
0.4
0.4
0.4
0.4
0.8
1.0
1.3
0.6
1.2
2.5
4.8
4.5
4.9
3.3
4.4
5.9
6.0
7.9
4.1
3.3
3.4
3.1
1991
0.8
1.2
0.6
0.6
0.8
0.9
0.9
0.9
1.3
1.6
1.9
1.2
1.8
3.1
5.3
5.1
5.5
4.0
5.1
6.6
6.7
8.4
5.1
4.5
4.9
4.8
4.4
2.6
2.6
4.8
13.8
1992
1.2
1.7
1.1
1.0
1.3
1.4
1.4
1.5
1.9
2.2
2.5
1.8
2.5
3.7
5.9
5.6
6.1
4.7
5.8
7.2
7.4
9.1
6.1
5.7
6.1
6.3
6.2
5.0
5.7
8.3
14.6 15.4
1993
1.9
2.4
1.8
1.8
2.1
2.2
2.3
2.3
2.8
3.1
3.4
2.8
3.5
4.7
6.8
6.7
7.2
5.9
7.1
8.5
8.8
10.4
7.8
7.6
8.2
8.6
8.9
8.3
9.5
1994
1.4
1.8
1.3
1.3
1.5
1.6
1.6
1.7
2.1
2.3
2.6
2.0
2.7
3.7
5.7
5.5
5.9
4.7
5.6
6.8
7.0
8.3
5.8
5.4
5.8
5.9
5.7
4.8
5.2
6.7
9.4
7.9
4.4 (13.8)
1995
1.8
2.2
1.7
1.7
2.0
2.0
2.1
2.1
2.5
2.8
3.1
2.6
3.2
4.2
6.1
6.0
6.4
5.2
6.2
7.4
7.5
8.8
6.5
6.2
6.6
6.8
6.7
6.1
6.6
8.1
10.5
9.7
1996
1.9
2.3
1.8
1.8
2.1
2.1
2.2
2.2
2.6
2.9
3.2
2.7
3.3
4.3
6.1
5.9
6.3
5.2
6.1
7.2
7.4
8.6
6.4
6.1
6.5
6.6
6.6
6.0
6.4
7.6
9.6
8.8
7.2
1.5
10.1
1997
2.2
2.6
2.1
2.1
2.4
2.5
2.6
2.7
3.0
3.3
3.6
3.1
3.7
4.7
6.5
6.3
6.7
5.7
6.6
7.7
7.8
9.0
7.0
6.8
7.1
7.3
7.3
6.9
7.4
8.6
10.4
9.8
8.8
4.7
1998
2.7
3.1
2.6
2.7
2.9
3.0
3.1
3.2
3.6
3.9
4.2
3.7
4.3
5.3
7.1
7.0
7.4
6.4
7.3
8.3
8.5
9.7
7.8
7.7
8.1
8.4
8.5
8.1
8.7
7.9
1999
2.5
2.9
2.4
2.4
2.7
2.8
2.9
2.9
3.3
3.6
3.8
3.4
4.0
4.9
6.6
6.4
6.8
5.9
6.6
7.6
7.8
8.8
7.0
6.8
7.2
7.3
7.4
7.0
7.4
8.3
9.7
9.2
8.4
5.6
10.0
8.7
9.9
7.4 (5.2)
2000
2.6
3.0
2.5
2.5
2.8
2.9
3.0
3.0
3.4
3.7
3.9
3.5
4.1
5.0
6.5
6.4
6.7
5.9
6.6
7.6
7.7
8.6
7.0
6.8
7.1
7.2
7.3
6.9
7.3
8.1
9.4
8.9
8.1
5.7
9.3
8.2
9.0
7.0
0.3
6.1
2001
2.5
2.9
2.5
2.5
2.7
2.8
2.9
3.0
3.3
3.6
3.8
3.4
3.9
4.8
6.3
6.2
6.5
5.6
6.4
7.2
7.3
8.2
6.6
6.5
6.7
6.8
6.8
6.4
6.8
7.5
8.5
8.0
7.2
5.0
8.1
6.9
7.3
5.3
0.4
3.3
0.6
2002
2.6
3.0
2.6
2.6
2.8
2.9
3.0
3.1
3.4
3.7
3.9
3.5
4.0
4.9
6.3
6.2
6.5
5.7
6.4
7.2
7.3
8.2
6.6
6.5
6.7
6.8
6.8
6.5
6.7
7.4
8.4
7.9
7.2
5.2
7.9
6.9
7.2
5.6
1.9
4.4
3.6
2003
2.5
2.9
2.5
2.5
2.7
2.8
2.9
3.0
3.3
3.5
3.7
3.4
3.8
4.7
6.1
5.9
6.2
5.4
6.1
6.8
6.9
7.7
6.3
6.1
6.3
6.3
6.3
6.0
6.2
6.8
7.6
7.1
6.4
4.6
6.8
5.8
5.9
4.4
1.3
3.0
2.0
2.7 (1.2)
2004
2.5
2.9
2.5
2.5
2.8
2.8
2.9
3.0
3.3
3.5
3.7
3.4
3.8
4.6
6.0
5.9
6.1
5.3
6.0
6.7
6.8
7.5
6.1
5.9
6.1
6.2
6.2
5.8
6.0
6.6
7.3
6.8
6.1
4.5
6.5
5.6
5.6
4.3
1.7
3.1
2.4
3.0
1.2
3.6
2005
2.6
3.0
2.6
2.6
2.8
2.9
3.0
3.0
3.4
3.6
3.8
3.4
3.9
4.7
6.0
5.9
6.1
5.4
6.0
6.7
6.7
7.5
6.1
6.0
6.1
6.2
6.2
5.8
6.0
6.5
7.2
6.8
6.1
4.6
6.5
5.6
5.7
4.5
2.3
3.6
3.1
3.7
2.8
4.8
2006
2.5
2.8
2.4
2.4
2.7
2.7
2.8
2.9
3.1
3.4
3.6
3.2
3.7
4.4
5.6
5.5
5.7
5.0
5.6
6.3
6.3
7.0
5.7
5.5
5.6
5.7
5.6
5.3
5.4
5.9
6.5
6.0
5.4
3.9
5.5
4.7
4.6
3.5
1.4
2.4
1.8
2.1
0.9
1.7
0.7 (4.4)
2007
2.4
2.8
2.4
2.4
2.6
2.7
2.7
2.8
3.1
3.3
3.5
3.2
3.6
4.3
5.5
5.4
5.6
4.9
5.4
6.1
6.1
6.7
5.5
5.3
5.4
5.5
5.4
5.1
5.2
5.6
6.1
5.7
5.1
3.7
5.2
4.4
4.3
3.3
1.4
2.3
1.7
1.9
1.0
1.5
2008
2.6
3.0
2.6
2.6
2.8
2.9
2.9
3.0
3.3
3.5
3.7
3.4
3.8
4.5
5.7
5.6
5.8
5.1
5.6
6.3
6.3
6.9
5.7
5.5
5.7
5.7
5.7
5.4
5.5
5.9
6.5
6.0
5.5
4.2
5.6
4.9
4.9
4.0
2.4
3.3
2.9
3.3
2.7
3.5
3.5
2.6
6.4
11.8
2009
2.5
2.8
2.5
2.5
2.7
2.7
2.8
2.9
3.1
3.3
3.5
3.2
3.6
4.3
5.4
5.3
5.5
4.8
5.3
5.9
6.0
6.5
5.4
5.2
5.3
5.3
5.3
5.0
5.1
5.4
5.9
5.5
4.9
3.7
5.0
4.3
4.3
3.4
1.9
2.6
2.2
2.4
1.8
2.3
2.1
1.1
3.0
4.0 (3.3)
2010
2.5
2.9
2.5
2.5
2.7
2.8
2.8
2.9
3.2
3.4
3.6
3.2
3.6
4.3
5.4
5.3
5.5
4.8
5.3
5.9
5.9
6.5
5.3
5.2
5.3
5.3
5.2
4.9
5.1
5.4
5.8
5.4
4.9
3.8
5.0
4.3
4.3
3.5
2.1
2.8
2.4
2.6
2.2
2.6
2.5
1.8
3.4
4.1
0.5
4.4
2011
2.8
3.1
2.8
2.8
3.0
3.1
3.1
3.2
3.4
3.6
3.8
3.5
3.9
4.6
5.7
5.5
5.7
5.1
5.6
6.2
6.2
6.8
5.7
5.5
5.7
5.7
5.6
5.4
5.5
5.8
6.3
5.9
5.5
4.4
5.6
5.0
5.0
4.3
3.1
3.8
3.6
3.9
3.6
4.2
4.3
4.0
5.8
6.9
5.4
10.0 15.8
2012
2.8
3.1
2.7
2.7
3.0
3.0
3.1
3.1
3.4
3.6
3.8
3.5
3.9
4.5
5.5
5.4
5.6
5.0
5.5
6.0
6.1
6.6
5.5
5.4
5.5
5.5
5.5
5.2
5.3
5.6
6.1
5.7
5.3
4.3
5.4
4.8
4.8
4.1
3.0
3.6
3.4
3.7
3.4
3.9
4.0
3.7
5.1
5.9
4.4
7.1
8.5
2013
2.5
2.8
2.5
2.5
2.7
2.7
2.8
2.8
3.1
3.3
3.4
3.1
3.5
4.1
5.1
5.0
5.2
4.6
5.0
5.5
5.6
6.0
5.0
4.9
4.9
4.9
4.9
4.6
4.7
5.0
5.3
5.0
4.5
3.5
4.5
3.9
3.9
3.2
2.1
2.6
2.4
2.5
2.1
2.5
2.4
1.9
2.8
3.1
1.5
2.7
2014
2.7
3.0
2.7
2.7
2.9
3.0
3.1
3.1
3.4
3.5
3.7
3.4
3.8
4.4
5.4
5.3
5.5
4.9
5.3
5.8
5.9
6.3
5.4
5.2
5.3
5.3
5.3
5.0
5.1
5.4
5.8
5.4
5.0
4.1
5.1
4.6
4.5
3.9
2.9
3.5
3.3
3.5
3.3
3.7
3.7
3.4
4.4
4.9
3.8
5.3
5.5
24 February 2015
12.1
5.1
6.7
6.0
1.6
2.3
2.6
16.4
189
1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
1961
88
1962
107
1963
109
124
102
1964
102
115
95
93
1965
101
115
95
93
100
1966
102
116
95
94
100
101
1967
102
116
95
94
100
101
100
1968
94
107
88
86
93
93
92
92
1969
90
102
84
83
89
89
88
88
121
96
1970
87
98
81
79
85
85
85
85
92
96
1971
101
115
94
93
99
100
99
99
107
112
117
1972
90
102
84
83
89
89
88
88
96
100
104
89
1973
74
84
69
68
73
73
73
73
79
83
86
74
82
1974
53
60
49
49
52
52
52
52
56
59
61
52
59
71
1975
58
66
54
53
57
57
57
57
62
64
67
57
64
78
110
1976
57
65
54
53
56
56
56
56
61
64
66
57
63
77
108
99
1977
74
84
69
68
73
73
73
72
79
82
86
73
82
100
140
128
129
1978
67
76
63
61
66
66
66
66
71
74
77
66
74
90
127
116
117
91
1979
59
68
56
55
59
59
58
58
63
66
69
59
66
80
112
103
104
80
89
1980
62
71
58
57
61
62
61
61
66
69
72
62
69
84
118
108
109
84
93
1981
57
64
53
52
56
56
56
56
60
63
66
56
63
76
107
98
99
77
85
95
91
1982
81
92
76
75
80
80
80
80
87
90
94
81
90
110
154
141
142
110
122
137
130
144
1983
90
102
84
82
88
88
88
88
95
99
104
89
99
121
169
155
156
121
134
151
143
158
110
1984
92
104
86
84
90
90
90
90
97
102
106
91
101
123
173
158
160
124
137
154
147
161
112
1985
96
109
90
88
95
95
94
94
102
107
111
95
107
129
182
166
168
130
143
162
154
169
118
107
105
1986
103
117
96
94
101
101
101
101
109
114
119
102
114
138
194
177
179
139
153
173
165
181
126
115
112
107
1987
115
131
108
106
113
114
113
113
123
128
133
114
128
155
218
199
201
156
172
194
185
203
142
129
126
120
112
1988
118
134
110
108
116
116
116
116
126
131
137
117
131
159
223
204
206
160
176
199
189
208
145
132
129
123
115
102
1989
116
132
109
107
114
114
114
114
123
129
134
115
129
156
220
200
203
157
173
195
186
205
143
130
127
121
113
101
98
1990
112
127
105
103
110
111
110
110
119
125
130
111
124
151
212
194
196
152
167
189
180
198
138
125
123
117
109
97
95
97
1991
128
145
119
117
126
126
125
125
136
142
148
126
142
172
241
220
223
173
191
215
204
225
157
142
139
133
124
111
108
110
114
1992
147
167
138
135
145
145
144
144
157
164
170
146
163
198
279
254
257
199
220
248
236
260
181
164
161
153
143
128
125
127
131
115
1993
186
211
174
171
183
184
183
182
198
207
215
184
206
251
352
321
325
252
278
313
298
328
229
208
203
194
181
162
158
160
166
146
126
1994
161
182
150
147
158
158
157
157
171
178
186
159
178
216
303
277
280
217
240
270
257
283
197
179
175
167
156
139
136
138
143
126
109
1995
185
210
173
170
182
182
181
181
197
205
214
183
205
249
350
319
323
250
276
311
296
326
227
206
202
193
180
160
157
159
165
145
126
99
115
1996
195
221
182
179
191
192
191
191
207
216
225
193
216
262
368
336
340
263
290
327
311
343
239
217
213
202
189
169
165
167
173
152
132
104
121
105
1997
224
255
210
206
221
221
220
220
238
249
259
222
248
302
424
387
391
303
335
377
359
395
275
250
245
233
218
194
190
193
200
176
152
120
140
121
115
1998
273
310
255
250
268
269
267
267
290
303
315
270
302
367
516
471
476
369
407
459
437
481
335
304
298
284
265
237
231
235
243
214
185
146
170
147
140
122
1999
259
294
242
237
254
255
254
253
275
287
299
256
287
348
489
446
451
350
386
435
414
456
317
288
282
269
251
224
219
223
230
202
175
139
161
140
133
115
95
2000
274
311
256
252
270
270
269
269
292
305
317
272
304
369
519
473
479
371
410
461
439
483
337
306
300
285
267
238
232
236
245
215
186
147
171
148
141
122
101
106
2001
276
314
258
253
272
272
271
270
293
306
319
273
306
371
522
476
482
373
412
464
442
487
339
308
302
287
269
239
234
238
246
216
187
148
172
149
142
123
101
107
101
2002
295
334
275
270
290
290
289
289
313
327
341
292
326
396
557
508
514
398
440
495
471
519
362
329
322
306
287
255
249
254
263
231
200
158
183
159
151
131
108
114
107
2003
291
330
272
267
286
287
285
285
309
323
336
288
322
391
550
502
508
393
434
489
466
513
357
325
318
303
283
252
246
251
259
228
197
156
181
157
150
130
107
113
106
105
99
2004
301
342
282
277
297
297
296
295
320
335
349
298
334
405
570
520
526
407
450
507
482
531
370
336
329
314
293
261
255
260
269
236
205
162
188
163
155
134
111
117
110
109
102
104
2005
320
363
299
293
315
315
313
313
340
355
370
316
354
430
604
552
558
432
477
537
512
563
392
357
349
333
311
277
271
275
285
250
217
172
199
173
164
143
117
124
117
116
109
110
106
2006
306
347
286
281
301
301
300
299
325
339
353
303
339
411
578
527
534
413
456
514
489
539
375
341
334
318
297
265
259
263
272
239
207
164
190
165
157
136
112
118
111
111
104
105
101
2007
309
351
289
284
304
305
303
303
329
343
357
306
343
416
584
534
540
418
462
520
495
545
380
345
338
322
301
268
262
266
276
242
210
166
193
167
159
138
113
120
113
112
105
106
103
97
101
2008
346
393
323
317
340
341
339
339
367
384
400
342
383
465
653
597
604
467
516
581
553
609
424
386
378
360
336
300
293
298
308
271
235
186
215
187
178
154
127
134
126
125
117
119
115
108
113
2009
335
380
313
307
329
330
328
328
355
371
387
331
371
450
632
577
584
452
499
562
535
589
411
373
365
348
325
290
283
288
298
262
227
180
208
181
172
149
123
129
122
121
114
115
111
105
109
108
97
2010
349
397
326
321
344
344
342
342
371
388
404
346
387
470
660
603
610
472
521
587
559
615
429
390
382
363
340
303
296
301
311
274
237
187
218
189
180
156
128
135
127
126
119
120
116
109
114
113
101
104
2011
405
459
378
371
398
399
397
396
430
449
468
400
448
544
764
698
706
547
604
680
647
713
496
451
442
421
393
351
342
348
360
317
274
217
252
219
208
180
148
156
147
146
137
139
134
127
132
131
117
121
116
2012
411
467
384
377
405
405
403
403
437
456
475
407
456
553
777
709
718
556
614
691
658
724
505
459
449
428
400
357
348
354
366
322
279
221
256
222
211
183
151
159
150
149
140
141
136
129
134
133
119
123
118
2013
372
422
347
341
366
366
364
364
395
412
430
368
412
500
702
641
649
502
555
625
595
655
456
415
406
387
361
322
315
320
331
291
252
199
231
201
191
166
136
144
135
135
126
128
123
116
122
120
107
111
106
92
90
2014
433
491
404
397
426
426
424
424
460
480
500
428
479
582
818
747
755
585
646
727
692
762
531
483
473
450
421
375
366
373
386
339
294
232
270
234
222
193
159
167
158
157
147
149
144
135
142
140
125
129
124
107
105
24 February 2015
105
102
86
107
96
112
102
116
190
0.7
1962
1.3
1.8
1963
1.5
1.8
1.9
1964
1.0
1.1
0.8 (0.4)
1965
1.1
1.2
1.1
0.7
1.7
1966
1.3
1.5
1.4
1.2
2.0
2.4
1967
1.6
1.8
1.8
1.8
2.5
2.9
3.4
1968
1.6
1.7
1.7
1.7
2.2
2.4
2.4
1.4
1969
1.8
1.9
1.9
1.9
2.4
2.6
2.6
2.3
3.1
1970
1.6
1.6
1.6
1.6
1.9
2.0
1.9
1.4
1.3 (0.4)
1971
1.2
1.2
1.1
1.1
1.3
1.2
0.9
0.4
1972
0.9
0.9
0.8
0.7
0.8
0.7
1973
0.7
0.7
0.6
0.5
0.6
0.4
1974
0.3
0.2
0.1 (0.1) (0.0) (0.2) (0.6) (1.1) (1.5) (2.4) (2.9) (3.0) (3.5) (5.5)
1975 (0.6) (0.7) (0.8) (1.1) (1.1) (1.4) (1.8) (2.4) (3.0) (4.0) (4.7) (5.2) (6.2) (8.5) (11.3)
1976 (0.7) (0.8) (1.0) (1.2) (1.3) (1.6) (2.0) (2.5) (3.0) (3.9) (4.4) (4.8) (5.4) (6.8) (7.4) (3.2)
1977 (0.8) (0.9) (1.1) (1.3) (1.4) (1.6) (2.0) (2.5) (2.9) (3.7) (4.1) (4.4) (4.8) (5.7) (5.7) (2.8) (2.4)
1978 (0.8) (0.9) (1.1) (1.3) (1.3) (1.5) (1.9) (2.3) (2.7) (3.3) (3.7) (3.8) (4.1) (4.6) (4.4) (2.0) (1.4) (0.3)
1979 (0.9) (1.0) (1.2) (1.4) (1.4) (1.7) (2.0) (2.4) (2.7) (3.3) (3.6) (3.7) (4.0) (4.4) (4.2) (2.3) (2.0) (1.8) (3.2)
1980 (0.8) (0.9) (1.0) (1.2) (1.2) (1.4) (1.7) (2.1) (2.4) (2.9) (3.1) (3.1) (3.3) (3.5) (3.2) (1.5) (1.1) (0.6) (0.8) 1.8
1981 (0.7) (0.8) (0.9) (1.0) (1.1) (1.3) (1.5) (1.8) (2.1) (2.5) (2.7) (2.7) (2.8) (2.9) (2.5) (1.0) (0.5) (0.1) (0.0) 1.7
1982 (0.4) (0.4) (0.5) (0.7) (0.7) (0.8) (1.0) (1.3) (1.5) (1.8) (1.9) (1.9) (1.9) (1.9) (1.4) 0.1
0.6
1.2
1.6
3.3
1.5
4.0 6.6
1983 (0.2) (0.2) (0.3) (0.4) (0.4) (0.5) (0.7) (0.9) (1.1) (1.4) (1.5) (1.4) (1.3) (1.3) (0.8) 0.6
1.2
1.8
2.2
3.6
4.2
5.6
4.6
1984
0.0
0.0 (0.1) (0.2) (0.1) (0.2) (0.4) (0.6) (0.7) (1.0) (1.0) (0.9) (0.8) (0.7) (0.3) 1.1
1.6
2.2
2.6
3.8
4.4
5.3
4.7
1985
0.3
0.2
0.2
0.1
0.1
0.3
1.5
2.1
2.6
3.1
4.2
4.7
5.4
5.1
5.3
5.8
1986
0.5
0.5
0.5
0.4
0.4
0.4
0.3
0.1
0.0
0.2
0.3
0.8
2.0
2.6
3.1
3.6
4.6
5.0
5.7
5.5
5.9
6.4
7.0
1987
0.7
0.7
0.7
0.6
0.7
0.6
0.5
0.4
0.3
0.2
0.2
0.4
0.6
0.7
1.2
2.3
2.8
3.4
3.8
4.7
5.1
5.7
5.6
5.8
6.2
6.3
5.7
1988
0.8
0.8
0.8
0.7
0.8
0.7
0.7
0.5
0.5
0.4
0.4
0.6
0.8
0.9
1.4
2.4
2.9
3.4
3.8
4.6
5.0
5.5
5.3
5.4
5.6
5.5
4.8
4.0
1989
1.0
1.0
1.0
1.0
1.0
1.0
0.9
0.8
0.8
0.7
0.7
0.9
1.1
1.2
1.7
2.7
3.2
3.7
4.0
4.8
5.1
5.6
5.4
5.6
5.8
5.7
5.3
5.2
6.4
1990
1.2
1.2
1.2
1.1
1.2
1.2
1.1
1.0
1.0
0.9
1.0
1.2
1.4
1.5
2.0
2.9
3.4
3.8
4.2
4.9
5.2
5.6
5.5
5.6
5.8
5.8
5.5
5.4
6.2
6.0
1991
1.3
1.4
1.4
1.3
1.4
1.4
1.3
1.3
1.3
1.2
1.2
1.4
1.6
1.8
2.3
3.2
3.6
4.1
4.4
5.1
5.4
5.8
5.7
5.8
5.9
6.0
5.8
5.8
6.4
6.4
6.8
1992
1.5
1.5
1.5
1.5
1.6
1.6
1.5
1.5
1.5
1.4
1.5
1.7
1.9
2.1
2.5
3.4
3.8
4.2
4.6
5.2
5.5
5.8
5.8
5.9
6.0
6.1
5.9
6.0
6.5
6.5
6.8
6.7
1993
1.6
1.6
1.6
1.6
1.7
1.7
1.6
1.6
1.6
1.5
1.6
1.8
2.0
2.1
2.6
3.4
3.8
4.2
4.5
5.1
5.3
5.7
5.6
5.7
5.8
5.8
5.6
5.6
5.9
5.8
5.8
5.3
3.9
1994
1.6
1.6
1.6
1.6
1.7
1.7
1.7
1.6
1.6
1.5
1.6
1.8
2.0
2.2
2.6
3.3
3.7
4.1
4.4
4.9
5.1
5.4
5.3
5.4
5.5
5.4
5.2
5.2
5.4
5.2
5.0
4.3
3.2
2.4
1995
1.7
1.7
1.7
1.7
1.7
1.7
1.7
1.7
1.7
1.6
1.7
1.9
2.1
2.2
2.6
3.4
3.7
4.1
4.3
4.8
5.0
5.3
5.2
5.2
5.3
5.2
5.0
4.9
5.1
4.9
4.6
4.1
3.2
2.9
3.4
1996
1.7
1.7
1.7
1.7
1.8
1.8
1.8
1.7
1.7
1.7
1.8
2.0
2.1
2.3
2.6
3.4
3.7
4.0
4.3
4.7
4.9
5.2
5.1
5.1
5.1
5.1
4.9
4.8
4.9
4.7
4.5
4.0
3.3
3.2
3.5
3.6
1997
1.7
1.8
1.8
1.8
1.8
1.8
1.8
1.8
1.8
1.7
1.8
2.0
2.2
2.3
2.7
3.4
3.7
4.0
4.2
4.7
4.8
5.0
4.9
5.0
5.0
4.9
4.7
4.6
4.7
4.5
4.3
3.9
3.3
3.2
3.4
3.4
3.1
1998
1.8
1.9
1.9
1.9
1.9
1.9
1.9
1.9
1.9
1.9
1.9
2.1
2.3
2.4
2.8
3.4
3.7
4.0
4.3
4.7
4.8
5.0
4.9
5.0
5.0
4.9
4.7
4.7
4.7
4.5
4.4
4.0
3.6
3.5
3.8
3.9
4.1
5.0
1999
1.9
1.9
1.9
1.9
2.0
2.0
2.0
1.9
2.0
1.9
2.0
2.2
2.3
2.5
2.8
3.4
3.7
4.0
4.2
4.6
4.8
5.0
4.9
4.9
4.9
4.8
4.7
4.6
4.6
4.5
4.3
4.0
3.6
3.5
3.8
3.9
3.9
4.4
3.7
2000
1.9
1.9
1.9
1.9
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.2
2.4
2.5
2.8
3.4
3.7
4.0
4.2
4.6
4.7
4.9
4.8
4.8
4.8
4.7
4.6
4.5
4.5
4.3
4.2
3.9
3.5
3.5
3.7
3.7
3.8
4.0
3.4
3.2
2001
2.0
2.0
2.0
2.0
2.1
2.1
2.1
2.1
2.1
2.0
2.1
2.3
2.4
2.6
2.9
3.5
3.8
4.0
4.2
4.6
4.7
4.9
4.8
4.8
4.8
4.7
4.6
4.5
4.5
4.4
4.2
4.0
3.7
3.7
3.8
3.9
4.0
4.2
3.9
4.0
4.8
2002
2.0
2.0
2.0
2.0
2.1
2.1
2.1
2.0
2.0
2.0
2.1
2.2
2.4
2.5
2.8
3.4
3.7
3.9
4.1
4.4
4.5
4.7
4.6
4.6
4.6
4.5
4.4
4.3
4.3
4.1
4.0
3.7
3.4
3.4
3.5
3.5
3.5
3.6
3.2
3.0
2.9
1.1
2003
1.9
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.1
2.2
2.3
2.5
2.8
3.3
3.6
3.8
4.0
4.3
4.4
4.5
4.4
4.4
4.4
4.3
4.1
4.1
4.1
3.9
3.7
3.5
3.2
3.1
3.2
3.2
3.1
3.1
2.7
2.5
2.3
1.0
0.9
2004
1.9
1.9
1.9
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.2
2.3
2.4
2.7
3.2
3.5
3.7
3.8
4.1
4.2
4.4
4.3
4.2
4.2
4.1
4.0
3.9
3.9
3.7
3.5
3.3
3.0
2.9
3.0
2.9
2.9
2.8
2.5
2.2
2.0
1.0
1.0
2005
1.9
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.2
2.3
2.4
2.7
3.2
3.4
3.7
3.8
4.1
4.2
4.3
4.2
4.2
4.1
4.1
3.9
3.8
3.8
3.6
3.5
3.3
3.0
2.9
3.0
2.9
2.8
2.8
2.5
2.3
2.1
1.5
1.6
1.9
2.7
2006
1.9
1.9
1.9
1.9
2.0
2.0
2.0
2.0
2.0
1.9
2.0
2.1
2.3
2.4
2.6
3.1
3.3
3.5
3.7
3.9
4.0
4.1
4.0
4.0
4.0
3.9
3.7
3.6
3.6
3.5
3.3
3.1
2.8
2.7
2.8
2.7
2.6
2.5
2.2
2.0
1.8
1.3
1.3
1.4
1.6
0.4
2007
1.9
1.9
1.9
1.9
2.0
2.0
2.0
1.9
2.0
1.9
2.0
2.1
2.2
2.4
2.6
3.1
3.3
3.5
3.6
3.9
3.9
4.0
3.9
3.9
3.9
3.8
3.6
3.5
3.5
3.4
3.2
3.0
2.7
2.7
2.7
2.6
2.5
2.5
2.2
2.0
1.8
1.3
1.4
1.5
1.6
1.1
1.8
2008
1.9
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.1
2.2
2.3
2.4
2.7
3.1
3.3
3.5
3.6
3.9
4.0
4.0
3.9
3.9
3.9
3.8
3.7
3.6
3.5
3.4
3.3
3.1
2.8
2.8
2.8
2.7
2.7
2.6
2.4
2.2
2.1
1.7
1.9
2.0
2.3
2.1
3.0
4.2
2009
1.9
1.9
1.9
1.9
1.9
2.0
1.9
1.9
1.9
1.9
2.0
2.1
2.2
2.3
2.5
3.0
3.2
3.3
3.5
3.7
3.8
3.8
3.7
3.7
3.7
3.6
3.4
3.3
3.3
3.1
3.0
2.8
2.6
2.5
2.5
2.4
2.3
2.3
2.0
1.8
1.7
1.3
1.3
1.4
1.5
1.2
1.4
1.2 (1.7)
2010
1.7
1.8
1.8
1.8
1.8
1.8
1.8
1.8
1.8
1.7
1.8
1.9
2.0
2.1
2.3
2.8
2.9
3.1
3.2
3.4
3.5
3.6
3.4
3.4
3.4
3.3
3.1
3.0
2.9
2.8
2.6
2.4
2.2
2.1
2.1
2.0
1.9
1.8
1.5
1.3
1.1
0.7
0.6
0.6
0.5
0.1
2011
1.6
1.6
1.6
1.6
1.7
1.7
1.7
1.6
1.6
1.6
1.7
1.8
1.9
1.9
2.2
2.6
2.7
2.9
3.0
3.2
3.2
3.3
3.2
3.1
3.1
3.0
2.8
2.7
2.6
2.5
2.3
2.1
1.8
1.7
1.7
1.6
1.4
1.3
1.0
0.8
0.6
0.2
0.1
2012
1.5
1.6
1.6
1.6
1.6
1.6
1.6
1.5
1.5
1.5
1.5
1.6
1.7
1.8
2.0
2.4
2.6
2.7
2.8
3.0
3.0
3.1
3.0
2.9
2.9
2.7
2.6
2.5
2.4
2.2
2.1
1.8
1.6
1.5
1.4
1.3
1.2
1.1
0.8
0.6
0.3 (0.1) (0.2) (0.3) (0.5) (0.9) (1.1) (1.7) (3.1) (3.6) (3.4) (2.7)
2013
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.4
1.4
1.4
1.5
1.6
1.6
1.7
1.9
2.3
2.4
2.6
2.7
2.8
2.9
2.9
2.8
2.7
2.7
2.6
2.4
2.3
2.2
2.0
1.9
1.7
1.4
1.3
1.2
1.1
1.0
0.8
0.6
0.3
0.1 (0.2) (0.4) (0.5) (0.7) (1.1) (1.3) (1.8) (3.0) (3.3) (3.0) (2.5) (2.3)
2014
1.4
1.4
1.4
1.4
1.5
1.5
1.4
1.4
1.4
1.4
1.4
1.5
1.6
1.6
1.8
2.2
2.3
2.5
2.6
2.7
2.8
2.8
2.7
2.6
2.5
2.4
2.3
2.1
2.1
1.9
1.7
1.5
1.3
1.2
1.1
1.0
0.9
0.7
0.5
0.2
0.0 (0.3) (0.4) (0.6) (0.7) (1.1) (1.3) (1.7) (2.7) (2.9) (2.6) (2.1) (1.7) (1.2)
24 February 2015
4.8
1.1
191
101
105 103
101 100
97
97
99
98
95
95
98
98
97
94
94
97
99
99
100
98
96
92
91
88
89
91
93
94
1975
92
91
90
88
88
87
85
82
81
78
79
81
83
84
89
1976
89
88
87
85
85
84
82
79
78
76
76
78
80
81
86
97
1977
87
86
85
83
83
82
80
77
76
74
74
76
78
79
84
94
98
1978
86
86
84
83
83
82
80
77
76
74
74
76
78
79
83
94
97
1979
84
83
82
80
80
79
77
75
74
71
72
74
75
76
81
91
94
96
97
1980
85
85
83
82
82
80
79
76
75
73
73
75
77
78
82
93
96
98
98
1981
86
86
84
83
83
82
80
77
76
74
74
76
78
79
83
94
97
100
100 103
102
1982
92
92
90
88
89
87
85
82
81
79
79
81
83
84
89
100
104 106
107 110
108 107
1983
96
96
94
92
93
91
89
86
85
82
83
85
87
88
93
105
108 111
111 115
113 111
105
99
97
97
95
93
90
89
86
87
89
91
92
98
110
114 116
117 121
119 117
110 105
99
95
94
91
92
94
96
98
103 116
120 123
124 128
126 124
116 111
106
102 101
98
98
110 124
129 132
132 137
134 132
124 119
113 107
117 132
136 139
140 144
142 140
131 125
121 137
141 145
145 150
148 145
136 130
129 145
150 154
155 160
157 155
145 139
111 106
137 154
159 163
164 169
166 164
154 147
117 113
106
146 165
170 174
175 181
178 175
164 157
125 120
113 107
156 176
182 186
187 193
190 187
175 167
134 128
121 114
107
162 183
189 193
194 200
197 194
182 174
139 133
125 118
111 104
166 187
193 198
199 205
202 199
186 178
142 137
129 121
171 193
200 205
205 212
209 205
193 184
147 141
133 125
103
178 200
207 212
213 220
216 213
200 191
152 146
138 130
107 104
183 207
214 219
220 227
223 220
206 197
157 151
142 134
111 107
103
192 217
224 230
231 238
234 231
216 207
165 159
149 141
116 112
108 105
200 225
233 238
239 247
243 239
224 214
171 165
155 146
120 116
112 109
206 232
240 246
247 255
250 247
231 221
177 170
160 151
124 120
116 112
107 103
216 243
251 258
258 267
262 258
242 232
185 178
167 158
130 126
121 118
218 246
254 261
261 270
265 261
245 234
187 180
169 160
132 127
123 119
101
220 248
257 263
264 273
268 264
247 237
189 182
171 161
133 128
124 120
102 101
223 251
259 266
267 276
271 267
250 239
191 184
173 163
134 130
125 121
103 102
101
229 258
267 273
274 283
278 274
257 246
196 189
177 167
138 133
129 125
106 105
104 103
230 259
268 274
275 284
279 275
258 247
197 189
178 168
139 134
129 125
106 105
104 103
100
234 264
272 279
280 289
284 280
263 251
200 193
181 171
141 136
132 128
108 107
106 105
102 102
244 275
284 291
292 302
296 292
274 262
209 201
189 178
147 142
137 133
113 112
111 109
107 106
104
240 270
279 286
287 297
291 287
269 257
205 197
186 175
144 140
135 131
111 110
109 108
105 104
102
98
230 259
268 274
275 284
279 275
258 247
197 189
178 168
139 134
129 125
106 105
104 103
100 100
98
94
96
220 248
257 263
264 273
268 264
248 237
189 182
171 161
133 128
124 120
102 101
100
99
96
96
94
90
92
96
214 242
250 256
257 265
261 257
241 230
184 177
166 157
129 125
121 117
99
98
97
96
94
93
92
88
90
93
97
210 236
244 250
251 259
255 251
235 225
180 173
162 153
126 122
118 114
97
96
95
94
92
91
90
86
87
91
95
98
207 233
241 247
248 256
252 248
233 222
177 171
160 151
125 121
117 113
96
95
94
93
91
90
89
85
86
90
94
97
24 February 2015
100
102
104
104
99
192
1.4
1962
2.4
3.4
1963
2.9
3.6
3.9
1964
2.4
2.7
2.4
0.9
1965
2.3
2.5
2.2
1.4
1.9
1966
2.4
2.6
2.4
1.9
2.5
3.0
1967
2.7
3.0
2.9
2.6
3.2
3.8
4.7
1968
2.6
2.7
2.6
2.4
2.8
3.1
3.1
1.5
1969
2.7
2.8
2.8
2.6
2.9
3.2
3.2
2.5
3.5
1970
2.5
2.6
2.5
2.3
2.5
2.6
2.5
1.8
2.0
1971
2.2
2.2
2.1
1.9
2.0
2.1
1.9
1.2
1972
2.0
2.1
2.0
1.7
1.8
1.8
1.6
1.0
0.9
1973
1.8
1.8
1.7
1.5
1.6
1.5
1.3
0.7
1974
1.2
1.1
1.0
0.7
0.7
0.6
1975
0.3
0.2 (0.0) (0.3) (0.4) (0.7) (1.1) (1.8) (2.2) (3.2) (3.9) (4.7) (6.3) (9.0) (11.1)
1976
0.0 (0.1) (0.3) (0.6) (0.7) (1.0) (1.4) (2.0) (2.4) (3.3) (3.9) (4.5) (5.7) (7.3) (7.6) (3.8)
0.6
0.1 (0.1)
0.5
1977 (0.0) (0.1) (0.4) (0.7) (0.8) (1.0) (1.4) (1.9) (2.3) (3.0) (3.5) (4.0) (4.8) (5.8) (5.5) (2.6) (1.3)
1978
0.0 (0.1) (0.3) (0.6) (0.7) (0.9) (1.2) (1.7) (2.0) (2.6) (3.0) (3.3) (3.9) (4.5) (3.9) (1.4) (0.2) 1.0
1979 (0.2) (0.3) (0.5) (0.8) (0.9) (1.1) (1.4) (1.9) (2.2) (2.8) (3.1) (3.4) (4.0) (4.5) (4.0) (2.1) (1.6) (1.7) (4.3)
1980 (0.2) (0.3) (0.5) (0.8) (0.9) (1.0) (1.3) (1.8) (2.0) (2.5) (2.8) (3.1) (3.5) (3.9) (3.4) (1.7) (1.2) (1.2) (2.2) (0.1)
1981 (0.2) (0.2) (0.4) (0.7) (0.8) (0.9) (1.2) (1.6) (1.8) (2.2) (2.5) (2.7) (3.0) (3.3) (2.8) (1.3) (0.8) (0.7) (1.2) 0.3
0.8
1982
0.1
0.1 (0.1) (0.3) (0.4) (0.5) (0.7) (1.1) (1.3) (1.6) (1.8) (1.9) (2.1) (2.3) (1.7) (0.3)
0.4
0.7
0.6
2.3
3.6
6.4
1983
0.3
0.2
0.1 (0.1) (0.1) (0.3) (0.4) (0.8) (0.9) (1.2) (1.3) (1.4) (1.6) (1.6) (1.1) 0.3
0.9
1.3
1.3
2.8
3.8
5.3
4.1
1984
0.5
0.5
0.3
0.2
0.1
0.0 (0.1) (0.4) (0.5) (0.8) (0.9) (0.9) (1.0) (1.0) (0.4) 0.8
1.4
1.8
2.0
3.2
4.1
5.2
4.6
5.2
1985
0.7
0.6
0.5
0.4
0.3
0.3
0.0
1.2
1.8
2.2
2.4
3.5
4.3
5.1
4.7
5.0
4.8
1986
0.9
0.9
0.8
0.6
0.6
0.6
0.4
0.2
0.6
1.7
2.3
2.7
2.9
3.9
4.6
5.4
5.2
5.5
5.7
6.6
1987
1.1
1.0
1.0
0.8
0.8
0.8
0.7
0.5
0.4
0.3
0.2
0.3
0.3
0.4
0.9
2.0
2.6
3.0
3.2
4.2
4.8
5.5
5.3
5.6
5.7
6.2
5.7
1988
1.1
1.1
1.0
0.9
0.9
0.8
0.7
0.5
0.5
0.3
0.3
0.4
0.4
0.4
1.0
2.0
2.5
2.8
3.0
3.9
4.4
4.9
4.6
4.7
4.6
4.5
3.5
1.4
1989
1.1
1.1
1.0
0.9
0.9
0.9
0.8
0.6
0.6
0.4
0.4
0.5
0.5
0.6
1.1
2.0
2.5
2.8
3.0
3.7
4.2
4.6
4.4
4.4
4.3
4.1
3.3
2.1
2.8
1990
1.2
1.2
1.1
1.0
1.0
1.0
0.9
0.7
0.7
0.5
0.5
0.6
0.6
0.7
1.2
2.1
2.5
2.8
2.9
3.6
4.0
4.4
4.1
4.1
4.0
3.8
3.1
2.2
2.6
2.5
1991
1.3
1.3
1.2
1.1
1.1
1.1
1.0
0.9
0.8
0.7
0.7
0.8
0.8
0.9
1.4
2.2
2.6
2.9
3.1
3.7
4.1
4.4
4.2
4.2
4.1
3.9
3.4
2.8
3.3
3.6
4.6
1992
1.5
1.5
1.4
1.3
1.3
1.3
1.2
1.1
1.1
1.0
1.0
1.1
1.1
1.2
1.7
2.5
2.9
3.2
3.3
4.0
4.3
4.6
4.4
4.5
4.4
4.3
4.0
3.6
4.2
4.6
5.7
6.8
1993
1.5
1.5
1.4
1.3
1.4
1.3
1.3
1.1
1.1
1.0
1.0
1.1
1.2
1.3
1.7
2.5
2.9
3.1
3.3
3.8
4.1
4.4
4.2
4.2
4.1
4.1
3.7
3.4
3.8
4.0
4.5
4.5
2.2
1994
1.5
1.5
1.4
1.3
1.3
1.3
1.3
1.1
1.1
1.0
1.0
1.1
1.1
1.2
1.7
2.4
2.7
3.0
3.1
3.6
3.9
4.1
3.9
3.9
3.8
3.7
3.3
3.0
3.3
3.4
3.6
3.2
1.5
0.8
1995
1.4
1.4
1.4
1.3
1.3
1.3
1.2
1.1
1.1
1.0
1.0
1.1
1.1
1.2
1.6
2.3
2.6
2.9
3.0
3.4
3.7
3.9
3.7
3.7
3.5
3.4
3.0
2.7
2.9
2.9
3.0
2.6
1.2
0.7
0.7
1996
1.4
1.4
1.3
1.3
1.3
1.3
1.2
1.1
1.1
1.0
1.0
1.1
1.1
1.2
1.5
2.2
2.5
2.7
2.8
3.2
3.4
3.6
3.4
3.4
3.2
3.1
2.7
2.4
2.5
2.5
2.5
2.1
0.9
0.5
0.4
1997
1.3
1.3
1.3
1.2
1.2
1.2
1.1
1.0
1.0
0.9
0.9
1.0
1.0
1.1
1.5
2.1
2.4
2.5
2.6
3.0
3.2
3.4
3.2
3.1
2.9
2.8
2.4
2.1
2.2
2.1
2.1
1.7
0.6
0.3
1998
1.4
1.4
1.4
1.3
1.3
1.3
1.2
1.1
1.1
1.0
1.0
1.1
1.1
1.2
1.6
2.2
2.4
2.6
2.7
3.1
3.3
3.4
3.2
3.2
3.0
2.9
2.6
2.3
2.4
2.4
2.3
2.0
1.2
1.0
1.1
1.2
1.8
1999
1.5
1.5
1.4
1.3
1.4
1.3
1.3
1.2
1.2
1.1
1.1
1.2
1.2
1.3
1.6
2.2
2.5
2.6
2.7
3.1
3.3
3.4
3.2
3.2
3.0
2.9
2.6
2.4
2.5
2.4
2.4
2.2
1.5
1.4
1.5
1.8
2.3
3.7
3.3
2000
1.5
1.5
1.4
1.4
1.4
1.4
1.3
1.2
1.2
1.2
1.2
1.2
1.3
1.3
1.7
2.2
2.5
2.6
2.7
3.1
3.2
3.4
3.2
3.1
3.0
2.9
2.6
2.4
2.5
2.4
2.4
2.2
1.6
1.6
1.7
1.9
2.3
3.3
2.9
2.5
2001
1.6
1.6
1.5
1.4
1.5
1.4
1.4
1.3
1.3
1.2
1.3
1.3
1.4
1.4
1.8
2.3
2.5
2.7
2.8
3.1
3.3
3.4
3.2
3.2
3.1
3.0
2.7
2.5
2.6
2.6
2.6
2.4
1.9
1.9
2.0
2.2
2.7
3.5
3.3
3.2
4.0
2002
1.5
1.5
1.5
1.4
1.4
1.4
1.4
1.3
1.3
1.2
1.2
1.3
1.3
1.4
1.7
2.2
2.5
2.6
2.7
3.0
3.1
3.2
3.1
3.0
2.9
2.8
2.6
2.4
2.4
2.4
2.4
2.2
1.8
1.7
1.8
2.0
2.3
2.9
2.5
2.3
2.2
0.4
2003
1.5
1.5
1.5
1.4
1.4
1.4
1.4
1.3
1.3
1.2
1.2
1.3
1.3
1.4
1.7
2.2
2.4
2.5
2.6
2.9
3.0
3.1
3.0
2.9
2.8
2.7
2.5
2.2
2.3
2.3
2.3
2.1
1.6
1.6
1.7
1.8
2.0
2.5
2.1
1.8
1.6
0.5
0.5
2004
1.5
1.5
1.4
1.4
1.4
1.4
1.3
1.2
1.2
1.2
1.2
1.3
1.3
1.3
1.6
2.1
2.3
2.5
2.5
2.8
2.9
3.0
2.9
2.8
2.7
2.6
2.4
2.2
2.2
2.2
2.1
2.0
1.6
1.5
1.6
1.7
1.9
2.2
1.9
1.6
1.4
0.6
0.6
0.7
2005
1.5
1.5
1.4
1.4
1.4
1.4
1.3
1.3
1.3
1.2
1.2
1.3
1.3
1.4
1.6
2.1
2.3
2.4
2.5
2.8
2.9
3.0
2.8
2.8
2.6
2.5
2.3
2.1
2.2
2.1
2.1
1.9
1.6
1.5
1.6
1.7
1.8
2.2
1.9
1.6
1.5
0.8
1.0
1.2
1.7
2006
1.5
1.5
1.4
1.4
1.4
1.4
1.3
1.2
1.2
1.2
1.2
1.2
1.3
1.3
1.6
2.0
2.2
2.3
2.4
2.6
2.8
2.8
2.7
2.6
2.5
2.4
2.2
2.0
2.1
2.0
2.0
1.8
1.5
1.4
1.4
1.5
1.7
1.9
1.6
1.4
1.2
0.7
0.7
0.8
0.8 (0.1)
2007
1.4
1.4
1.4
1.3
1.3
1.3
1.3
1.2
1.2
1.1
1.2
1.2
1.2
1.3
1.6
2.0
2.2
2.3
2.3
2.6
2.7
2.8
2.6
2.5
2.4
2.3
2.1
1.9
2.0
1.9
1.9
1.7
1.4
1.3
1.4
1.4
1.6
1.8
1.5
1.3
1.1
0.7
0.7
0.8
0.8
0.3
0.7
2008
1.4
1.4
1.4
1.3
1.3
1.3
1.3
1.2
1.2
1.1
1.1
1.2
1.2
1.3
1.5
1.9
2.1
2.2
2.3
2.5
2.6
2.6
2.5
2.4
2.3
2.2
2.0
1.9
1.9
1.8
1.8
1.6
1.3
1.3
1.3
1.3
1.4
1.6
1.4
1.1
1.0
0.6
0.6
0.6
0.6
0.2
0.3 (0.1)
2009
1.3
1.3
1.3
1.2
1.2
1.2
1.2
1.1
1.1
1.0
1.0
1.1
1.1
1.2
1.4
1.8
2.0
2.1
2.1
2.3
2.4
2.5
2.3
2.3
2.1
2.0
1.8
1.7
1.7
1.6
1.6
1.4
1.1
1.0
1.1
1.1
1.2
1.3
1.0
0.8
0.6
0.2
0.2
0.1
2010
1.2
1.2
1.2
1.1
1.1
1.1
1.1
1.0
1.0
0.9
0.9
0.9
1.0
1.0
1.2
1.6
1.8
1.9
1.9
2.1
2.2
2.2
2.1
2.0
1.9
1.8
1.6
1.4
1.4
1.3
1.3
1.1
0.8
0.7
0.7
0.7
0.7
0.8
0.6
0.3
0.1 (0.3) (0.4) (0.5) (0.7) (1.2) (1.5) (2.2) (3.2) (4.4)
2011
1.1
1.1
1.0
1.0
1.0
1.0
0.9
0.8
0.8
0.8
0.8
0.8
0.8
0.9
1.1
1.4
1.6
1.7
1.7
1.9
2.0
2.0
1.9
1.8
1.6
1.5
1.3
1.1
1.1
1.1
1.0
0.8
0.5
0.4
0.4
0.4
0.4
0.5
0.2 (0.1) (0.3) (0.7) (0.8) (1.0) (1.3) (1.7) (2.1) (2.8) (3.6) (4.4) (4.4)
2012
1.0
1.0
1.0
0.9
0.9
0.9
0.8
0.8
0.7
0.7
0.7
0.7
0.7
0.8
1.0
1.3
1.5
1.5
1.6
1.8
1.8
1.8
1.7
1.6
1.5
1.4
1.2
1.0
1.0
0.9
0.8
0.6
0.3
0.2
0.2
0.2
0.2
0.2 (0.0) (0.3) (0.5) (0.9) (1.0) (1.2) (1.5) (1.9) (2.2) (2.8) (3.4) (3.9) (3.6) (2.8)
2013
1.0
1.0
0.9
0.8
0.8
0.8
0.8
0.7
0.7
0.6
0.6
0.6
0.6
0.7
0.9
1.2
1.4
1.4
1.5
1.6
1.7
1.7
1.6
1.5
1.3
1.2
1.0
0.9
0.8
0.8
0.7
0.5
0.2
0.1
0.1
0.0
0.0
0.1 (0.2) (0.4) (0.7) (1.0) (1.2) (1.3) (1.6) (2.0) (2.2) (2.7) (3.2) (3.5) (3.2) (2.6) (2.4)
2014
0.9
0.9
0.9
0.8
0.8
0.8
0.7
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.8
1.2
1.3
1.4
1.4
1.5
1.6
1.6
1.5
1.4
1.3
1.1
0.9
0.8
0.7
0.7
0.6
0.4
0.1
0.0
0.0 (0.0) (0.0) (0.0) (0.3) (0.5) (0.7) (1.1) (1.2) (1.3) (1.5) (1.9) (2.1) (2.5) (2.9) (3.1) (2.7) (2.2) (1.9) (1.3)
24 February 2015
0.2
4.2
193
101
105 103
100
99
100
99
97
96
93
92
93
92
93
96
95
93
91
87
85
82
82
83
82
83
89
1976 101
99
96
92
92
90
87
83
82
79
79
79
79
80
85
96
1977
99
98
95
91
90
89
86
82
81
78
78
78
78
79
84
95
99
1978 100
99
96
92
91
89
87
83
82
79
79
79
79
79
85
96
100 101
99
1979
96
95
91
88
87
86
83
79
78
76
75
76
75
76
81
92
95
97
96
1980
96
94
91
88
87
86
83
79
78
76
75
76
75
76
81
92
95
97
96
100
1981
97
95
92
89
88
86
84
80
79
76
76
76
76
76
82
92
96
97
96
101
98
94
94
92
89
85
84
81
81
81
81
81
87
98
102 104
103 107
107 106
98
97
96
93
89
87
84
84
84
84
85
91
102
106 108
107 112
112 111
104
98
93
92
89
88
89
88
89
96
108
112 113
112 117
117 117
109 105
98
96
93
92
93
93
93
100 113
117 119
118 123
123 122
115 110
105
104 103
99
99
99
99
100
107 120
125 127
126 131
131 130
122 118
112 107
113 127
132 134
133 139
139 138
129 124
115 129
134 136
135 141
141 140
131 126
118 132
138 140
138 144
145 144
135 130
104 103
121 136
141 143
142 148
148 147
138 133
107 105
102
126 142
148 150
148 155
155 154
145 139
112 110
107 105
135 152
158 160
158 166
166 164
154 148
119 118
115 112
107
138 155
161 163
162 169
169 168
158 152
122 120
117 114
109 102
139 156
163 165
163 170
171 169
159 153
123 121
118 115
140 157
164 166
164 172
172 170
160 154
124 122
119 116
101
140 158
164 166
165 172
172 171
160 154
124 122
119 116
101 100
139 157
163 165
164 171
171 170
160 153
123 122
118 115
100 100
99
145 163
170 172
170 178
178 177
166 160
128 127
123 120
104 104
104 104
150 169
175 178
176 184
184 183
172 165
133 131
127 124
108 107
107 108
154 173
180 182
180 189
189 187
176 169
136 134
130 127
111 110
110 110
106 102
160 180
187 189
188 196
196 195
183 176
141 139
136 132
115 114
114 115
160 181
188 190
188 197
197 196
184 176
142 140
136 133
116 115
115 115
100
161 181
189 191
189 198
198 197
185 177
143 141
137 134
116 115
115 116
101 101
162 183
190 193
191 199
200 198
186 179
144 142
138 135
117 116
116 117
102 101
101
165 186
193 196
194 203
203 201
189 182
146 144
140 137
119 118
118 119
103 103
102 102
165 186
193 196
194 203
203 201
189 182
146 144
140 137
119 118
118 119
103 103
102 102
100
166 187
194 197
195 204
204 203
190 183
147 145
141 138
120 119
119 119
104 104
103 102
101 101
166 187
194 197
195 204
204 202
190 183
147 145
141 138
120 119
119 119
104 103
103 102
101 101
163 183
190 193
191 200
200 198
186 179
144 142
138 135
117 116
116 117
102 101
101 100
98
98
98
98
155 175
182 184
183 191
191 189
178 171
138 136
132 129
112 111
111 112
97
97
96
96
94
94
94
94
96
149 167
174 176
175 182
183 181
170 163
132 130
126 123
107 106
106 107
102
99
97
93
93
92
92
90
90
89
90
91
96
144 163
169 171
170 177
177 176
165 159
128 126
123 120
104 103
103 104
100
96
94
90
90
90
89
87
88
87
87
89
93
97
141 159
165 167
166 173
173 172
161 155
125 123
120 117
102 101
101 101
97
94
92
88
88
87
87
85
85
85
85
87
91
95
98
139 156
163 165
163 171
171 170
159 153
123 121
118 115
100
99
96
93
91
87
87
86
86
84
84
84
84
86
89
94
96
24 February 2015
101
101
99
100
103
100
99
194
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
1983
(4.3)
1984
(1.2)
1.9
1985
(2.7)
(1.9)
(5.5)
1986
(1.5)
(0.5)
(1.7)
2.3
1987
(0.6)
0.4
(0.1)
2.7
3.1
1988
0.6
1.6
1.5
3.9
4.8
6.5
1989
1.4
2.3
2.4
4.5
5.3
6.4
6.3
1990
0.6
1.3
1.2
2.7
2.7
2.6
0.8
(4.5)
1991
0.6
1.3
1.2
2.3
2.3
2.2
0.7
(1.9)
1992
1.9
2.6
2.7
3.9
4.2
4.4
3.9
3.2
7.2
14.1
1993
3.3
4.1
4.4
5.7
6.2
6.7
6.8
6.9
11.0
16.5
18.9
1994
2.1
2.7
2.8
3.8
3.9
4.1
3.7
3.1
5.2
6.7
3.1
(10.5)
1995
2.6
3.2
3.3
4.2
4.4
4.6
4.3
4.0
5.8
7.1
4.9
(1.5)
8.5
1996
2.7
3.2
3.3
4.2
4.4
4.5
4.3
4.0
5.5
6.5
4.6
0.3
6.2
1997
3.1
3.7
3.8
4.6
4.8
5.0
4.8
4.7
6.0
7.0
5.6
2.5
7.3
6.7
9.4
1998
3.9
4.5
4.7
5.5
5.8
6.1
6.0
6.0
7.4
8.3
7.4
5.3
9.6
10.0
13.2
1999
3.9
4.4
4.6
5.4
5.6
5.8
5.7
5.7
6.9
7.7
6.8
4.9
8.3
8.3
9.7
9.9
3.2
2000
3.7
4.2
4.3
5.0
5.2
5.4
5.3
5.2
6.2
6.8
5.9
4.2
6.9
6.6
7.2
6.5
1.6
0.1
2001
3.4
3.8
4.0
4.6
4.7
4.8
4.7
4.6
5.5
5.9
5.1
3.5
5.6
5.2
5.4
4.4
0.5
(0.7)
(1.6)
2002
3.5
3.9
4.0
4.6
4.7
4.9
4.7
4.6
5.4
5.9
5.1
3.6
5.6
5.2
5.4
4.6
1.7
1.2
1.7
5.1
2003
3.5
3.9
4.0
4.6
4.7
4.8
4.7
4.6
5.3
5.7
5.0
3.7
5.4
5.0
5.1
4.4
2.1
1.8
2.4
4.5
3.9
2004
3.6
4.0
4.1
4.6
4.7
4.8
4.7
4.6
5.3
5.6
5.0
3.8
5.3
5.0
5.1
4.5
2.6
2.4
3.0
4.6
4.4
4.9
2005
3.7
4.1
4.2
4.7
4.8
4.9
4.8
4.7
5.4
5.7
5.1
4.0
5.5
5.2
5.3
4.8
3.2
3.2
3.8
5.2
5.2
5.8
6.7
2006
3.5
3.8
3.9
4.4
4.5
4.5
4.4
4.3
4.9
5.2
4.6
3.5
4.8
4.5
4.5
4.0
2.5
2.4
2.8
3.7
3.3
3.1
2.2
(2.1)
2007
3.4
3.7
3.8
4.2
4.3
4.4
4.3
4.2
4.7
4.9
4.4
3.4
4.5
4.2
4.3
3.7
2.4
2.3
2.6
3.3
2.9
2.7
2.0
(0.3)
1.4
2008
3.2
3.5
3.5
3.9
4.0
4.1
3.9
3.8
4.3
4.5
3.9
3.0
4.1
3.7
3.7
3.2
1.9
1.8
2.0
2.5
2.1
1.7
0.9
(0.9)
(0.4)
2009
3.2
3.4
3.5
3.9
4.0
4.0
3.9
3.8
4.2
4.4
3.9
3.0
4.0
3.7
3.7
3.2
2.0
1.9
2.1
2.6
2.2
2.0
1.4
0.1
0.8
0.5
3.1
2010
3.2
3.5
3.6
4.0
4.0
4.1
4.0
3.9
4.3
4.5
4.0
3.2
4.1
3.8
3.8
3.4
2.3
2.2
2.4
2.9
2.6
2.4
2.0
1.1
1.9
2.1
4.2
5.3
2011
3.6
3.9
4.0
4.3
4.4
4.5
4.4
4.3
4.8
5.0
4.5
3.8
4.7
4.4
4.5
4.1
3.2
3.2
3.5
4.0
3.9
3.9
3.7
3.2
4.3
5.0
7.5
9.8
14.4
2012
3.5
3.8
3.8
4.2
4.3
4.3
4.2
4.1
4.5
4.7
4.3
3.6
4.4
4.2
4.2
3.8
3.0
2.9
3.2
3.6
3.5
3.4
3.3
2.8
3.6
4.0
5.6
6.5
7.1
0.2
2013
3.2
3.5
3.5
3.9
3.9
4.0
3.9
3.8
4.2
4.3
3.9
3.2
4.0
3.7
3.7
3.3
2.5
2.4
2.6
3.0
2.8
2.7
2.4
1.9
2.5
2.7
3.7
3.8
3.3
(1.8)
(3.9)
2014
3.6
3.8
3.9
4.2
4.3
4.3
4.3
4.2
4.6
4.7
4.3
3.7
4.4
4.2
4.2
3.9
3.2
3.2
3.4
3.8
3.7
3.7
3.5
3.2
3.9
4.2
5.3
5.8
5.9
3.2
4.7
24 February 2015
2013
0.7
4.0
17.1
(2.1)
14.0
195
1983
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
96
1984
98
102
1985
92
96
1986
94
98
97
102
1987
97
102
100
105
103
1988
103
108
106
112
110
106
1989
110
115
113
119
117
113
106
1990
105
110
108
114
111
108
102
1991
106
111
108
115
112
109
102
96
101
1992
121
126
124
131
128
124
117
110
115
114
1993
144
150
147
156
152
148
139
130
137
136
119
1994
128
134
132
139
136
132
124
117
122
121
106
1995
139
146
143
151
148
143
135
127
133
132
115
97
108
1996
145
151
148
157
154
149
140
132
138
137
120
101
113
104
1997
158
166
162
172
168
163
153
144
151
150
131
110
123
114
109
1998
186
194
190
201
197
191
179
169
177
175
154
129
144
133
128
1999
191
200
196
208
203
197
185
174
182
181
158
133
149
137
132
121
103
2000
192
200
196
208
203
197
185
174
182
181
159
133
149
138
132
121
103
2001
189
197
193
205
200
194
182
171
179
178
156
131
147
135
130
119
102
99
98
2002
198
207
203
215
210
204
191
180
189
187
164
138
154
142
137
125
107
104
103
105
2003
206
215
211
223
218
212
199
187
196
194
170
143
160
148
142
130
111
108
107
109
104
2004
216
226
221
234
229
222
209
196
206
204
179
150
168
155
149
136
116
113
113
115
109
2005
231
241
236
250
244
237
223
210
219
218
191
161
179
165
159
145
124
120
120
122
116
112
107
2006
226
236
231
245
239
232
218
205
215
213
187
157
176
162
156
142
122
118
118
120
114
110
105
98
2007
229
239
235
248
243
236
221
208
218
216
190
160
178
164
158
144
123
120
119
121
116
111
106
99
2008
224
234
230
243
238
231
217
204
213
212
186
156
175
161
155
141
121
117
117
119
113
109
104
97
99
98
2009
231
241
237
251
245
238
223
210
220
218
191
161
180
166
160
146
125
121
121
123
117
112
107
100
102
101
103
2010
243
254
249
264
258
250
235
221
232
230
202
170
190
175
168
154
131
127
127
129
123
118
113
106
108
106
109
105
2011
278
291
285
302
295
287
269
253
265
263
231
194
217
200
192
176
150
146
145
148
141
135
129
121
123
122
124
121
114
2012
279
292
286
303
296
287
270
254
266
264
231
194
217
200
193
176
150
146
146
148
141
136
129
121
124
122
124
121
115
2013
268
280
275
291
285
276
259
244
255
254
222
187
209
193
185
169
145
140
140
142
135
130
124
116
119
117
120
116
110
96
96
2014
306
320
314
332
324
315
296
278
291
289
253
213
238
220
211
193
165
160
160
162
154
149
142
133
136
134
136
132
126
110
110
24 February 2015
94
95
89
117
100
105
101
100
114
196
Analyst Certification
We, Michael Gavin, Keith Parker, Tal Shapsa, Guillermo Felices, Krishna Goradia, Jim McCormick, Michael Gapen, Joseph Abate, Eric Gross, Jeffrey Meli,
Brian Monteleone, Conor Pigott, Hamish Pepper, Nikolaos Sgouropoulos, Sreekala Kochugovindan and Siddhartha Sanyal, hereby certify (1) that the
views expressed in this research report accurately reflect our personal views about any or all of the subject securities or issuers referred to in this research
report and (2) no part of our compensation was, is or will be directly or indirectly related to the specific recommendations or views expressed in this
research report.
Each research report excerpted herein was certified under SEC Regulation AC by the analyst primarily responsible for such report as follows: I hereby
certify that: 1) the views expressed in this research report accurately reflect my personal views about any or all of the subject securities referred to in this
report and; 2) no part of my compensation was, is or will be directly or indirectly related to the specific recommendations or views expressed in this
report.
Important Disclosures:
Barclays Research is a part of the Corporate and Investment Banking division of Barclays Bank PLC and its affiliates (collectively and each individually,
"Barclays"). For current important disclosures regarding companies that are the subject of this research report, please send a written request to: Barclays
Research Compliance, 745 Seventh Avenue, 14th Floor, New York, NY 10019 or refer to http://publicresearch.barclays.com or call 212-526-1072.
Barclays Capital Inc. and/or one of its affiliates does and seeks to do business with companies covered in its research reports. As a result, investors
should be aware that Barclays may have a conflict of interest that could affect the objectivity of this report. Barclays Capital Inc. and/or one of its affiliates
regularly trades, generally deals as principal and generally provides liquidity (as market maker or otherwise) in the debt securities that are the subject of
this research report (and related derivatives thereof). Barclays trading desks may have either a long and/or short position in such securities, other financial
instruments and/or derivatives, which may pose a conflict with the interests of investing customers. Where permitted and subject to appropriate
information barrier restrictions, Barclays fixed income research analysts regularly interact with its trading desk personnel regarding current market
conditions and prices. Barclays fixed income research analysts receive compensation based on various factors including, but not limited to, the quality of
their work, the overall performance of the firm (including the profitability of the investment banking department), the profitability and revenues of the
Fixed Income, Currencies and Commodities Division and the potential interest of the firms investing clients in research with respect to the asset class
covered by the analyst. To the extent that any historical pricing information was obtained from Barclays trading desks, the firm makes no representation
that it is accurate or complete. All levels, prices and spreads are historical and do not represent current market levels, prices or spreads, some or all of
which may have changed since the publication of this document. Barclays produces various types of research including, but not limited to, fundamental
analysis, equity-linked analysis, quantitative analysis, and trade ideas. Recommendations contained in one type of research may differ from
recommendations contained in other types of research, whether as a result of differing time horizons, methodologies, or otherwise. Unless otherwise
indicated, Barclays trade ideas are provided as of the date of this report and are subject to change without notice due to changes in prices. In order to
access
Barclays
Statement
regarding
Research
Dissemination
Policies
and
Procedures,
please
refer
to
https://live.barcap.com/publiccp/RSR/nyfipubs/disclaimer/disclaimer-research-dissemination.html. In order to access Barclays Research Conflict
Management Policy Statement, please refer to: https://live.barcap.com/publiccp/RSR/nyfipubs/disclaimer/disclaimer-conflict-management.html.
The MSCI sourced information is the exclusive property of MSCI Inc. (MSCI). Without prior written permission of MSCI, this information and any other
MSCI intellectual property may not be reproduced, redisseminated or used to create any financial products, including any indices. This information is
provided on an as is basis. The user assumes the entire risk of any use made of this information. MSCI, its affiliates and any third party involved in, or
related to, computing or compiling the information hereby expressly disclaim all warranties of originality, accuracy, completeness, merchantability or
fitness for a particular purpose with respect to any of this information. Without limiting any of the foregoing, in no event shall MSCI, any of its affiliates or
any third party involved in, or related to, computing or compiling the information have any liability for any damages of any kind. MSCI and the MSCI
indexes are services marks of MSCI and its affiliates.
Disclaimer:
This publication has been prepared by the Corporate and Investment Banking division of Barclays Bank PLC and/or one or more of its affiliates
(collectively and each individually, "Barclays"). It has been issued by one or more Barclays legal entities within its Corporate and Investment Banking
division as provided below. It is provided to our clients for information purposes only, and Barclays makes no express or implied warranties, and expressly
disclaims all warranties of merchantability or fitness for a particular purpose or use with respect to any data included in this publication. Barclays will not
treat unauthorized recipients of this report as its clients. Prices shown are indicative and Barclays is not offering to buy or sell or soliciting offers to buy or
sell any financial instrument.
Without limiting any of the foregoing and to the extent permitted by law, in no event shall Barclays, nor any affiliate, nor any of their respective officers,
directors, partners, or employees have any liability for (a) any special, punitive, indirect, or consequential damages; or (b) any lost profits, lost revenue, loss
of anticipated savings or loss of opportunity or other financial loss, even if notified of the possibility of such damages, arising from any use of this
publication or its contents.
Other than disclosures relating to Barclays, the information contained in this publication has been obtained from sources that Barclays Research believes
to be reliable, but Barclays does not represent or warrant that it is accurate or complete. Barclays is not responsible for, and makes no warranties
whatsoever as to, the content of any third-party web site accessed via a hyperlink in this publication and such information is not incorporated by
reference.
The views in this publication are those of the author(s) and are subject to change, and Barclays has no obligation to update its opinions or the information
in this publication. The analyst recommendations in this publication reflect solely and exclusively those of the author(s), and such opinions were prepared
independently of any other interests, including those of Barclays and/or its affiliates. This publication does not constitute personal investment advice or
take into account the individual financial circumstances or objectives of the clients who receive it. The securities discussed herein may not be suitable for
all investors. Barclays recommends that investors independently evaluate each issuer, security or instrument discussed herein and consult any
independent advisors they believe necessary. The value of and income from any investment may fluctuate from day to day as a result of changes in
relevant economic markets (including changes in market liquidity). The information herein is not intended to predict actual results, which may differ
substantially from those reflected. Past performance is not necessarily indicative of future results.
This material has been issued and approved for distribution in the UK and European Economic Area (EEA) by Barclays Bank PLC. It is being made
available primarily to persons who are investment professionals as that term is defined in Article 19 of the Financial Services and Markets Act 2000
(Financial Promotion) Order 2005. It is directed at, and therefore should only be relied upon by, persons who have professional experience in matters
relating to investments. The investments to which it relates are available only to such persons and will be entered into only with such persons. Barclays
LAST PAGE
Bank PLC is authorised by the Prudential Regulation Authority and regulated by the Financial Conduct Authority and the Prudential Regulation Authority
and is a member of the London Stock Exchange.
The Corporate and Investment Banking division of Barclays undertakes U.S. securities business in the name of its wholly owned subsidiary Barclays
Capital Inc., a FINRA and SIPC member. Barclays Capital Inc., a U.S. registered broker/dealer, is distributing this material in the United States and, in
connection therewith accepts responsibility for its contents. Any U.S. person wishing to effect a transaction in any security discussed herein should do so
only by contacting a representative of Barclays Capital Inc. in the U.S. at 745 Seventh Avenue, New York, New York 10019.
Non-U.S. persons should contact and execute transactions through a Barclays Bank PLC branch or affiliate in their home jurisdiction unless local
regulations permit otherwise.
Barclays Bank PLC, Paris Branch (registered in France under Paris RCS number 381 066 281) is regulated by the Autorit des marchs financiers and the
Autorit de contrle prudentiel. Registered office 34/36 Avenue de Friedland 75008 Paris.
This material is distributed in Canada by Barclays Capital Canada Inc., a registered investment dealer and member of IIROC (www.iiroc.ca).
Subject to the conditions of this publication as set out above, the Corporate & Investment Banking Division of Absa Bank Limited, an authorised financial
services provider (Registration No.: 1986/004794/06. Registered Credit Provider Reg No NCRCP7), is distributing this material in South Africa. Absa Bank
Limited is regulated by the South African Reserve Bank. This publication is not, nor is it intended to be, advice as defined and/or contemplated in the
(South African) Financial Advisory and Intermediary Services Act, 37 of 2002, or any other financial, investment, trading, tax, legal, accounting, retirement,
actuarial or other professional advice or service whatsoever. Any South African person or entity wishing to effect a transaction in any security discussed
herein should do so only by contacting a representative of the Corporate & Investment Banking Division of Absa Bank Limited in South Africa, 15 Alice
Lane, Sandton, Johannesburg, Gauteng 2196. Absa Bank Limited is a member of the Barclays group.
In Japan, foreign exchange research reports are prepared and distributed by Barclays Bank PLC Tokyo Branch. Other research reports are distributed to
institutional investors in Japan by Barclays Securities Japan Limited. Barclays Securities Japan Limited is a joint-stock company incorporated in Japan with
registered office of 6-10-1 Roppongi, Minato-ku, Tokyo 106-6131, Japan. It is a subsidiary of Barclays Bank PLC and a registered financial instruments
firm regulated by the Financial Services Agency of Japan. Registered Number: Kanto Zaimukyokucho (kinsho) No. 143.
Barclays Bank PLC, Hong Kong Branch is distributing this material in Hong Kong as an authorised institution regulated by the Hong Kong Monetary
Authority. Registered Office: 41/F, Cheung Kong Center, 2 Queen's Road Central, Hong Kong.
Information on securities/instruments that trade in Taiwan or written by a Taiwan-based research analyst is distributed by Barclays Capital Securities
Taiwan Limited to its clients. The material on securities/instruments not traded in Taiwan is not to be construed as 'recommendation' in Taiwan. Barclays
Capital Securities Taiwan Limited does not accept orders from clients to trade in such securities. This material may not be distributed to the public media
or used by the public media without prior written consent of Barclays.
This material is distributed in South Korea by Barclays Capital Securities Limited, Seoul Branch.
All equity research material is distributed in India by Barclays Securities (India) Private Limited (SEBI Registration No: INB/INF 231292732 (NSE), INB/INF
011292738 (BSE) | Corporate Identification Number: U67120MH2006PTC161063 | Registered Office: 208 | Ceejay House | Dr. Annie Besant Road |
Shivsagar Estate | Worli | Mumbai - 400 018 | India, Phone: + 91 22 67196363). Other research reports are distributed in India by Barclays Bank PLC, India
Branch.
Barclays Bank PLC Frankfurt Branch distributes this material in Germany under the supervision of Bundesanstalt fr Finanzdienstleistungsaufsicht (BaFin).
This material is distributed in Malaysia by Barclays Capital Markets Malaysia Sdn Bhd.
This material is distributed in Brazil by Banco Barclays S.A.
This material is distributed in Mexico by Barclays Bank Mexico, S.A.
Barclays Bank PLC in the Dubai International Financial Centre (Registered No. 0060) is regulated by the Dubai Financial Services Authority (DFSA).
Principal place of business in the Dubai International Financial Centre: The Gate Village, Building 4, Level 4, PO Box 506504, Dubai, United Arab Emirates.
Barclays Bank PLC-DIFC Branch, may only undertake the financial services activities that fall within the scope of its existing DFSA licence. Related financial
products or services are only available to Professional Clients, as defined by the Dubai Financial Services Authority.
Barclays Bank PLC in the UAE is regulated by the Central Bank of the UAE and is licensed to conduct business activities as a branch of a commercial bank
incorporated outside the UAE in Dubai (Licence No.: 13/1844/2008, Registered Office: Building No. 6, Burj Dubai Business Hub, Sheikh Zayed Road,
Dubai City) and Abu Dhabi (Licence No.: 13/952/2008, Registered Office: Al Jazira Towers, Hamdan Street, PO Box 2734, Abu Dhabi).
Barclays Bank PLC in the Qatar Financial Centre (Registered No. 00018) is authorised by the Qatar Financial Centre Regulatory Authority (QFCRA).
Barclays Bank PLC-QFC Branch may only undertake the regulated activities that fall within the scope of its existing QFCRA licence. Principal place of
business in Qatar: Qatar Financial Centre, Office 1002, 10th Floor, QFC Tower, Diplomatic Area, West Bay, PO Box 15891, Doha, Qatar. Related financial
products or services are only available to Business Customers as defined by the Qatar Financial Centre Regulatory Authority.
This material is distributed in the UAE (including the Dubai International Financial Centre) and Qatar by Barclays Bank PLC.
This material is distributed in Saudi Arabia by Barclays Saudi Arabia ('BSA'). It is not the intention of the publication to be used or deemed as
recommendation, option or advice for any action(s) that may take place in future. Barclays Saudi Arabia is a Closed Joint Stock Company, (CMA License
No. 09141-37). Registered office Al Faisaliah Tower, Level 18, Riyadh 11311, Kingdom of Saudi Arabia. Authorised and regulated by the Capital Market
Authority, Commercial Registration Number: 1010283024.
This material is distributed in Russia by OOO Barclays Capital, affiliated company of Barclays Bank PLC, registered and regulated in Russia by the FSFM.
Broker License #177-11850-100000; Dealer License #177-11855-010000. Registered address in Russia: 125047 Moscow, 1st Tverskaya-Yamskaya str.
21.
This material is distributed in Singapore by the Singapore branch of Barclays Bank PLC, a bank licensed in Singapore by the Monetary Authority of
Singapore. For matters in connection with this report, recipients in Singapore may contact the Singapore branch of Barclays Bank PLC, whose registered
address is One Raffles Quay Level 28, South Tower, Singapore 048583.
Barclays Bank PLC, Australia Branch (ARBN 062 449 585, AFSL 246617) is distributing this material in Australia. It is directed at 'wholesale clients' as
defined by Australian Corporations Act 2001.
IRS Circular 230 Prepared Materials Disclaimer: Barclays does not provide tax advice and nothing contained herein should be construed to be tax advice.
Please be advised that any discussion of U.S. tax matters contained herein (including any attachments) (i) is not intended or written to be used, and
cannot be used, by you for the purpose of avoiding U.S. tax-related penalties; and (ii) was written to support the promotion or marketing of the
transactions or other matters addressed herein. Accordingly, you should seek advice based on your particular circumstances from an independent tax
advisor.
Copyright Barclays Bank PLC (2015). All rights reserved. No part of this publication may be reproduced or redistributed in any manner without the prior
written permission of Barclays. Barclays Bank PLC is registered in England No. 1026167. Registered office 1 Churchill Place, London, E14 5HP. Additional
information regarding this publication will be furnished upon request.
LAST PAGE