Text

INTERNATIONAL MONETARY FUND
GLOBAL
FINANCIAL
STABILITY
REPORT
Bridge to Recovery
2020
OCT
INTERNATIONAL MONETARY FUND
GLOBAL
FINANCIAL
STABILITY
REPORT
Bridge to Recovery
2020
OCT
©2020 International Monetary Fund
IMF CSF Creative Solutions Division

Composition: AGS, An RR Donnelley Company; and The Grauel Group
Cataloging-in-Publication Data
IMF Library
Names: International Monetary Fund.

Title: Global financial stability report.
Other titles: GFSR | World economic and financial surveys, 0258-7440
Description: Washington, DC : International Monetary Fund, 2002- | Semiannual | Some issues also have thematic
titles. | Began with issue for March 2002.
Subjects: LCSH: Capital market—Statistics—Periodicals. | International finance—Forecasting—Periodicals. |
Economic stabilization—Periodicals.
Classification: LCC HG4523.G557
ISBN 978-1-51355-422-8 (Paper)

978-1-51355-851-6 (ePub)
978-1-51355-855-4 (PDF)
Disclaimer: The Global Financial Stability Report (GFSR) is a survey by the IMF staff published twice
a year, in the spring and fall. The report draws out the financial ramifications of economic issues high-
lighted in the IMF’s World Economic Outlook (WEO). The report was prepared by IMF staff and has
benefited from comments and suggestions from Executive Directors following their discussion of the
report on September 30, 2020. The views expressed in this publication are those of the IMF staff and
do not necessarily represent the views of the IMF’s Executive Directors or their national authorities.
Recommended citation: International Monetary Fund. 2020. Global Financial Stability Report: Bridge
to Recovery. Washington, DC, October.
Please send orders to:

International Monetary Fund, Publications Services
P.O. Box 92780, Washington, DC 20090, U.S.A.
Tel.: (202) 623-7430 Fax: (202) 623-7201
E-mail: publications@imf.org
www.bookstore.imf.org
www.elibrary.imf.org
CONTENTS
Assumptions and Conventions vi

Further Information vii
Preface viii
Foreword ix
Executive Summary xi
IMF Executive Board Discussion of the Outlook, October 2020 xvi
Chapter 1 Global Financial Stability Overview: Bridge to Recovery 1

Box 1.1. Are Financial Stability Risks Rising in Commercial Real Estate Markets? 28
Box 1.2. The Behavior of Investment Funds during COVID-19 Market Turmoil 30
Box 1.3. Interlinkages among Local Government, Corporate, and Bank Vulnerabilities in China 32
References 34
Online Annex 1.1. Technical Note
Chapter 2 Emerging and Frontier Markets: A Greater Set of Policy Options to Restore Stability 35
References 52
Chapter 3 Corporate Funding: Liquidity Strains Cushioned by a Powerful Set of Policies 53

References 69
Online Annex 3.1. Data Sources
Online Annex 3.2. Firms’ Choice of Debt Financing Instrument
Online Annex 3.3. Identification of Aggregate Credit Supply Shocks
Online Annex 3.4. Estimation of the Excess Bond Premium
Online Annex 3.5. Firm-Level Stock Market Performance
Online Annex 3.6. Policy Intervention Analysis
Online Box 3.1. Group of Seven Policy Measures that Supported Corporate Funding during the
Containment Phase of the COVID-19 Pandemic
Online Box 3.2. Corporate Debt Financing Choice and Firm Characteristics
Chapter 4 Bank Capital: COVID-19 Challenges and Policy Responses 71

Box 4.1. The Role of Corporate and Consumer Risk in the Evolution of Banks’ Loan
Loss Provisions 85
References 86
Chapter 5 Corporate Sustainability: Firms’ Environmental Performance and the COVID-19 Crisis 87
Box 5.1. Climate Index Based on Firms’ Earnings Calls 95
References 96
Online Annex 5.1. Data Sources
Online Annex 5.2. Financial Constraints and Firms’ Environmental Performance
Online Annex 5.3. Firms’ Environmental Performance, Financial, Economic, and Oil Market Shocks
Online Annex 5.4. Climate Change and Disaster Indices
International Monetary Fund | October 2020 iii

GLOBAL FINANCIAL STABILIT Y REPORT: BRIDGE TO RECOVERY
Tables
Table 1.1. Monetary and Financial Policy Road Map 23
Table 2.1. Asset Purchase Programs in Response to COVID-19 in Emerging Market Economies 39
Figures
Figure 1. Proportion of Systemically Important Countries with Elevated Vulnerabilities, by Sector xii
Figure 2. Key Drivers of Global Financial Conditions Indices xii
Figure 3. Near-Term Global Growth Forecast Densities xii
Figure 4. Capital Flows at Risk: Near-term Forecasts of Portfolio Flows xiii
Figure 5. Stock Market Performance in 2020: Sectoral Contributions xiii
Figure 6. Bond Spread Misalignment xiii
Figure 7. Publicly Listed Firms: Debt at Risk xiv
Figure 8. Aggregate Nonfinancial Corporate Debt xiv
Figure 9. Distribution of Bank Assets by Capital Ratio under the October 2020 WEO
Adverse Scenario, with Policy Mitigation xiv
Figure 10. Nonbank Financial Institutions: Financial Vulnerability Indices and Sector Size xv
Figure 11. Corporate, Bank, and Sovereign Vulnerabilities in 29 Jurisdictions with
Systemically Important Financial Sectors xv
Figure 12. Change in Local Currency Government Bonds Outstanding by Holder,
end-February–June 2020 xv
Figure 1.1. GDP Growth: The COVID-19 Crisis versus the Global Financial Crisis 2
Figure 1.2. Global Growth at Risk 3
Figure 1.3. Central Bank Measures in Major Advanced Economies—Game Changer 4
Figure 1.4. Global Financial Conditions 5
Figure 1.5. Global Equity Markets: Impact of COVID-19 on Countries and Sectors 6
Figure 1.6. Equity Market Valuations 7
Figure 1.7. Market Volatility and Economic Uncertainty 8
Figure 1.8. Credit Market Valuations 9
Figure 1.9. Global Financial Vulnerabilities: High and Rising 11
Figure 1.10. Easier Funding Conditions and Rising Debt 13
Figure 1.11. Solvency Risks in the Corporate Sector 14
Figure 1.12. Solvency Risks in the Household Sector 15
Figure 1.13. Banking Sector: Potential Losses in the October 2020 World Economic
Outlook Adverse Scenario 16
Figure 1.14. Vulnerabilities in the Nonbank Financial Sector 17
Figure 1.15. Financial Leverage and Global Cross-Asset Correlations 18
Figure 1.16. Sovereign Vulnerabilities and Interconnectedness 19
Figure 1.17. Emerging Market Financing: Challenges, Options, and Risks 21
Figure 1.18. Emerging and Frontier Market Economy Spreads and Market Access 22
Figure 1.1.1. Trends and Developments in Commercial Real Estate Markets 29
Figure 1.2.1. Vulnerabilities of Fixed-Income Funds Exposed during the March 2020
Market Turmoil 31
Figure 1.3.1. Interlinkages among Local Government, Corporate, and Bank Vulnerabilities
in China 33
Figure 2.1. Emerging Market Policy Response to the COVID-19 Pandemic 36
Figure 2.2. FX Interventions and Reserve Operations 38
Figure 2.3. Central Bank Asset Purchases in Emerging Markets 40
Figure 2.4. Stress in Local Currency Bond and FX Markets 41
Figure 2.5. Bond Stress and Asset Purchase Programs in Emerging Market Economies 43
Figure 2.6. Market Reaction to Domestic Asset Purchase Program Announcements 44
iv International Monetary Fund | October 2020

Contents
Figure 2.7. Asset Purchase Program Announcements and Sovereign Bond Yields 45
Figure 2.8. Asset Purchase Program Announcements and Domestic Currencies 46
Figure 2.9. Considerations for Asset Purchase Programs 47
Figure 2.10. Frontier Economies Have a Challenging Road Ahead 49
Figure 2.11. Large Shares of Senior Creditors Could Lead to Higher Spreads 51
Figure 3.1. Bank Lending to Nonfinancial Firms and Government Liquidity Support 56
Figure 3.2. Developments in Syndicated Loan Markets 57
Figure 3.3. Corporate Bond and Commercial Paper Issuance 58
Figure 3.4. Change in Corporate Cash-to-Assets Ratio and Corporate Bank Deposits 61
Figure 3.5. Evolution of Credit Supply Conditions 62
Figure 3.6. Firm-Level Stock Market Performance 65
Figure 3.7. The Effect of Policies on Vulnerable Firms 67
Figure 4.1. Historical Context: Magnitude of the Current Crisis and the Ex Ante
Position of Banks 73
Figure 4.2. Mitigation Policies Announced since February 1, 2020, by Category and Jurisdiction 74
Figure 4.3. Magnitude of Announced Mitigation Policies 75
Figure 4.4. Scenarios for Stress Test Simulation 77
Figure 4.5. Bank Solvency under COVID-19 without Policy Mitigation 78
Figure 4.6. Bank Solvency under COVID-19 with Policy Mitigation 81
Figure 4.1.1. Additional Quarterly Provisioning 85
Figure 5.1. The Energy Transition during the COVID-19 Crisis 88
Figure 5.2. The COVID-19 Crisis and Green Investments 89
Figure 5.3. Financial Constraints, Financial Stress, and Environmental Performance 91
Figure 5.4. Economic Shocks and Environmental Performance 92
Figure 5.5. Oil Market Shocks and Environmental Performance 93
Figure 5.1.1. Climate Index 95
Editor's Note (November 13, 2020)

This online version of the GFSR has been updated to reflect the following changes to the version published online on
October 13 and 23:
- Chapter 1: Figure 1.12 Solvency Risks in Small and Medium-Sized Enterprises: Model-Based Scenario Analysis, which appeared
on page 15 of the October 13 version was deleted, along with text related to this figure that appeared on pages 12 and 14 and
section C of Online Annex 1.1.
- Chapter 2, Figure 2.6, panel 4: The title was updated to Intraday Asset Price Movement on the Days of Asset Purchase
Program Announcements for Selected EMs (South Africa, India, and Poland)
International Monetary Fund | October 2020 v

ASSUMPTIONS AND CONVENTIONS
The following conventions are used throughout the Global Financial Stability Report (GFSR):
. . . to indicate that data are not available or not applicable;
— to indicate that the figure is zero or less than half the final digit shown or that the item does not exist;
– between years or months (for example, 2019–20 or January–June) to indicate the years or months covered,
including the beginning and ending years or months;
/ between years or months (for example, 2019/20) to indicate a fiscal or financial year.
“Billion” means a thousand million.
“Trillion” means a thousand billion.
“Basis points” refers to hundredths of 1 percentage point (for example, 25 basis points are equivalent to ¼ of
1 percentage point).
If no source is listed on tables and figures, data are based on IMF staff estimates or calculations.
Minor discrepancies between sums of constituent figures and totals shown reflect rounding.
As used in this report, the terms “country” and “economy” do not in all cases refer to a territorial entity that is a state
as understood by international law and practice. As used here, the term also covers some territorial entities that are
not states but for which statistical data are maintained on a separate and independent basis.
The boundaries, colors, denominations, and any other information shown on the maps do not imply, on the part
of the International Monetary Fund, any judgment on the legal status of any territory or any endorsement or
acceptance of such boundaries.
vi International Monetary Fund | October 2020

FURTHER INFORMATION
Corrections and Revisions

The data and analysis appearing in the Global Financial Stability Report are compiled by the IMF staff at the
time of publication. Every effort is made to ensure their timeliness, accuracy, and completeness. When errors are
discovered, corrections and revisions are incorporated into the digital editions available from the IMF website and
on the IMF eLibrary (see below). All substantive changes are listed in the online table of contents.
Print and Digital Editions

Print
Print copies of this Global Financial Stability Report can be ordered from the IMF bookstore at imfbk.st/29273.
Digital
Multiple digital editions of the Global Financial Stability Report, including ePub, enhanced PDF, and HTML,
are available on the IMF eLibrary at www.elibrary.imf.org/OCT20GFSR.
Download a free PDF of the report and data sets for each of the charts therein from the IMF website at
www.imf.org/publications/gfsr or scan the QR code below to access the Global Financial Stability Report web page
directly:
Copyright and Reuse

Information on the terms and conditions for reusing the contents of this publication are at www.imf.org/
external/terms.htm.
International Monetary Fund | October 2020 vii

PREFACE
The Global Financial Stability Report (GFSR) assesses key vulnerabilities the global financial system is exposed
to. In normal times, the report seeks to play a role in preventing crises by highlighting policies that may mitigate
systemic risks, thereby contributing to global financial stability and the sustained economic growth of the IMF’s
member countries.
The analysis in this report was coordinated by the Monetary and Capital Markets (MCM) Department under
the general direction of Tobias Adrian, Director. The project was directed by Fabio Natalucci, Deputy Director,
as well as by Claudio Raddatz, former Advisor, Anna Ilyina, Division Chief, Evan Papageorgiou, Deputy Division
Chief, Mahvash Qureshi, Division Chief, and Jérôme Vandenbussche, Deputy Division Chief. It benefited from
comments and suggestions from the senior staff in the MCM Department.
Individual contributors to the report were Sergei Antoshin, Romain Bouis, John Caparusso, Yingyuan Chen,
Dan Cheng, Fabio Cortes, Reinout De Bock, Andrea Deghi, Xioadan Ding, Dimitris Drakopoulos, Kelly
Eckhold, Ibrahim Ergen, Salih Fendoglu, Ken (Zhi) Gan, Deepali Gautam, Rohit Goel, Pierpaolo Grippa,
Marco Gross, Pierre Guérin, Sanjay Hazarika, Frank Hespeler, Henry Hoyle, Mohamed Jaber, Phakawa Jeasakul,
Oksana Khadarina, Piyusha Khot, Annamaria Kokenyne, Ivo Krznar, Dimitrios Laliotis, Fabian Lipinsky, Pavel
Lukyantsau, Elizabeth Mahoney, Sheheryar Malik, Samuel Mann, Manuel Perez, Dmitri Petrov, Nicola Pierri,
Thomas Piontek, Umang Rawat, Jochen Markus Schmittmann, Patrick Schneider, Dulani Seneviratne, Can Sever,
Juan Solé, Felix Suntheim, Thierry Tressel, Tomohiro Tsuruga, Germán Villegas Bauer, Jeffrey Williams, Yizhi Xu,
Dmitry Yakovlev, Akihiko Yokoyama, and Xingmi Zheng. Input was provided by Hee Kyong Chon, Alan Feng,
Caio Ferreira, Alejandro Lopez, Luc Riedweg, and Julia Xueliang Wang. Magally Bernal, Monica Devi, Leroy
Perumal, and Andre Vasquez were responsible for word processing.
Gemma Diaz from the Communications Department led the editorial team and managed the report’s produc-
tion with editorial assistance from Christine Ebrahimzadeh, David Einhorn, Lucy Scott Morales, Katy Whipple/
The Grauel Group, AGS, and Vector Talent Resources.
This issue of the GFSR draws in part on a series of discussions with banks, securities firms, asset management
companies, hedge funds, standard setters, financial consultants, pension funds, central banks, national treasuries,
and academic researchers.
This GFSR reflects information available as of September 29, 2020. The report benefited from comments and
suggestions from staff in other IMF departments, as well as from Executive Directors following their discussions of
the GFSR on September 30, 2020. However, the analysis and policy considerations are those of the contributing
staff and should not be attributed to the IMF, its Executive Directors, or their national authorities.
viii International Monetary Fund | October 2020

FOREWORD
T
he COVID-19 pandemic has triggered a evolution of the economy and the assessment of
global economic crisis of unprecedented risk in financial markets, reflecting in part investor
magnitude. The World Economic Outlook expectations of continued policy support. Corporate
forecasts a sharp global economic con- debt is rising, and it is estimated to be at record
traction for 2020. Despite the expected rebound levels relative to gross domestic product in most
in growth in 2021, the level of global output is countries. Despite the resilience exhibited so far,
anticipated to remain below precrisis levels for several there is a weak tail of fragile banks in some coun-
years. The swift, aggressive, and broad economic tries. Fragilities in the nonbank financial sector
policy response has contained the immediate damage, became clearly evident during the financial market
providing a bridge to recovery. Central banks have strains in March, with market volatility jumping,
eased monetary policy across the globe, with a nearly margin calls rising, and liquidity in even the most
$7.5 trillion balance sheet expansion to date in G10 liquid and deep bond markets drying up. Further-
countries, and with about 20 emerging market central more, sovereign debt is at historically high levels.
banks deploying asset purchases for the first time. The This is a critical issue for many low-income coun-
post-2008 regulatory framework has been put to the tries and some emerging market economies, where a
test for the first time, and has been proven largely suc- debt crisis might be inevitable without prompt and
cessful, as the global banking system entered the crisis decisive policy action—a theme that is explored at
with relatively high capital and liquidity buffers. In length in the Fiscal Monitor.
addition, a fiscal policy response of $12 trillion glob- Policymakers face stark trade-offs between short-
ally has provided substantial support to households term support and medium-term macro-financial
and firms. stability risks, and they need to closely monitor
As a result of these policy actions, the adverse any potential unintended consequences of their
macro-financial feedback loops that were so preva- unprecedented support. In the corporate sector,
lent and pernicious in the 2008 crisis have largely massive liquidity may lead to significantly higher
been contained. Financial conditions have eased debt and medium-term resource misallocation,
significantly and rapidly since late March, allowing potentially allowing insolvent firms to survive for
countries and firms to benefit from continued access years. For banks, the usage of buffers may lead to
to capital market and bank funding, and preventing too little capital being available in the future to
liquidity pressures from turning into broad-based cushion shocks. In capital markets, the easing of
insolvencies. Capital flows to emerging markets have financial conditions may fuel future vulnerabili-
started to rebound, with many economies regaining ties. For emerging markets, limited policy space
market access. While insolvency risks still loom large, can prevent optimal policies in the short and the
widespread corporate and banking distress has, to medium term. For many frontier economies and
date, been contained. In fact, the global banking sys- low-income countries—many of which continue to
tem remains fairly well capitalized against additional be shut out of international markets—further pan-
adverse shocks. demic pressure and the challenging global economic
But financial vulnerabilities are rising, putting environment are formidable headwinds for their
medium-term macro-financial stability and growth macro-financial stability.
at risk. Stretched valuations in risk asset markets In this Global Financial Stability Report, we take
persist, despite the September repricing in equity stock of key recent market developments and present
markets, giving rise to a disconnect between the a forward-looking analysis of banks, nonbank financial
International Monetary Fund | October 2020 ix

institutions, nonfinancial firms, and emerging mar- to gauge the extent to which the crisis may result in a
ket capital flows that can help policymakers navigate reversal of the gains posted in recent years. The analysis
difficult policy trade-offs in the next phases of the underscores the importance of climate policies and
pandemic and recovery. We also attempt to quantify green investment packages to support a green recovery
the impact of policies in our asset valuation assess- and the transition to a low-carbon economy.
ments, which can help policymakers better assess risks
to financial conditions. In addition, we assess the pan- Tobias Adrian
demic’s impact on firms’ environmental performance Financial Counsellor
x International Monetary Fund | October 2020

EXECUTIVE SUMMARY
Bridge to Recovery: October 2020 Global Financial Stability Report at a Glance

• Near-term global financial stability risks have been con- • While the global banking system is well capitalized, there
tained for now. Unprecedented and timely policy response is a weak tail of banks, and some banking systems may
has helped maintain the flow of credit to the economy and experience capital shortfalls in the October 2020 World
avoid adverse macro-financial feedback loops, creating a Economic Outlook adverse scenario even with the currently
bridge to recovery. deployed policy measures.
• However, vulnerabilities are rising, intensifying financial stabil- • Some emerging and frontier market economies face financ-
ity concerns in some countries. Vulnerabilities have increased ing challenges, which may tip some of them into debt
in the nonfinancial corporate sector, as firms have taken on distress or lead to financial instability and may require
more debt to cope with cash shortages, and in the sovereign official support.
sector, as fiscal deficits have widened to support the economy. • As economies reopen, accommodative policies will be
• As the crisis unfolds, corporate liquidity pressures may morph essential to ensure that the recovery takes hold and
into insolvencies, especially if the recovery is delayed. Small becomes sustainable—see the following Policy Road Map.
and medium-sized enterprises are more vulnerable than The post-pandemic financial reform agenda should focus
large firms with access to capital markets. The future path of on strengthening the regulatory framework for the non-
defaults will ultimately be shaped by the extent of continued bank financial sector and stepping up prudential supervi-
policy support and the pace of the recovery, which is expected sion to contain excessive risk taking in a lower-for-longer
to be uneven across sectors and countries. interest-rate environment.
Monetary and Financial Policy Road Map after the Great Lockdown
Gradual Reopening under Uncertainty
Monetary policy—Maintain accommodation to support the recovery

Liquidity support—Maintain support but adjust pricing to incentivize a gradual exit
Credit provision—Encourage banks to use capital and liquidity buffers to continue lending
Nonfinancial private sector—Extend moratoria on debt service only if necessary to prevent widespread insolvencies, support
viable firms through restructuring and efficient out-of-court workouts to reduce the debt burden, as well as by providing
solvency support (as appropriate)
Multilateral support—Provide support to emerging and frontier market economies facing financing difficulties
Pandemic under Control
Monetary policy—Maintain accommodation until monetary policy objectives are achieved

Liquidity support—Gradually withdraw
Credit provision—Require banks to gradually rebuild capital and liquidity buffers, develop credible plans to reduce problem
assets, and create markets for problem assets
Nonfinancial private sector—Recapitalize, restructure, or resolve nonviable firms
Green recovery—Encourage more proactive management of climate-related risks and green investments
Digitalization—Encourage greater digital investment to enhance financial sector efficiency and inclusion
Post-pandemic Financial Reform Agenda
Nonbank financial sector—Strengthen the regulatory framework to address vulnerabilities exposed during the coronavirus
disease (COVID-19) crisis
Lower for longer—Implement prudential measures to contain risk-taking in the lower-for-longer interest-rate environment
International Monetary Fund | October 2020 xi

Figure 1. Proportion of Systemically Important Countries Confronted with a global health and economic crisis,
with Elevated Vulnerabilities, by Sector
(Percent of countries with high and medium-high vulnerabilities, by GDP policymakers have taken extraordinary measures to
or assets; number of vulnerable countries in parentheses)
protect people, the economy, and the financial system.
Other financial 100% Sovereigns However, prospects for recovery remain highly uncertain
institutions (12)
More
(16) 80%
and will depend on the availability of reliable COVID-19
vulnerable 60%
treatments and vaccines. In addition, many countries have
40%
Asset
Nonfinancial
firms
entered the crisis with elevated preexisting vulnerabilities
20%
(21)
managers
(11) in some sectors—asset management, nonfinancial firms,
and sovereigns—and vulnerabilities are rising, represent-
ing potential headwinds for the recovery (Figure 1).
Insurers Households
Since the June 2020 Global Financial Stability
(12) (16)
Update, global financial conditions have remained
Oct. 2020 GFSR
Apr. 2020 GFSR Banks accommodative on the back of continued policy sup-
Global financial crisis (12)
port. In advanced economies, low interest rates and a
Sources: Bank for International Settlements; Haver Analytics; national authorities; recovery in risk asset markets have continued to sup-
Standard & Poor’s; WIND Information Co.; and IMF staff calculations.
Note: Based on 29 jurisdictions with systemically important financial sectors (see
Chapter 1 for details). “Global financial crisis” reflects the maximum 2007–08
port further easing in financial conditions (Figure 2).
vulnerability value. GFSR = Global Financial Stability Report. Financial conditions have generally eased also in
Figure 2. Key Drivers of Global Financial Conditions Indices emerging markets (excluding China) over the same
(Standard deviations from mean)
period, although external costs for many countries are
EM external costs
Interest rates
Corporate valuations
Index
House prices
still above pre–COVID-19 levels (Figure 2). In China,
2.5 United States Euro area Other
advanced
China Other
emerging
financial conditions have remained broadly stable, as
2.0 authorities have scaled back expectations for further
1.5
interest rate reductions amid improving economic activ-
ity and rising financial sector risks.
1.0
Although the sharp easing of financial conditions
0.5 since late March has helped prevent a financial crisis and
0.0
cushion the impact of COVID-19 on the economy, the
deterioration of the economic outlook has shifted the
–0.5
expected distribution of global growth in 2020 deeply
–1.0 into negative territory (Figure 3). Looking ahead, the
Dec.
Mar.
June
Sep.
Dec.
Mar.
June
Sep.
Dec.
Mar.
June
Sep.
Dec.
Mar.
June
Sep.
Dec.
Mar.
June
Sep.
global economy is expected to grow by 5.2 percent in

Sources: Bank for International Settlements; Bloomberg Finance L.P.; Haver
Analytics; IMF, International Financial Statistics database; and IMF staff 2021, according to the October 2020 World Economic
calculations.
Note: Higher number indicates a tightening of financial conditions. See Chapter 1 Outlook (WEO). This expected rebound and easy finan-
for details. EM = emerging market.
cial conditions imply that the odds of negative growth
Figure 3. Near-Term Global Growth Forecast Densities next year are low, though the balance of risks is tilted to
(Probability densities)
the downside (Figure 3).
0.35
Density for year
2021: at 2020:Q3
Unprecedented policy actions taken in response
Density for
0.30 year 2020: to the pandemic have been successful in boosting
Density for year at 2020:Q1
0.25
2020: at 2020:Q2 investor sentiment and maintaining the flow of credit
to the economy. To cope with cash flow pressures,
Probability density
0.20
Fifth percentiles firms have stepped up bond issuance, tapped bank
0.15 credit lines (most notably in the United States), and
0.10 taken advantage of government-guaranteed loans (see
Chapter 3).
0.05
Hard currency bond issuance in emerging markets
0.00
–12 –10 –8 –6 –4 –2 0 2 4 6 8 has been strong as well. Aggregate portfolio flows have
Global growth rate (percent) recovered from their March lows, though about half of
Sources: Bank for International Settlements; Bloomberg Finance L.P.; Haver
Analytics; IMF, International Financial Statistics database; and IMF staff
emerging market economies have continued to experi-
calculations.
Note: Forecast density estimates are centered around the respective World
ence outflows over the past three months. Easy financial
Economic Outlook forecasts for 2020 and 2021. Given the unprecedented nature
of the current crisis, model-based growth-at-risk estimates are inevitably subject conditions have improved the outlook for portfolio
to larger-than-usual uncertainty bounds.
flows to emerging markets, with the probability of
xii International Monetary Fund | October 2020
Executive Summary
outflows over the next three quarters falling from about Figure 4. Capital Flows at Risk: Near-term Forecasts of
Portfolio Flows
60 percent at the peak of market turmoil to 25 percent (Probability density function)
in September (Figure 4), though still above its pre– 0.30

March 23, 2020
COVID-19 level. September 29, 2020
0.25 Aggregate
Global equity markets have rebounded strongly from portfolio outflows 60% probability of
in 2020:Q1 an outflow
pandemic lows, with notable differentiation across coun- 0.20
tries depending on the spread of the virus, the scope of

0.15 25% probability of
policy support, and sectoral composition. Equity markets an outflow
in China and the United States have outperformed other 0.10
markets, driven by technology stocks (dark and light

0.05
green bars, Figure 5), notwithstanding the market correc-
tion in September. More contact-intensive sectors (hotels, 0.00
–6 –5 –4 –3 –2 –1 0 1 2 3 4 5 6 7 8 9 10
restaurants, leisure) have been hurt by lockdowns and Portfolio flows as a percent of GDP
social distancing. The underperformance of the energy Sources: Bloomberg Finance L.P.; Haver Analytics; IMF, World Economic Outlook
database; JP Morgan estimates; national sources; and IMF staff estimates.
and financial sectors (red and yellow bars, Figure 5) Note: Based on debt and equity portfolio flows for 19 largest emerging markets;
near term = next 3 quarters. See Chapter 1 for details.
reflects investors’ assessments of weaker growth prospects.
Figure 5. Stock Market Performance in 2020: Sectoral
The disconnect between rising market valuations Contributions
and the evolution of the economy, discussed in the (Percent, year to date)
June 2020 Global Financial Stability Update, persists. 20
For example, analysis of year-to-date US stock market 15
10
performance shows that a sharp decline in the corpo-
5 Information technology
rate earnings outlook has been more than offset by Health care
0 Financials
lower risk-free rates and a compression of the equity –5
Telecom
Consumer
risk premium, reflecting central banks’ policy rate –10
Industrials
Energy
Other
cuts and other measures that have boosted investor –15 Overall performance (YTD)
sentiment despite higher economic uncertainty (see –20
Chapter 1). Similarly, the decline in corporate bond –25
–30
yields has been driven by the fall in risk-free rates and US Japan Euro UK China India Brazil
area
the compression in credit spreads—in many cases
Sources: Bloomberg Finance L.P.; MSCI; and IMF staff calculations.
below values estimated to be consistent with economic Note: All country indices are local currency MSCI sub-indices. Overall performance
is based on aggregation of sectoral indices. “Consumer” is the sum of the
fundamentals (Figure 6). The spread compression can consumer discretionary and consumer staples sectors and “other” is the sum of
the utilities, materials, and real estate sectors. UK = United Kingdom; US = United
be partly attributed to policy support and, in the case States; YTD = year to date.
of emerging markets, it can also be traced to policy Figure 6. Bond Spread Misalignment
(Deviation from fair value per unit of risk, left scale; percentile based on
easing by central banks in advanced economies. If mar- 1995–2020, right scale)
kets believe that policy support will be maintained or 6 100
Misalignment
scaled up in response to deterioration in the economic Percentile (right scale)
4
outlook, current risk asset valuations could be sus- 80
tained for some time. However, if investors reassess the 2
scope for policy support or if the recovery is delayed, 0

60
the odds of a sharp adjustment may rise. –2 40

Nonfinancial firms have come under significant liquid-
–4
ity strains following the COVID-19 outbreak. More 20
vulnerable firms—with weaker solvency and liquidity –6
positions, as well as smaller firms—have experienced –8 0

Q4
Q1
Q3 to date
Q4
Q1
Q3 to date
Q4
Q1
Q3 to date
Q4
Q1
Q3 to date
Q4
Q1
Q3 to date
Q4
Q1
Q3 to date
Q4
Q1
Q3 to date
Q4
Q1
Q3 to date
greater financial stress than their peers in the early stages

of the crisis (see Chapter 3). To cope with cash shortages, IG HY IG HY IG HY IG HY
many firms—notably those whose earnings fell short of United States Euro area EM hard
currency
EM local
currency
their interest expenses—have increased their borrowing Sources: Bloomberg Finance L.P.; Consensus Economics; Haver Analytics;
(Figure 7), adding to the already high corporate debt Refinitiv I/B/E/S; and IMF staff calculations.
Note: Misalignment is the difference between market- and model-based values
levels in several economies (Figure 8). Default rates have scaled by the standard deviation of monthly changes in spreads; negative values
on the left scale indicate overvaluation. See Chapter 1 for details. EM = emerging
market; HY = high yield; IG = investment grade.
been on the rise as well. As the crisis continues to unfold,
International Monetary Fund | October 2020 xiii
Figure 7. Publicly Listed Firms: Debt at Risk and especially if a sustainable recovery is delayed, liquidity
(Percent of debt of sample firms)
pressures may morph into insolvencies.
ICR < 1 in 2019:Q4
ICR < 1 in 2020:Q2
ICR < 1 in 2020:Q2 and an increase in net debt between Q4 and Q2
Barring a significant tightening in funding conditions,
50 large firms with access to capital markets are likely to
avoid significant solvency pressures. Firms in sectors most
40
affected by the pandemic, however, are facing weaker
30
growth prospects and greater liquidity strains, and hence a
higher risk of default and insolvency. Small and medium-
20 sized enterprises, which are generally more vulnerable,
could be a significant channel for transmission of the
10 economic shock. Furthermore, small and medium-sized
enterprises tend to dominate some of the most contact-
0
intensive sectors (hotels, restaurants, entertainment),
North America
Euro area
Other advanced
Latin America
EMEA
Emerging Asia
which have taken a beating from COVID-19.

Banks entered the COVID-19 crisis with signifi-
Sources: Bank for International Settlements; Bloomberg L.P.; Haver Analytics; cantly stronger capital and liquidity buffers than they
Institute of International Finance; S&P Global Ratings; S&P Leveraged Commentary
and Data; and IMF staff calculations.
Note: The sample includes firms with quarterly statements. The bars show the
had in 2008–09. This has allowed them to continue
share of debt at firms with ICR < 1 and with an increase in net debt as a share of
total debt in the sample. EMEA = Europe, Middle East, and Africa; ICR = interest
to provide credit to the economy. Policies aimed at
coverage ratio.
supporting borrowers and encouraging banks to use
Figure 8. Aggregate Nonfinancial Corporate Debt
the flexibility built into the regulatory framework have
(Percent of GDP) likely supported banks’ willingness and ability to lend.
180 Range over past 10 years 2019:Q4 2020:Q1 However, some banks are already starting to tighten
150
their lending standards, which could have adverse
implications for the recovery. A forward-looking
120
analysis of bank solvency in 29 countries (not includ-
90
ing China) shows that in the October 2020 WEO
baseline scenario most banks will be able to absorb
60
losses and maintain capital buffers above the minimum
30 capital requirements (see Chapter 4). In the WEO
adverse scenario characterized by a deeper recession
0
and a weaker recovery, a sizable weak tail of banks
ESP
POL
CAN
JPN
USA
DEU
CHN
RUS
TUR
IND
MEX
FRA
GBR
ITA
BRA
could see their capital buffers depleted to the levels

Sources: Bank for International Settlements; Haver Analytics; and IMF staff
calculations.
Note: For France, corporate debt is reported on an unconsolidated basis. Data
that could constrain their lending capacity (Figure 9).
labels use International Organization for Standardization (ISO) country codes.
These weak banks’ capital shortfall relative to broad
Figure 9. Distribution of Bank Assets by Capital Ratio under
regulatory requirements—which include the counter-
the October 2020 WEO Adverse Scenario, with Policy cyclical capital buffer, capital conservation buffer, and
Mitigation
(CET1 ratio, percent) systemic buffers—could reach $220 billion, even after
100
< 4.5% < 6% < 8% < 10% < 12% ≥ 12% accounting for borrower- and bank-oriented mitigation
policies (see Chapter 4).
80 Nonbank financial institutions (NBFIs) have entered
the crisis with elevated vulnerabilities (Figure 10).
60
They have managed to cope with the pandemic-
induced market turmoil thanks to policy support, but
40
fragilities remain high. Asset managers, for example,
20 could be forced into fire sales if portfolio losses are
larger and redemptions last longer than during the
0
19 T 22 19 T 22 19 T 22 19 T 22 19 T 22
March sell-off. NBFIs play a growing role in credit
Global AEs EMs GSIB Non-GSIB markets, including riskier segments, and the increased
Sources: Bloomberg Finance L.P.; Fitch; IMF, World Economic Outlook database;
and IMF staff estimates.
links between NBFIs and banks imply that fragilities
Note: The scenario takes into account mitigation policies (see Chapter 4 for details).
AEs = advanced economies; CET1 = common equity Tier 1; EMs = emerging could spread through the financial system.
markets; GSIB = global systemically important bank; T = trough year.
xiv International Monetary Fund | October 2020

Executive Summary
Sovereign vulnerabilities have increased because Figure 10. Nonbank Financial Institutions: Financial
Vulnerability Indices and Sector Size
countries have expanded fiscal support, and sovereigns
may face a sharp rise in contingent liabilities. Vulner- China
1.0
United States Other AEs Euro area Other EMs
abilities have increased across multiple sectors, with 0.9 OFIs
Financial vulnerability indices (percentile score)

6 out of 29 jurisdictions with systemically important 0.8
OFIs AMs
financial sectors now showing elevated vulnerabili- 0.7
ties in the corporate, banking, and sovereign sectors 0.6
(Figure 11). 0.5
Because of the pandemic, the financing needs of 0.4
AMs
AMs
AMs
OFIs
OFIs
emerging markets have risen sharply. Concerns about 0.3
0.2
new debt supply and weak domestic fundamentals may
AMs
OFIs
0.1
have curtailed demand for local currency bonds from
0.0
foreign investors (Figure 12), especially where they 0 10 20 30 40 50 60 70 80 90 100 110
Sector size (trillions of US dollars)
hold large shares of debt and where the domestic inves-
Sources: Banco de Mexico; European Central Bank; Haver Analytics; Reserve Bank
tor base may not be sufficiently deep. Some emerging of India; Securities and Exchange Commission of Brazil; WIND Information Co.; and
IMF staff calculations.
market central banks purchased a substantial share Note: See Chapter 1 for details. AEs = advanced economies; AMs = asset
managers; EMs = emerging markets; OFIs = other financial institutions.
of bonds in the secondary market to stabilize market
conditions (see Chapter 2). Frontier market economies Figure 11. Corporate, Bank, and Sovereign Vulnerabilities
in 29 Jurisdictions with Systemically Important Financial
face even greater financing challenges, as the COVID- Sectors
19 shock pushed borrowing costs for many to prohibi-
medium-high — high
tive levels—calling for official support.
As policymakers build a bridge to recovery, poli-
cies will have to adjust, depending on the evolution of
the pandemic and the pace of the economic rebound
(see Policy Road Map in the at-a-glance box at the
beginning of this Executive Summary). At each step,
Higher bank vulnerabilities
low — medium-low
policymakers should be mindful of intertemporal

trade-offs and of unintended consequences—the ben-
efits of using available buffers today should be carefully
balanced against the possible need for further support low — medium-low medium-high — high
in the future, as well as the risk of exacerbating future Higher corporate vulnerabilities
vulnerabilities. Sources: Bank for International Settlements; Haver Analytics; Institute of

International Finance; IMF, October 2020 World Economic Outlook; and IMF staff
As economies reopen, continued monetary policy estimates.
Note: Based on the data underlying Figure 1; red dots denote countries with
medium-high or high sovereign vulnerabilities.
accommodation and targeted liquidity support will
be essential for sustaining the recovery. A robust Figure 12. Change in Local Currency Government Bonds
Outstanding by Holder, end-February–June 2020
framework for debt restructuring will be critical for (Percent of GDP)
reducing debt overhangs and resolving nonviable 7
Other domestic
firms. Low-income countries with financing difficul- 6
Nonresidents
Domestic banks
ties may require multilateral support. Despite its 5
Central bank
Total change in LC
adverse effect on firms’ environmental performance, 4
debt outstanding
the COVID-19 crisis also presents an opportunity 3
to engineer a transition to a greener economy (see 2
Chapter 5). 1
After the pandemic is fully under control, policy 0
support can be gradually withdrawn and policy priori- –1
ties should focus on rebuilding bank buffers, strength- –2
ening regulation of nonbank financial institutions –3

POL ZAF IDN ROU HUN BRA COL MEX
and stepping up prudential supervision to contain Sources: Bloomberg Finance L.P.; Haver Analytics; IMF, World Economic Outlook
excessive risk taking in a lower-for-longer interest-rate database; national sources; and IMF staff estimates.
Note: Data are not adjusted for inflation-linked debt. South Africa total differs
slightly from aggregated component changes. Indonesia central bank holdings of
environment. government securities reported as net of monetary operations by source. Data
labels use International Organization for Standardization (ISO) country codes.
LC = local currency.
International Monetary Fund | October 2020 xv

IMF EXECUTIVE BOARD DISCUSSION OF THE OUTLOOK,
OCTOBER 2020
The following remarks were made by the Chair at the conclusion of the Executive Board’s discussion of the
Fiscal Monitor, Global Financial Stability Report, and World Economic Outlook on September 30, 2020.
E
xecutive Directors broadly concurred with the setbacks to human capital accumulation. Most Direc-
assessment of the global economic outlook, tors also saw the crisis as an opportunity to stimulate
risks, and policy priorities. While noticing the innovation, develop the digital infrastructure, and to
stronger-than-expected economic activity in transition to lower carbon emissions using different
the second quarter, especially in advanced economies, climate tools, such as green investment and a gradual
they agreed that the path to prepandemic activity will increase of the carbon price, with due consideration to
be long and precarious with persistent scarring effects offsetting negative social impact.
on output and employment. They noted that the Directors welcomed the unprecedented fiscal actions
projections assume that social distancing will con- in response to the pandemic. Directors emphasized
tinue into 2021 and then fade over time as therapies that, as economies tentatively reopen, governments
improve and vaccines become more broadly available. should ensure that lifelines are not withdrawn prema-
Directors noted with concern that the pandemic is turely. Support should gradually shift from protect-
having dramatic effects on vulnerable people, leading ing jobs to helping displaced workers find new jobs
to higher inequality, and a sharp increase in the num- through retraining and reskilling. Directors noted that
ber of people living in extreme poverty. when the pandemic is under control, governments
Directors agreed that the uncertainty surrounding will need to address the legacies of the crisis, including
the baseline projections remains exceptionally large record deficits and public debt levels, elevated unem-
as the economic recovery will be shaped primarily ployment, and increased poverty. Directors agreed
by the path of the pandemic, the efficacy of contain- that public investment should play a crucial role in
ment measures, and pharmaceutical innovations. supporting the postpandemic recovery, noted its siz-
More rapid development of new therapeutics and able job creation potential, and underlined that good
wide distribution of effective vaccines could acceler- governance, budget execution, and communication,
ate the economic recovery, while medical setbacks remain crucial to reap the full benefits of fiscal support
and new waves of infections could require new and maintain public trust.
lockdowns. Other important sources of uncertainty Directors emphasized that governments will need
include the extent of global spillovers, the damage to do more with less and prepare credible and equi-
to the supply potential, the efficacy and duration table measures to reduce fiscal deficits and debts over
of policy support, and potential shifts in financial the medium term. Countries with limited fiscal space
market sentiment. Directors also noted prepandemic should protect public investment and support lower-
risks stemming from trade and technology tensions, income households that have been disproportionately
geopolitical challenges, and climate change. hit by the pandemic. Governments could consider
Directors agreed that effective and decisive policy increasing progressive taxation as well as reforms to
support is needed to ensure stronger, more equitable, modernize business taxation, including multilateral
and resilient growth. Key near-term priorities include cooperation on the design of international corporate
supporting the economic recovery, protecting vulner- taxation to respond to the challenges of the digital
able people, and strengthening health care systems. economy. LICs in particular are faced with significant
They stressed the need to reduce the scarring effects of financing constraints, and many countries will require
the crisis on potential output and employment and to external support, including in the form of debt relief,
reverse the development toward greater inequality and grants, and concessional financing.
xvi International Monetary Fund | October 2020

IMF executive board discussion of the outlook, October 2020
Directors agreed that bold policy actions taken by policy support can be gradually withdrawn. The
central banks to ease monetary policy, provide ample postpandemic financial reform agenda should focus
liquidity, and maintain the flow of credit have helped on strengthening the regulatory framework to address
contain the near-term risks to global financial stabil- vulnerabilities in the nonbank financial sector exposed
ity. They noted, however, that vulnerabilities are rising, by the crisis and stepping up prudential supervision
most notably in the nonfinancial corporate sector to contain excessive risk taking in the lower-for-longer
as liquidity pressures may morph into insolvencies, interest rate environment.
especially for small and medium-sized enterprises. Directors underscored the importance of inter-
The credit outlook will ultimately be shaped by the national cooperation in the fight against the global
extent of continued policy support and the pace of the health and economic crisis. A key priority is to scale up
recovery, which is expected to be uneven across sectors production capacity and develop distribution channels
and countries. Rising defaults could lead to significant to ensure that all countries have access to an effective,
losses at banks and nonbank financial institutions. affordable, and safe vaccine. Directors noted that sev-
While the global banking system is overall well capital- eral emerging market and developing countries require
ized, some banks and banking systems may experi- international assistance through debt relief, grants, and
ence aggregate capital shortfalls in the WEO adverse concessional financing. They pointed out that the IMF
scenario. Directors also highlighted the importance has rapidly scaled up its lending facilities since the
of improving access of emerging markets and frontier onset of the pandemic, providing swift financial assis-
economies to capital markets. tance to more than 80 countries. Directors discussed
Directors emphasized that as economies reopen, opportunities for multilateral cooperation to alleviate
accommodative policies and the continued flow of trade and technology tensions between countries and
credit to borrowers will be essential to sustaining to collectively implement climate change mitigation
the recovery. Once the pandemic is under control, policies.
International Monetary Fund | October 2020 xvii

1
CHAPTER
GLOBAL FINANCIAL STABILITY OVERVIEW
BRIDGE TO RECOVERY
Chapter 1 at a Glance
•• Near-term global financial stability risks have been contained for now. Unprecedented and timely
policy response has helped maintain the flow of credit to the economy and avoid adverse macro-financial
feedback loops, creating a bridge to recovery.
•• However, vulnerabilities are rising, intensifying financial stability concerns in some countries. Vulnera-
bilities have increased in the nonfinancial corporate sector as firms have taken on more debt to cope with
cash shortages and in the sovereign sector as fiscal deficits have widened to support the economy.
•• As the crisis unfolds, corporate liquidity pressures may morph into insolvencies, especially if the
recovery is delayed. Small and medium-sized enterprises (SMEs) are more vulnerable than large firms with
access to capital markets. The future path of defaults will be shaped by the extent of continued policy
support and the pace of the recovery, which may be uneven across sectors and countries.
•• While the global banking system is well capitalized, there is a weak tail of banks, and some banking
systems may experience capital shortfalls in the October 2020 World Economic Outlook adverse scenario
even with the currently deployed policy measures.
•• Some emerging and frontier market economies face financing challenges, which may tip some of them
into debt distress or lead to financial instability, and may require official support.
•• As economies reopen, continued policy support remains critical. Accommodative monetary and
financial conditions, credit availability, and targeted solvency support will be essential to sustaining the
recovery, facilitating the necessary structural transformation and transition to a greener economy.
•• The post-pandemic financial reform agenda should focus on addressing fragilities unmasked by the
coronavirus disease (COVID-19) crisis, strengthening the regulatory framework for the nonbank finan-
cial sector and stepping up prudential supervision to contain excessive risk taking in a lower-for-longer
interest-rate environment.
Outlook (WEO). Both advanced and emerging market

The COVID-19 Pandemic Has Led to a economies will suffer deep and broad-based declines,
Deep Recession with more than 85 percent of countries around the
The COVID-19 pandemic has led to an unprec- world expected to see subzero growth this year (red
edented contraction in economic activity globally, shaded area in Figure 1.1). Confronted with a global
with global growth projected at –4.4 percent this health and economic crisis, policymakers have taken
year, according to the October 2020 World Economic extraordinary measures to protect people, the econ-
omy, and the financial system. Despite forceful policy
Prepared by staff from the Monetary and Capital Markets action, however, the prospects for recovery remain
Department (in consultation with other departments): The authors highly uncertain.
of this chapter are Anna Ilyina (Division Chief ), Evan Papageorgiou
The October 2020 WEO baseline global growth
(Deputy Division Chief ), Sergei Antoshin, Yingyuan Chen, Fabio
Cortes, Rohit Goel, Phakawa Jeasakul, Sanjay Hazarika, Kelly forecast of +5.2 percent for 2021 assumes that
Eckhold, Frank Hespeler, Henry Hoyle, Piyusha Khot, Sheheryar continued unprecedented monetary policy accom-
Malik, Thomas Piontek, Akihiko Yokoyama, and Xingmi Zheng, modation and large fiscal lifelines will keep financial
under the guidance of Fabio Natalucci (Deputy Director). Magally
Bernal and Andre Vasquez were responsible for word processing and conditions easy and help offset COVID-19–related
the production of this report. cash flow pressures on firms and households, thus
International Monetary Fund | October 2020 1

GLOBAL FINANCIAL STABILITY REPORT: Bridge to Recovery
Figure 1.1. GDP Growth: The COVID-19 Crisis versus the the worsening in the WEO baseline forecast between
Global Financial Crisis April and June.1
GDP Growth Rates – Average and Distribution
Looking ahead, current economic and financial
(Share of countries [left scale]; GDP growth in percent [right scale]) conditions, combined with the expected rebound
< –3% –3 to 0% 0 to 1% 1 to 2% 2 to 3% of 5.2 percent in global GDP growth next year,
3 to 4% 4 to 5% > 5% Global growth (right scale) imply that the 2021 growth forecast distribution will
0% 8
shift back into positive territory (shown in green in
Figure 1.2, panel 1). Nonetheless, the shape of the
5.4 5.5 5.5 5.4 6
5.2 2021 growth distribution suggests that there are still
20%
4.3
4.9 3.8 significant downside risks. For example, the probability
3.5 3.4 3.8 3.5 4.2 4
3.0 3.5 3.5 3.3 3.6 3.5 of global growth falling below zero in 2021 is close to
40% 2.8 5 percent, indicating that risks are elevated by histori-
2
cal standards (Figure 1.2, panel 2).
0
Several possible developments could delay the recov-
60% –0.1 ery and lead to worse-than-expected growth outcomes,
–2
putting financial stability at risk. A resurgence of the
virus in some countries may require partial lockdowns
80%
–4 and more prolonged social distancing, leading to job
–4.4
losses and renewed pressures on corporate and financial
100% –6 sector balance sheets (see the WEO Scenario Box).
2004 06 08 10 12 14 16 18 20 22 24 Policy missteps, such as a premature withdrawal of pol-
icy support (as discussed in the October 2020 WEO),
Sources: IMF, October 2020 World Economic Outlook; and IMF staff calculations.
could trigger investor reassessment of risks, market
turbulence, and tightening of financial conditions. For
example, market participants have been increasingly
attuned to the progress on Brexit negotiations given
keeping insolvencies at bay. Nevertheless, some vul- the looming deadline, a development that could lead
nerable firms (such as SMEs) and sectors (notably the to increased market volatility.
contact-intensive sectors) will experience greater dis-
tress. Furthermore, if the recovery were delayed, liquid-
ity pressures could reemerge and insolvencies could rise Unprecedented Policy Support Has
sharply and become more widespread. Such an adverse Helped Buy Time
scenario would entail repricing of risk in credit markets Unprecedented policy actions taken in response
and a tightening of financial conditions—ultimately to the pandemic have been successful in boosting
testing the resilience of the financial system, as well as investor sentiment and maintaining the flow of
the capacity of country authorities to provide addi- credit to the economy. Central banks’ interventions
tional policy support. have stabilized key markets by lifting investor risk
The deterioration of the global economic outlook appetite through both anticipated and actual central
early in the year shifted the expected distribution of bank demand for safe and risk assets (Figure 1.3).
global growth in 2020 deeply into negative territory
(Figure 1.2, panel 1). Besides changes in the WEO 1The growth-at-risk framework assesses the downside risks to
baseline global growth forecast, around which these financial stability by gauging how the range of severely adverse
distributions are centered, these shifts reflect changes growth outcomes (5th percentile of the growth distribution) shifts
in response to changes in financial conditions and vulnerabilities
in financial conditions, and hence are heavily influ-
(see Chapter 3 of the October 2017 GFSR for details). Assump-
enced by investor perceptions and assessment of future tions pertaining to policy responses or macroeconomic shocks are
growth outcomes. The massive easing of financial captured in the growth-at-risk framework to the extent that they
conditions (discussed in the June 2020 Global Finan- affect the current economic and financial conditions, or the baseline
growth forecast. Given the unprecedented nature of the current
cial Stability Report [GFSR] Update) has helped contain crisis, model-based growth-at-risk estimates are inevitably subject to
downside risks to growth and financial stability despite larger-than-usual uncertainty bounds.
2 International Monetary Fund | October 2020

CHAPTER 1 GLOBAL FINANCIAL STABILITY OVERVIEW: Bridge to Recovery
Figure 1.2. Global Growth at Risk

The unprecedented policy support helped reduce the downside risks ... risks are expected to remain tilted to the downside and within the
to growth and financial stability, but even with growth projected to danger zone.
rebound next year ...
1. Near-Term Growth Forecast Densities 2. Near-Term Growth-at-Risk Forecasts
(Probability density) (Percentile rank)
0.35 Quintiles
Density for year
2021: at 2020:Q3
Density for Worst Best
0.30 Density for year year 2020: 100
2020: at 2020:Q2 at 2020:Q1
0.25
80
Probability density
0.20
Fifth percentiles 60
0.15
40
0.10
20
0.05
0 0
–12 –10 –8 –6 –4 –2 0 2 4 6 8 2008 09 10 11 12 13 14 15 16 17 18 19 20
Global growth rate (percent)
Sources: Bank for International Settlements; Bloomberg Finance L.P.; Haver Analytics; IMF, International Financial Statistics database; and IMF staff calculations.
Note: Forecast density estimates are centered around the respective World Economic Outlook forecasts for 2020 and 2021. In panel 2, the black line traces the
evolution of the 5th percentile threshold (growth-at-risk) of near-term growth forecast densities. The color of the shading depicts the percentile rank for the
growth-at-risk metric, from 1991 onwards. See the April 2018 Global Financial Stability Report for details.
Many emerging market central banks have, for the easing in financial conditions (Figure 1.4, panel 2).
first time, engaged in asset purchases to stabilize With nominal yields already at low levels, central bank
their local currency bond markets or to ease domes- measures have driven real yields down to historic lows.
tic financial conditions (see Chapter 2). Unprece- Market-implied inflation expectations for the near to
dented policy support has been a game changer—it medium term have recovered since the March sell-off but
has lessened risks to financial stability and bought remain slightly below pre–COVID-19 levels (see Online
time for country authorities to take steps to address Annex 1.1).2 In other emerging markets (excluding
the health crisis and contain its economic fallout. China), financial conditions have generally eased since
However, these policy measures may have unin- June (Figure 1.4, panels 3 and 4), more so in emerging
tended consequences, for example, by contributing market economies in Asia and Latin America than in
to stretched asset valuations or fueling financial those in Europe, the Middle East, and Africa. External
vulnerabilities (see subsequent sections), especially spreads for many emerging markets remain above the
if these policies remain in place for an extended pre–COVID-19 levels, reflecting a deterioration in
period of time and investors become used to them. domestic economic activity.3
These considerations should be taken into account
as central banks plan for the eventual withdrawal of
2While the decline in real yields has mechanically pushed up infla-
support (see the policy section).
tion breakevens (given stable nominal yields), this appears to have
Since the June 2020 GFSR Update, global financial been driven in part by liquidity and technical factors.
conditions have remained accommodative on the back 3IMF staff analysis, using the fundamentals-based JP Morgan
of continued policy support (Figure 1.4, panel 1). In Emerging Market Bond Index Global model, shows that the key
driver of widening of spreads in 2020 has been the deterioration
advanced economies, low interest rates and a recovery in domestic factors, following the deep and sudden recession in
in risk asset markets have continued to support further most economies.

Figure 1.3. Central Bank Measures in Major Advanced Economies—Game Changer

(Index, left scale; number of policy announcements, right scale)
Central bank actions were forceful and swift, and targeted a range of key markets using an array of policy tools.
120 50
VIX (left scale) MOVE (left scale) United States United Kingdom
Canada ECB Japan Australia
45
100
40
35
80
30
60 25
20
40
15
10
20
5
0 0
OM
FX
GB
CB
QG
PR
OM
FX
GB
CB
QG
PR
OM
FX
GB
CB
QG
PR
OM
FX
GB
CB
QG
PR
OM
FX
CB
GB
QG
PR
OM
FX
GB
CB
QG
PR
OM
FX
GB
CB
QG
PR
Mar. 2–Mar. 15 Mar. 16–Mar. 29 Mar. 30–Apr. 12 Apr. 13–May 10 May 11–June 7 June 8–July 5 July 6–Aug. 2
Sources: Bloomberg Finance L.P.; central bank websites; Haver Analytics; and IMF staff calculations.
Note: Intervention types refer to expansion/enhancement of OMs, FX, GBs, CBs, QGs, and PRs. Each dot refers to an announced enhancement or new operation or
facility. The policy intervention types correspond to the economic nature of the interventions undertaken, even though in some cases the technical mechanism varies.
CB = commercial paper, asset-backed securities, and corporate bond purchases; ECB = European Central Bank; FX = foreign exchange swap lines and foreign
exchange lending operations; GB = government securities purchase; MOVE = Merrill Lynch Option Volatility Estimate; OM = open market operation, collateral
framework, and standing liquidity facility; PR = reduced policy rate; QG = purchase of quasi government or government-guaranteed/-supported securities;
VIX = Chicago Board Options Exchange Volatility Index.
In China, financial conditions have remained The Pandemic Has Hit Some Economic Sectors
broadly stable over the summer (Figure 1.4, Harder than Others
panels 1 and 2). After initially cutting policy rates Behind the broad rebound in risk asset prices there
and deploying measures to directly increase bank are clear signs of differentiation across sectors. Some
credit, authorities in May scaled back expectations sectors (such as airlines, hotels, energy, and financials)
for further interest rate reductions, leading to a have been more affected by the lockdown and social
rebound in bond and money market yields (Fig- distancing, whereas those that are less contact-intensive
ure 1.4, panels 1 and 2). The policy shift came (information technology, communications) have been
amid improving economic activity but also concerns faring better. Equity market indices with a larger share
about rising financial sector risks. Rapid increases in of sectors less affected by COVID-19 have seen a
risky asset management product borrowing contrib- stronger rebound (Figure 1.5, panel 1).
uted to large swings in interest rates, whereas most Market analysts’ earnings forecasts may provide
banks saw limited pass-through from policy rates an indication of the likely pace of recovery from
to funding costs, posing risks to bank profitability the pandemic across sectors and countries. Certain
(see Online Annex Box 2.1). Other People’s Bank of sectors—notably consumer services (hotels, restaurants,
China measures have helped direct credit to vulner- leisure), industrials (capital goods), and financials
able borrowers and support the economy, but these (banks)—have seen large swing in their 2020–21
may be adding to nonfinancial sector vulnerabilities earnings per share forecasts, the large dispersion of
(Figure 1.9, panel 2). forecasts across analysts, and significant downgrades

Figure 1.4. Global Financial Conditions
Global financial conditions have eased further since the June 2020 ... on the back of a continued decline in interest rates and recovery in
GFSR Update ... risk asset markets.
1. Global Financial Conditions Indices 2. Key Drivers of Global Financial Conditions Indices
(Standard deviations from mean) (Standard deviations from mean)
EM external costs Corporate valuations House prices
Interest rates Index
United Euro Other China Other
6 States area advanced emerging 2.5
United April
5 States 2020 2.0
Tightening GFSR
4
1.5
3
Euro Other
2 area China advanced 1.0
1 economies 0.5
0
0.0
–1 Other emerging
–2 market economies –0.5
–3 –1.0
2007 08 09 10 11 12 13 14 15 16 17 18 19 20
Dec.
Mar.
June
Sep.
Dec.
Mar.
June
Sep.
Dec.
Mar.
June
Sep.
Dec.
Mar.
June
Sep.
Dec.
Mar.
June
Sep.
In emerging market economies, financial conditions have eased as External funding costs have declined but remain elevated relative to
well. pre-COVID-19 levels.
3. Financial Conditions Indices for Emerging Market Regions 4. Key Drivers of Emerging Market Financial Conditions Indices
(Standard deviations from mean) (Standard deviations from mean)
EM external costs Corporate valuations House prices
Interest rates Index
4 1.5
April
Asia, 2020
3 excluding Tightening 1.0
GFSR
China Europe,
2
Middle East, 0.5
1 and Africa
0.0
0
Asia, Europe,
–1 –0.5
excluding Latin America Middle East,
Latin America China and Africa
–2 –1.0
2007 08 09 10 11 12 13 14 15 16 17 18 19 20
Dec.
Mar.
June
Sep.
Dec.
Mar.
June
Sep.
Dec.
Mar.
June
Sources: Bank for International Settlements; Bloomberg Finance L.P.; Haver Analytics; IMF, International Financial Statistics database; and IMF staff calculations. Sep.
Note: Panels 1 and 3 show quarterly averages for 2007–2019 and monthly averages for 2020; panels 2 and 4 show monthly averages. In panels 2 and 4, the interest
rate component contains real short-term interest rates, term spreads or medium-term interest rates, and interbank spreads. See the April 2018 Global Financial
Stability Report (GFSR) for details. EM = emerging market.
of long-term earnings per share growth forecasts since sectors means that some countries may recover faster
the outbreak (Figure 1.5, panel 2). The downward than others.
revisions for financials likely reflect the subdued
growth outlook and low interest rates. Furthermore,
banks in major economies have significant exposure to Risk Assets Have Rebounded despite High
commercial real estate, which has been hit particularly Economic Uncertainty
hard by the pandemic as the shift to working remotely The disconnect between rising market valuations
has sharply reduced demand for commercial properties and weak economic activity, discussed in the June
(see Box 1.1). The differential global recovery across 2020 GFSR Update, has persisted notwithstanding

Figure 1.5. Global Equity Markets: Impact of COVID-19 on Countries and Sectors
Countries and regions with a higher share of less contact-intensive Some sectors (such as consumer services, industrials, and financials)
sectors (such as information technology and telecommunications) have have seen large fluctuations in their near-term forecasts as well as
done better, whereas energy and financial stocks have been a drag on notable downward revisions of the long-term earnings per share
stock market performance. forecasts.
1. Stock Market Performance in 2020: Sectoral Contributions 2. Long-term EPS Growth Forecasts: United States, Euro Area, and Japan
(Percent, year to date) (Percent, simple average)
20 20
Consumer Retail
15 durables
Long-term EPS growth forecast (current)

10 Insurance Biotech 15
Information Software
technology
5 Tech hardware
Health care
Financials Health care
0 10
Telecom
Consumer Materials
–5 Capital goods
Industrials
Energy Telecom (industrials)
–10 Households 5
Other
Overall Food
–15
performance Banks
(YTD) Energy
–20 0
Automobiles
–25 Consumer services
–30 –5
US Japan Euro UK China India Brazil 0 5 10 15 20
area Long-term EPS growth forecast (2019)
Sources: Bloomberg Finance L.P.; MSCI; Refinitiv I/B/E/S; and IMF staff calculations.
Note: In panel 1, all country indices are local currency MSCI sub-indices. Overall performance is based on aggregation of sectoral indices. “Consumer” is the sum of
the consumer discretionary and consumer staples sectors and “other” is the sum of the utilities, materials, and real estate sectors. In panel 2, red dots denote the
largest downward forecast revisions. Long-term forecasts cover three- to five-year horizon. All indices are national benchmark indices by sector. UK = United
Kingdom; US = United States; YTD = year to date.
the September correction in equity markets. Despite Factors such as the sectoral composition, investor base,
subdued activity and a highly uncertain outlook, global and other technical factors have also played a role in driving
equity markets have rebounded from the March lows, equity valuations.4 For example, US stock market perfor-
though with notable differentiation across countries, mance has been boosted by a large share of tech firms in
depending on the spread of the virus, the scope of pol- the S&P 500 index, as the pandemic has had pronounced
icy support, and sectoral composition (see Figure 1.6, implications for work and consumption behavior that
panels 1 and 2). are expected to encourage spending on new technologies
The stock market recovery has been largely driven (Figure 1.6, panel 4). Despite the September sell-off, five
by policy support. A simple decomposition of the S&P tech giants have significantly outperformed the rest of
500 year-to-date performance into the contributions the index since June 2020, benefiting from their business
of three factors—earnings (current and projected), the models and diversified business revenues (Figure 1.6,
risk-free rate, and the equity risk premium—shows panel 5).5 In addition, in some countries, retail investors,
that a sharp deterioration in the corporate earnings who tend to chase growth and technology stocks, have
outlook has contributed negatively to stock market
performance (Figure 1.6, panel 3). But such a negative
4For example, the US stock market is dominated by sectors and
contribution has been more than offset by a lower
large firms that have been less affected by the pandemic than the
risk-free rate (green bars) and a compression of the broader economy. SMEs, which are not publicly listed but play an
equity risk premium (shown as a positive contribu- important role in the economy, could also account for some of this
tion in gray), reflecting the Federal Reserve’s policy disconnect between stock market and the broader economy.
5The top five S&P stocks by market cap (AAPL, AMZN,
rate cuts and other policy measures that have boosted GOOG, FB, MSFT) account for about 25 percent of total market
risk sentiment. capitalization.

Figure 1.6. Equity Market Valuations
Markets rebounded on strong policy support, but with clear differentiation across countries and sectors.
1. Global Equity Markets: Countries and Regions 2. Global Equity Markets: Economic Sectors
(Percent) (Percent)
Trough 2020–current End 2019–trough YTD Trough 2020–current End 2019–trough YTD
China Information technology

US Consumer discretionary
Japan Telecom
India Health care
Euro area Materials
EM Asia Consumer staples
Brazil Industrial
EMEA Utilities
UK Financials
EM Latam Energy
–80 –60 –40 –20 0 20 40 60 –80 –60 –40 –20 0 20 40 60 80
Falling risk-free rates and equity premium compression have In the United States, a few large firms have significantly outperformed
supported equity market performance, despite the drag from a weaker the rest of the stock market since the COVID-19 outbreak.
earnings outlook.
3. S&P 500: Decomposition of Equity Market Performance 4. US Stock Market Performance
(Percent contribution to cumulative returns) (Indices; February 19, 2020 = 100)
20 150
Top five tech firms
10 140
S&P 500 ex. top five
130
0 Russell 2000
120
–10 110
100
–20 90
Earnings (current and projected) 80
–30
Equity risk premiums 70
–40 Risk-free rate Price return 60
–50 50
Jan. 2020 Mar. 20 May 20 July 20 Sep. 20 Jan. Feb. Mar. Apr. May June July Aug. Sep.
2020 20 20 20 20 20 20 20 20
These top five firms tend to dominate certain sectors (information Valuations in major equity markets have become increasingly stretched
technology, telecommunications, consumer discretionary) and have by historical standards.
large international exposures.
5. Stock Market Performance and Shares of Foreign Revenues and of 6. Equity Market Misalignments
Top Five Tech Firms by Sector (Deviation from fair value per unit of risk, left scale; percentile based
(Price changes in percent since February 19, 2020, shares in percent) on 1995–2020 period, right scale)
Foreign revenues Top five tech firms Price change (right scale) Misalignment Percentile (right scale)
60 20 United Euro UK Japan China Brazil India
50 10 States area
0 6 100
40 5 90
–10 4 80
30 3 70
–20
20 2 60
–30 1 50
10 –40 0 40
–1 30
0 –50 –2 20
S&P 500
Consumer
discretionary
Information
technology
Materials
Telecom
Health care
Consumer
staples
Industrial
Utilities
Financials
Energy
–3 10
–4 0
Mar.
June
Sep.
Mar.
June
Sep.
Mar.
June
Sep.
Mar.
June
Sep.
Mar.
June
Sep.
Mar.
June
Sep.
Mar.
June
Sep.
Sources: Bloomberg Finance L.P.; Consensus Economics; Haver Analytics; Refinitiv I/B/E/S; and IMF staff calculations.
Note: In panel 3, the decomposition is based on a standard three-stage dividend discount model. See Panigirtzoglou (2002). In panel 4 and 5, the top five firms are
Alphabet (Google), Amazon, Apple, Facebook, and Microsoft. In panel 6, misalignment is the difference between market- and model-based values scaled by the
standard deviation of weekly returns; positive values indicate overvaluation. Intuitively, this measure indicates how many standard deviations of weekly returns (or
“units of risk”) it would take to get back to fair value. Misalignment in the euro area, Japan, and the United States is measured at the sector level and aggregated to
the index level by market capitalization. For other countries, misalignment is measured at the index level, due to data limitations. EM = emerging market;
EMEA = Europe, Middle East, and Africa; ex. = excluding; Latam = Latin America; UK = United Kingdom; US = United States.

Figure 1.7. Market Volatility and Economic Uncertainty
Despite an uncertain earnings outlook, the VIX and realized market ... as central banks’ actions have stabilized market conditions.
volatility have declined ...
1. VIX, Historical Market Volatility, and 12-Month-Forward Earnings 2. Drivers of US Option-Implied Equity Volatility
per Share Forecast Dispersion (Standard deviations from mean)
(Standard deviations from mean)
7 4.5
VIX External factors
Realized volatility Funding and liquidity conditions 4.0
6 EPS dispersion Corporate performance
Macroeconomic fundamentals 3.5
VIX
5
Model-fitted VIX 3.0
4 2.5
2.0
3
1.5
2 1.0
0.5
1
0
0
–0.5
–1 –1.0
Jan. Feb. Mar. Apr. May June July Aug. Sep. Oct. Jan. May Sep. Jan. May Sep. Jan. May
2020 20 20 20 20 20 20 20 20 20 2018 18 18 19 19 19 20 20
Sources: Bloomberg Finance L.P.; Consensus Economics; Refinitiv I/B/E/S; and IMF staff calculations.
Note: In panel 1, EPS dispersion is the standard deviation of EPS forecasts across analysts. Panel 2 is based on the VIX model presented in the October 2019 Global
Financial Stability Report (see Figure 1.2). EPS = earnings per share; VIX = Chicago Board Options Exchange Volatility Index.
significantly increased their participation in the stock mar- economic uncertainty and compressed equity market
ket in recent months, likely providing further support to volatility, though this gap has narrowed during the
equity prices.6 According to market analysts, the unwind September sell-off. For example, both option-implied
of retail positions, including in derivatives markets, may volatility (Chicago Board Options Exchange Volatil-
have contributed to the correction in the tech sector. ity Index [VIX]) and realized market volatility have
Has the stock market rebound gone too far? The IMF declined sharply in late March-April, reflecting improve-
staff’s equity valuation models suggest that overvalu- ment in funding and liquidity conditions following
ations are at historically high levels in some countries policy interventions, even though uncertainty about
(see Figure 1.6, panel 6).7 This disconnect has also earnings outlook has remained elevated for some time
been evident in a notable divergence between elevated (Figure 1.7). Although these misalignments could be
partially an unintended outcome of policy measures
6For example, in China, margin trading outstanding, which is
aimed at supporting investor sentiment and keeping
often cited as an indicator of retail investors’ activities, has increased
sharply since last year. In the United States, E*TRADE, Fidelity, markets open, it is difficult to separate intended from
Schwab, Robinhood, and Interactive Brokers all reported increased unintended effects quantitatively.
activity, new account sign-ups, or both. Trading on Robinhood Yields in credit markets have declined since the start
tripled in March 2020 compared with March 2019.
7The extent of equity price misalignments—the difference of the pandemic, reflecting both the decline in risk-free
between the actual price and the model-based value—can be rates and the compression in credit spreads on the back
interpreted as the portion of the equity risk premium that cannot of continued policy support. For example, the IMF staff’s
be explained by the explanatory variables included in the model:
valuation model for US investment-grade corporate bonds
expected corporate earnings (the mean earnings per share forecasts),
uncertainty about future earnings (the dispersion of earnings per suggests that central bank policy rate cuts and “other
share forecasts), term spreads, and interest rates (see the October policy support” (including asset purchases and other
2019 GFSR Online Annex 1.1 for details). The model relies on facilities) have partly offset the impact of the deterioration
12-month- and 18-month-ahead earnings forecasts and does not
capture the impact of the longer-term earning growth expectations in economic fundamentals that has occurred since the
on equity valuations. outbreak and that would have otherwise pushed bond

Figure 1.8. Credit Market Valuations
Much of the decline in the US investment-grade corporate bond yield Most bond spreads appear to be too compressed relative to
since March has been driven by policy support. fundamentals across both advanced and emerging markets.
1. Decomposition of Changes in US Investment-Grade Corporate 2. Bond Spread Misalignments
Bond Yields (Deviation from fair value per unit of risk, left scale; percentile based
(Basis points, left scale; percentage points, right scale) on 1995–2020, right scale)
Fundamentals Nominal risk-free rates Misalignment Percentile (right scale)
Other policy support Residual 6 100
Yield (right scale)
400 1.2 90
4
80
300
0.8 2 70
200 60
0
0.4 50
100 –2
40
0 0 –4 30
20
–100 –6
–0.4 10
–200 –8 0
Q4
Q1
Q3 to date
Q4
Q1
Q3 to date
Q4
Q1
Q3 to date
Q4
Q1
Q3 to date
Q4
Q1
Q3 to date
Q4
Q1
Q3 to date
Q4
Q1
Q3 to date
Q4
Q1
Q3 to date
–0.8
–300
–400 –1.2 IG HY IG HY IG HY IG HY
Jan. Mar. May July Sep. United States Euro Area EM hard EM local
2020 20 20 20 20 currency currency
Sources: Bloomberg Finance L.P.; Consensus Economics; Haver Analytics; Refinitiv I/B/E/S; and IMF staff calculations.
Note: The corporate bond valuation model in panel 1 is based on four groups of explanatory variables: economic (firm value) factors, uncertainty measures, leverage
metrics, and policy support factors. The group of policy support factors includes five variables: the size of the Federal Reserve’s balance sheet, the number of
announced policy measures, a dummy (0 before March 2020 and 1 thereafter), the amount of the Federal Reserve US dollar swap lines used (flow), and the
outstanding amount of the Federal Reserve US dollar swap lines (stock). The estimates are based on extreme bound analysis (see Durham 2002), which entails
running a large number of regressions covering all possible linear combinations of the explanatory variables in each of the four groups. The final model-implied bond
spread corresponds to the weighted average fitted value estimated across the various model combinations, in which the weights correspond to the R-squared
obtained from the respective regression. In panel 2, misalignment is the difference between market- and model-based values scaled by the standard deviation of
monthly changes in spreads; negative values on the left scale indicate overvaluation. Historical data go back to 1995 or earliest available. Latest data are through
September 29, 2020. The valuation model for the United States and the euro area is based on three groups of explanatory variables: economic factors, uncertainty
measures, and leverage metrics. For details, see October 2019 Global Financial Stability Report Online Annex 1.1. EM = emerging market; HY = high yield;
IG = investment grade.
yields higher (Figure 1.8, panel 1).8 More broadly, credit yields since March can also be traced to policy support,
spreads appear to be too compressed relative to economic including the spillovers from policy easing in advanced
fundamentals across both advanced and emerging markets economies. Rough estimates of the pass-through of US
(Figure 1.8, panel 2).9 In emerging markets, the decline policy actions to emerging market yields suggest that US
in hard currency bond spreads and in local currency bond policy actions since the COVID-19 sell-off account for
about one-quarter to one-half of the decline in emerging
8The corporate bond valuation model in Figure 1.8, panel 1,
markets’ long-term interest rates (see Online Annex 1.1).
is based on four groups of explanatory variables: economic (firm In local currency bond markets, both conventional and
value) factors, uncertainty measures, leverage metrics, and policy unconventional policies, such as asset purchases by emerg-
support factors. ing market central banks, have helped push short rates and
9The measures of misalignment shown in Figure 1.8, panel 2, for
advanced economy corporate bond spreads and emerging market

long-term yields lower (see Chapter 2).
sovereign bond spreads/yields may partly reflect the unprece- The sharp rebound in asset valuations, even if it is
dented policy support. Adding the policy support proxies to the partially the intended outcome of policies aimed at
corporate bond valuation model (as shown in Figure 1.8, panel 1)
creating a bridge to recovery, does raise concerns about
can help explain some, but not all, of the misalignments shown in
Figure 1.8, panel 2. the possibility of a market correction—as witnessed,

for example, with respect to tech stocks in Septem- industries, are much more vulnerable than large
ber. Current market valuations may be sustained for firms with access to capital markets.
some time, as long as there is a perception in markets •• Credit losses could deplete banks’ capital buffers,
that policy support will be maintained or scaled up affecting their ability and willingness to provide
in response to deterioration in economic conditions. credit to households and firms. Although the
Valuations may also continue to rise if pandemic- and global banking system is well capitalized, there is
policy-related uncertainties decline. However, the risk a weak tail of banks, and some banking systems
of a sharp adjustment in asset prices or periodic bouts may experience capital shortfalls in the adverse
of volatility remains and may rise should investors WEO scenario even with the currently deployed
reassess the extent or duration of policy support or if policy measures.
the recovery is delayed. •• Fragilities in the nonbank financial sector have aggra-
vated market dislocations during the March sell-off.
Central bank support has limited the fallout from
Global Financial Vulnerabilities Have Increased these fragilities but has not eliminated them. Market
since the COVID-19 Outbreak expectation that central banks will extend policy
The COVID-19 pandemic could be a major resil- support in response to adverse shocks may encour-
ience test for the global financial system. Before the age risk taking over and above desired levels.
outbreak, financial vulnerabilities were already ele- •• As policy space shrinks, the public-sector capacity to
vated in several sectors—including asset management continue to provide a backstop to the private sector
companies, nonfinancial firms, and sovereigns—across may come into question, especially where vulnera-
29 jurisdictions with systemically important financial bilities are high and rising across several sectors of
sectors (henceforth, S29) (see Figure 1.9) and likely the economy.
contributed to stress in financial markets during the •• External financing challenges facing emerging and
March sell-off (see the April 2020 GFSR).10 frontier markets may tip some of them into debt
Since the COVID-19 outbreak, vulnerabilities have distress or lead to financial instability.
continued to rise. Triggers such as new virus outbreaks,
policy missteps, or other shocks could interact with The rest of this section will focus on each of these
preexisting vulnerabilities and tip the economy into a areas. The rise in financial vulnerabilities increases the
more adverse scenario (see the October 2020 WEO). likelihood of adverse macro-financial feedback loops
In such a scenario, more widespread bankruptcies could in response to negative shocks, potentially requiring
lead to a repricing of credit risk, tightening of bank further liquidity and solvency policy measures.
lending standards, and a renewed sharp tightening of
financial conditions (see Chapter 3 for an analysis of this
dynamic in March). Solvency Risks in the Nonfinancial Sector Have
As the crisis continues to unfold, rising vulnerabili- Been Mitigated by Policy Support So Far
ties may create headwinds to recovery: Nonfinancial firms in many systemically import-
•• Widespread bankruptcies have been avoided so far ant economies entered the COVID-19 recession
thanks to large and frontloaded policy support. with elevated vulnerabilities, with the share of S29
However, as firms have borrowed more to cope with economies with high or medium-high corporate
cash shortages, some solvency risks have shifted into sector vulnerabilities already close to 80 percent (by
the future. SMEs, especially in contact-intensive GDP) before the pandemic (Figure 1.9).11 After the
outbreak, cash flows took a hit as economic activity
declined sharply. More vulnerable firms—those with
10The S29 include the euro area economies (Austria, Belgium,
France, Germany, Ireland, Italy, Luxembourg, The Netherlands,

Finland, Spain), other systemically important advanced econ- 11For example, the increased share of BBB-rated companies
omies (Australia, Canada, Denmark, Hong Kong SAR, Japan, among investment-grade borrowers in global credit markets and the
Korea, Norway, Singapore, Sweden, Switzerland, the United rapid expansion of risky credit markets raise the risk that credit rat-
Kingdom, the United States), and systemically important emerging downgrades and corporate defaults in the current downturn will
ing market economies (Brazil, China, India, Mexico, Poland, surpass levels observed during previous recessions. For details, see the
Russia, Turkey). April 2019, October 2019, and April 2020 GFSR issues.

Figure 1.9. Global Financial Vulnerabilities: High and Rising

Vulnerabilities have increased across more regions in the corporate and sovereign sectors as corporate borrowing surged amid the COVID-19
pandemic, whereas vulnerabilities in the nonbank financial sectors remain elevated.
1. Proportion of Systemically Important Countries with Elevated Vulnerabilities, by Sector
(Percent of countries with high and medium-high vulnerabilities, by GDP [assets of banks, asset managers, other financial institutions, and insurers];
number of vulnerable countries in parentheses)
Other financial 100% Sovereigns
institutions (12)
(16) 80%
More
vulnerable 60%
40%
Asset managers 20% Nonfinancial firms

(11) (21)
Insurers Households
(12) (16)
Oct. 2020 GFSR
Apr. 2020 GFSR
Global financial crisis
Banks
(12)
2. Financial Vulnerabilities by Sector and Region

Quintiles
Worst Best
Nonfinancial Asset Other Financial

Sovereigns Households Banks Insurers
Firms Managers Institutions
Apr. Oct. Apr. Oct. Apr. Oct. Apr. Oct. Apr. Oct. Apr. Oct. Apr. Oct.
2020 2020 2020 2020 2020 2020 2020 2020 2020 2020 2020 2020 2020 2020
Advanced Economies
United States
Euro area
Other advanced
Emerging Market Economies
China
Other emerging
Sources: Banco de Mexico; Bank for International Settlements; Bank of Japan; Bloomberg Finance L.P.; China Insurance Regulatory Commission; European Central
Bank; Haver Analytics; IMF, Financial Soundness Indicators database; Reserve Bank of India; S&P Global Market Intelligence; S&P Leveraged Commentary and Data;
Securities and Exchange Commission of Brazil; WIND Information Co.; and IMF staff calculations.
Note: In panel 1, “global financial crisis” reflects the maximum vulnerability value during 2007–08. In panel 2, dark red shading indicates a value in the top 20 percent
of pooled samples (advanced and emerging market economies pooled separately) for each sector during 2000–20 (or longest sample available), and dark green
shading indicates values in the bottom 20 percent. In panels 1 and 2, for households, the debt service ratio for emerging market economies is based on all private
nonfinancial corporations and households. Other systemically important advanced economies comprise Australia, Canada, Denmark, Hong Kong Special Administrative
Region, Japan, Korea, Norway, Singapore, Sweden, Switzerland, and the United Kingdom. Other systemically important emerging market economies are Brazil, India,
Mexico, Poland, Russia, and Turkey. Even though the latest readings for the insurance sectors in the United States and Japan and asset managers in China—based on
the available data—put them slightly below the threshold for the “medium-high vulnerability category” as of 2020:Q1, given the exceptionally high uncertainty these
sectors are categorized as “medium-high” in this assessment. The assessment for the insurance sector in the April 2020 GFSR was also revised as a result of a
change in Japan’s reading to “medium-high,” based on an update of the data available at the time. GFSR = Global Financial Stability Report.

weaker solvency and liquidity positions as well as uncertainty about the evolution of the pandemic
of smaller size—experienced greater financial stress and corporate credit quality. At the same time, credit
than their peers in the early stages of the crisis (see market pricing suggests a notably more sanguine
Chapter 3). Taking advantage of the massive easing in picture, likely reflecting expectations of continued
financial conditions, firms in advanced and emerging policy support.
market economies stepped up their bond issuance The future path of defaults and bankruptcies will
(Figure 1.10, panels 1–3), and also increased their critically depend on the evolution of the pandemic
borrowing from banks (Figure 1.10, panel 4) to cope and on policymakers’ capacity to maintain accommo-
with cash shortages, refinance their debt, or build pre- dative funding conditions and continue to provide
cautionary cash buffers. The rapid expansion of bank fiscal support to viable firms (see the October 2020
credit in the first half of this year partly reflects sizable Fiscal Monitor). Large firms with access to capital
credit line drawdowns, especially in the United States, markets can likely avoid a significant erosion of their
as well as government guaranteed loans and lending equity positions unless there is a significant tightening
under government-supported programs (Figure 1.10, in funding conditions. However, SMEs are much more
panel 5). The share of firms that had to raise new debt vulnerable (as discussed in Chapter 2 of the October
because they could not generate enough cash to cover 2019 GFSR), as they tend to have thin equity cush-
their debt service costs rose sharply (Figure 1.10, panel ions, low liquidity buffers (lack of precautionary credit
6). In all likelihood, without the policy support that lines and liquid and noncore assets), limited financing
facilitated such borrowing, nonfinancial firms would options, and nondiversified revenues. Furthermore,
have seen a sharp rise in bankruptcies. However, this the COVID-19 shock was particularly damaging for
further expansion of corporate debt has added to SMEs because they tend to dominate some of the
already high debt levels in several economies (Fig- most contact-intensive sectors (hotels, restaurants,
ure 1.10, panel 7). entertainment). Widespread insolvencies among SMEs
As the crisis continues to unfold, liquidity pressures could have a significant direct macroeconomic impact
may morph into insolvencies. Increased net borrowing as well as adverse implications for the health of the
has helped reduce liquidity pressures and mitigated an banking sector. Notably in Europe, SMEs account for
otherwise larger increase in defaults for now. However, more than half of total output and about two-thirds
rising debt may lead to a deterioration in repayment of employment and thus can affect financial stability
capacity over the medium term, putting solvency through macro-financial linkages. Because SMEs rely
at risk. Corporate credit quality has already shown almost entirely on bank financing, they could be a
signs of deterioration—credit rating downgrades source of vulnerability, especially for regional and
initially spiked and year-to-date speculative-grade small banks.
defaults have risen quickly, particularly in the United In the household sector, the COVID-19 pandemic
States (Figure 1.11, panel 1). Missed debt payments has resulted in unprecedented job losses, especially in
were reported as the leading cause of defaults in the United States, as well as in some emerging market
2020 to date. Firms in sectors most affected by the economies, where unemployment support has been
pandemic—air travel, retail, hospitality, and energy— more limited (see the October 2020 Fiscal Monitor).12
have seen higher default rates (Figure 1.11, panel With sharply reduced personal income of the affected
2). Looking across the credit spectrum, the largest households, their indebtedness has risen to cover lost
increase has been among high-yield bond issuers, income, further weakening their debt servicing capac-
followed by leveraged loans and middle-market loans, ity in the future. The new buildup of debt is taking
even though defaults are still significantly lower place on top of already elevated household leverage in
than in 2008–09 (Figure 1.11, panel 3). The pace a number of major economies (Figure 1.12, panel 1).
of defaults has recently slowed in the United States Historically, higher unemployment portends more
and has remained relatively subdued in Europe.
12A number of jurisdictions, notably in the euro area, have
Looking ahead, the range of speculative-grade default
implemented job retention schemes aimed at sustaining employment
forecasts for 2021 by credit rating agencies is fairly levels and mitigating financial vulnerabilities potentially arising
wide (Figure 1.11, panel 4), which reflects significant from households.

Figure 1.10. Easier Funding Conditions and Rising Debt
Bond markets have reopened for a broad range of issuers, with lower-rated issuers paying spreads higher than those before COVID-19.
1. Advanced Economy Corporate Bond and 2. Emerging Market Hard Currency Corporate 3. Advanced Economy and Emerging Market
Leverage Loan Issuance and Sovereign Bond Issuance Bond Spreads CEMBI IG+
(Billions of US dollars) (Billions of US dollars) (Basis points) EMBIG IG
400 Investment grade High yield Investment grade High yield 80 GABI IG 1,400
Non-rated Leveraged loans Non-rated CEMBI HY+
350 70 1,200
EMBIG HY
300 60 GABI HY 1,000
250 50
800
200 40
600
150 30
100 20 400
50 10 200
0 0 0
2019 Avg.
Jan.
Feb.
Mar.
Apr.
May
June
July
Aug.
Sep.
2019 Avg.
Jan.
Feb.
Mar.
Apr.
May
June
July
Aug.
Sep.
Jan. 2020
Feb.
Mar.
Apr.
May
July
Aug.
Sep.
Bank lending to nonfinancial firms was strong in the first half of the ... in part driven by credit line drawdowns and government guarantees.
year ...
4. Bank Credit Growth in Advanced and Emerging Market Economies, 5. New Loans, Credit Lines, and Government Guarantees
2020:Q2 in Major Advanced Economies, since March 2020
(Percent) (Billions of US dollars)
700
20 2020:Q2 2015–19 average Credit line drawdowns
Lending to corporate sector Lending to household sector Guaranteed loan commitments 600
16 New business loan volume 500
Changes in outstanding loans
12 400
300
8
200
4 100
0 0
US Euro Other China Other US Euro Other China Other USA GBR DEU FRA ITA ESP
area AE EM area AE EM
Increased borrowing helped firms cope with liquidity pressures as ... and has pushed aggregate corporate debt levels to new highs in
earnings collapsed following the outbreak ... several countries.
6. Publicly Listed Firms: Share of Debt with ICR < 1 and 7. Aggregate Nonfinancial Corporate Debt
Increased Net Debt (Percent of GDP)
(Percent of debt of sample firms) Range over past 10 years 2019:Q4 2020:Q1
ICR < 1 in 2019:Q4 ICR < 1 in 2020:Q2 180
60
ICR < 1 in 2020:Q2 and an increase in net debt between Q4 and Q2 150
50
40 120
30 90
20 60
10 30
0 0
North Euro Other Latin EMEA Emerging
FRA
CAN
JPN
ESP
USA
GBR
ITA
DEU
CHN
RUS
TUR
IND
POL
BRA
MEX
America area advanced America Asia
Sources: Banca D’Italia; Bank aus Verantwortung (KfW); Bank for International Settlements; Bank of England; Bank of Japan; Bloomberg Finance L.P.; BondRadar;
Dealogic; Emerging Portfolio Fund Research Global; Federal Reserve; French Ministry of the Economy and Finance; Haver Analytics; JPMorgan Chase & Co.; S&P
Global Market Intelligence; S&P Leveraged Commentary and Data; Spanish Instituto de Credito Oficial (ICO); and IMF staff calculations.
Note: In panel 5, the credit line draw downs are cumulative since 2019:Q4. New business loan volume and changes in outstanding loans are as of 2020:Q2. The
guaranteed loan commitment is as of July for United Kingdom and Italy, and as of August for the other countries. In panel 6, the sample includes firms with quarterly
statements. The bars show the share of debt at firms with ICR < 1 and with an increase in net debt as a share of total debt in the sample. In panel 7, for France,
corporate debt is reported on an unconsolidated basis. Data labels in panels 5 and 7 use International Organization for Standardization (ISO) country codes.
AE = advanced economy; CEMBI = JP Morgan Corporate Emerging Market Bond Index; EM = emerging market; EMBIG = JP Morgan Emerging Markets Bond Index
Global; EMEA = Europe, Middle East, and Africa; GABI = JP Morgan Global Aggregate Bond Index; HY = high yield; ICR = interest coverage ratio; IG = investment
grade; US = United States.

Figure 1.11. Solvency Risks in the Corporate Sector
Liquidity pressures and weaker credit quality have led to a rapid rise in Global consumer services and energy sector default rates have been
corporate defaults. more pronounced.
1. Global Speculative-Grade Corporate Defaults 2. Global Speculative-Grade Corporate Default Rates
(Year-to-date number of defaults) (Trailing 12-month rate, percent)
US Europe Other developed Emerging markets All sectors
200 All sectors excluding energy and consumer services 25
180 Consumer services
Energy
160 20
140
120 15
100
80 10
60
40 5
20
0 0
2008 09 10 11 12 13 14 15 16 17 18 19 20 2008 09 10 11 12 13 14 15 16 17 18 19 20
Defaults have risen across risky markets, with the largest increase ... and rating agencies have revised their default forecasts up, though
among high-yield bond issuers, followed by leveraged loans and the range of forecasts is fairly wide.
middle-market loans ...
3. US Speculative-Grade Corporate Default Rates by Market 4. US Speculative-Grade Default Rate: Actual and Forecasts by
(Percent) Credit Rating Agencies
(Trailing 12-month rate, percent)
16 US middle market leveraged loan default rate Rating 14
US large corporate leveraged loan default rate agencies’
14 12
US high yield bond default rate forecast
12 range 10
Recession
10
Actual 8
8 default rate
6
6
4 4
2 2
0 0
2008 09 10 11 12 13 14 15 16 17 18 19 20
1982
84
86
88
90
92
94
96
98
2000
02
04
06
08
10
12
14
16
18
20
Sources: Fitch; Haver Analytics; International Institute of Finance; Moody’s; S&P Global Ratings; S&P Leveraged Commentary and Data; and IMF staff calculations.
Note: In panel 4, the range in the projection period corresponds to the forecasts from Fitch, Moody’s, and Standard & Poor’s.
delinquencies and larger bank losses on unsecured Most Banks Will Be Able to Absorb Losses, but
consumer credit. For example, delinquencies on There Is a Weak Tail
US credit cards already started to accelerate in the Banks entered the COVID-19 crisis with signifi-
first quarter of this year, whereas delinquencies on cantly stronger capital and liquidity buffers than they
mortgages remain low (Figure 1.12, panel 2). In the had at the time of the global financial crisis thanks
housing markets, real house price growth was posi- to regulatory reforms (see Figure 1.9). Policies aimed
tive in most advanced economies in the first quarter, at supporting borrowers and at encouraging banks
boosted by broad policy support, particularly lower to use the flexibility built into the regulatory frame-
mortgage rates and moratoriums on interest pay- work have likely further supported their willingness
ments, foreclosures, and evictions. In emerging market to continue to provide credit to the economy. How-
economies, year-over-year real house prices declined ever, banks in some countries have started tightening
in China and India—following notable appreciation their lending standards in response to deterioration in
in previous years—but continued to rise in other economic conditions and borrowers’ financial positions
major economies. (see Chapter 4).

Figure 1.12. Solvency Risks in the Household Sector
Household debt is elevated relative to the size of the economy in ... and rising unemployment may portend higher delinquencies on
several advanced economies and in China ... loans to households.
1. Aggregate Household Debt 2. US Unemployment Rate and Delinquency Rates on Credit Card and
(Percent of GDP) Mortgage Loans
(Percent)
120 16
Range over past 10 years 2019:Q4 2020:Q1 Unemployment rate Credit card delinquencies
Mortgage delinquencies
14
100
12
80
10
60 8
6
40
4
20
2
0 0
2000 02 04 06 08 10 12 14 16 18 20
CAN
GBR
USA
FRA
JPN
ESP
DEU
ITA
CHN
POL
BRA
RUS
MEX
TUR
IND
Sources: Bank for International Settlements; Federal Reserve; US Bureau of Labor Statistics; and IMF staff calculations.
Note: Data labels in panel 1 use International Organization for Standardization (ISO) country codes.
Looking ahead, the resilience of banks will depend capital falls below regulatory minimum (Figure 1.13,
on the depth and duration of the COVID-19 recession, panel 1).14 Global systemically important banks tend to
governments’ ability to continue to support the private fare better, while banks in emerging markets appear to
sector, and the pace of loss recognition. Chapter 4 pres- be less resilient than their peers in advanced economies
ents a forward-looking bank solvency analysis based on (Figure 1.13, panel 1).
the October 2020 WEO baseline and adverse scenarios, In the October 2020 WEO adverse scenario,
taking into account announced policies to mitigate the capital shortfall relative to minimum capital
borrower distress and support bank capital levels.13 requirements is about $110 billion, whereas the
In the baseline scenario, most banks are able to absorb overall capital shortfall relative to broad capital
losses and maintain capital buffers above the minimum requirements—which include the countercyclical
regulatory capital requirements. In the adverse sce- capital buffer, the capital conservation buffer, and
nario, characterized by a deeper recession and a weaker systemic risk buffers—could reach $220 billion, after
recovery, there is a sizable weak tail of banks whose accounting for policy support (Figure 1.12, panel 2,
and Chapter 4). This implies that the average capital
shortfall in the adverse scenario is close to 1 percent of
13The analysis is carried out for about 350 banks accounting for
GDP. For comparison, the median government bank
about 75 percent of global banking assets. The exercise covers 29
jurisdictions, comprising Australia, Austria, Belgium, Brazil, Canada, recapitalization during the global financial crisis was
Denmark, Finland, France, Germany, Greece, Hong Kong SAR, about 3.6 percent of GDP. That said, the full fiscal
India, Indonesia, Ireland, Italy, Japan, Korea, Luxembourg, Mexico, cost of ensuring that banks are adequately capitalized
The Netherlands, Norway, Portugal, Singapore, South Africa, Spain,
Sweden, Switzerland, the United Kingdom, and the United States. must also include the direct fiscal support to firms and
In each jurisdiction, the largest banks covering up to 80 percent
of banking assets are included. Therefore, the simulation does 14The regulatory minimum is the “Pillar 1” requirement—4.5 per-
not include the consequences of the scenarios for the solvency of cent of risk-weighted assets—plus the mandatory buffers required of
small banks. each global systemically important bank.

Figure 1.13. Banking Sector: Potential Losses in the October 2020 World Economic Outlook Adverse Scenario
In the adverse scenario, the weak tail of banks is large, especially in Policy mitigation helps cushion some of the capital depletion and has
emerging markets. been stronger in advanced economies.
1. Distribution of Bank Assets by Capital Ratio under Adverse Scenario, 2. Broad Capital Shortfall under Adverse Scenario
with Policy Mitigation (Billions of US dollars)
(CET1 ratio, percent)
< 4.5% < 6% < 8% < 10% < 12% ≥ 12% Barebone: 4.5% + GSIB buffer
100 Broad: fully loaded 500
Barebone: 4.5% + GSIB buffer, with mitigation
Broad: fully loaded, with mitigation
80 400
60 300
40 200
20 100
0 0
19 T 22 19 T 22 19 T 22 19 T 22 19 T 22 Global AE EM GSIB Non-GSIB
Global AE EM GSIB Non-GSIB
Sources: Bloomberg Finance L.P.; Fitch; IMF, October 2020 World Economic Outlook; and IMF staff estimates.
Note: In panel 2, the shortfall is measured against bank-specific and fully loaded capital requirements effective August 2020, which include a minimum CET1 of
4.5 percent, a GSIB buffer, a systemic risk buffer, a stress capital buffer, a conservation capital buffer, and a countercyclical capital buffer, where applicable.
AE = advanced economy; CET1 = common equity Tier 1; EM = emerging market; GSIB = global systemically important bank; T = trough year.
households, which effectively reduced bank recapital- In combination with higher credit risk and increased
ization needs ex ante, and which may also adversely leverage in other financial institutions, this could
affect the fiscal capacity to provide additional support lead to larger potential losses in the event of renewed
in the future if needed. Furthermore, a more severe market stress.
adverse scenario that would entail larger losses for the During the March sell-off, fixed-income funds saw
banking sector cannot be ruled out, given the high a surge in redemptions, which led to selling pressures
degree of uncertainty around the depth and duration revealing some weaknesses in market infrastructures
of the COVID-19 recession. and dealers’ intermediation capacity (see April 2020
GFSR). Jurisdictions with swing pricing reportedly
saw less price pressure from redemptions.15 Fund
Fragilities in Nonbank Financial Institutions flows have generally recovered, reflecting the rebound
Remain Elevated in asset markets on the back of strong policy support
Asset managers in advanced economies entered the (Figure 1.14, panel 2). Insurance companies and pension
pandemic crisis with already elevated vulnerabilities funds, which experienced portfolio losses during the
(Figure 1.14, panel 1), including sizable liquidity mis- March sell-off, have also seen the value of their portfo-
matches (see April 2020 GFSR). After the outbreak, lios recover.
they faced increased credit risk and became more inter-
connected with banks. Exposures through investment
positions, including bank deposits and money market 15Swing pricing is the adjustment of a fund’s net asset value
fund shares, have risen. Borrowing from banks has with the aim to pass on the trading costs generated by purchases or
increased, as funds reportedly tapped into credit lines. redemptions to the shareholders who initiate those transactions.

Figure 1.14. Vulnerabilities in the Nonbank Financial Sector
Asset managers’ vulnerabilities remain elevated in China, the euro During the March 2020 sell-off, fixed-income funds experienced large
area, and the United States, and grew in OFIs in other advanced outflows, which have subsequently reversed.
economies.
1. Financial Vulnerability Indices and Sector Size 2. Cumulative Monthly Fund Flows
(Percent of assets under management)
China United States Other AEs Euro area Other EMs Fixed income Mixed Money market
1.0 2 25
0.9 OFIs
Financial vulnerability indices (percentile score)
1
0.8 20
OFIs AMs
0.7 0
0.6 15
–1
0.5
–2
0.4 10
AMs
AMs
AMs
OFIs
OFIs
0.3 –3
0.2 5
AMs
OFIs
–4
0.1
0.0 –5 0
0 10 20 30 40 50 60 70 80 90 100 110
Jan. 2020
Feb. 20
Mar. 20
Apr. 20
May 20
June 20
July 20
Aug. 20
Jan. 2020
Feb. 20
Mar. 20
Apr. 20
May 20
June 20
July 20
Aug. 20
Sector size (trillions of US dollars)
Sources: Banco de Mexico; European Central Bank; Haver Analytics; Morningstar; Reserve Bank of India; Securities and Exchange Commission of Brazil; WIND
Information Co.; and IMF staff calculations.
Note: Data in panel 1 are lagged at the end of the series by 18 months for UK AMs, by 15 months for Indian AMs, and by 3 months for Russian AMs as more recent
data are not yet available. For OFIs, data are lagged at the end of the series by 15 months for Switzerland and by 3 months for Russia. The financial vulnerability
indices reported are the base for the heatmaps reported in Figure 1.9. Panel 2 shows cumulative changes since December 2019. Data included for fixed income
funds, mixed funds, and money market funds covered 73%, 57%, and 75% of assets reported by the International Investment Funds Association for the respective
global fund sectors (as of end June). AEs = advanced economies; AMs = asset managers; EMs = emerging markets; OFIs = other financial institutions.
Looking ahead, risks from nonbank financial insti- due to or in conjunction with a lack of sufficient
tutions could stem from their portfolio rebalancing in policy support:
response to investor redemptions and market losses or •• First, liquidity mismatches in the asset management
from their decision to pull back from certain markets. sector remain elevated, especially in some fragile
In recent years, nonbank financial institutions have segments.16 The analysis of the March sell-off (see
been playing an increasingly important role in credit Box 1.2) shows that fixed-income funds facing large
markets, including in riskier segments (leveraged loans redemptions reacted primarily by reducing liquid
and private debt), which means that they could face assets, but also by selling less-liquid assets. The
sizable credit losses in the event of a surge in defaults sell-off of riskier assets contributed to price dislo-
and insolvencies (as discussed in Chapter 2 of the April cations in the underlying markets and could have
2020 GFSR). These losses could, in turn, lead them to resulted in larger-scale fire sales had central banks
step back from providing credit to these segments of not intervened quickly to backstop the key segments
the corporate sector, which would exacerbate strains on of the financial system. However, these interventions
borrowers and lead to worse macro-financial outcomes. have masked but not eliminated the pressure points.
Existing fragilities in the nonbank financial sec- A more prolonged liquidity shock in the future,
tor (Figure 1.14, panel 1) could have significant
implications for the financial system if a more pro- 16See Box 3.1 of the October 2019 GFSR, which presents the liquid-
longed period of market stress were to occur, possibly ity stress test for fixed-income funds in Europe and the United States.

Figure 1.15. Financial Leverage and Global Cross-Asset Correlations
Volatility-targeting investors have been re-leveraging as volatility Cross-asset correlations remain near the historic highs reached during
normalized following the March sell-off. the COVID-19 crisis.
1. Theoretical Leverage of a Volatility-Targeting Portfolio 2. Global Median Cross-Asset Correlation
(Total investment exposure to net asset value) (One-year rolling, weekly)
3.5 0.9
Leverage of theoretical volatility targeting portfolio COVID-19 crisis
Historical average of theoretical volatility targeting portfolio
0.8
3.0
Global financial crisis 0.7
2.5
0.6
2.0 0.5
1.5 0.4
0.3
1.0
0.2
Re-leveraging
0.5
0.1
0 0
2010 11 12 13 14 15 16 17 18 19 20 2003 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20
Sources: Bloomberg Finance L.P.; and IMF staff calculations.

Note: In panel 1, the leverage calculation for a theoretical volatility-targeting investment strategy assumes a theoretical investment portfolio consisting of 60 percent
global equities/40 percent bonds and an annual return volatility target of 10 percent. Leverage is defined as total investment exposure divided by the net asset value
of the portfolio. The MSCI World Equity Index is used as a proxy for equity investments; the Bloomberg Barclays Global Aggregate Total Return Value Unhedged index
is used as a proxy for bond investments. Panel 2 shows the median cross-asset correlation across nine global risky assets: global equities (proxied by the MSCI
World Equity Index), emerging market equities (proxied by the MSCI Emerging Markets Index), investment-grade credit (proxied by the Bloomberg Barclays Global
Aggregate Credit Total Return Index), high-yield credit (proxied by the Bloomberg Barclays Global High Yield Total Return Index), leveraged loans (proxied by the S&P
Global Leveraged Loan Index), mortgages (proxied by the Bloomberg Barclays Global Aggregate-Mortgages Index), emerging market sovereign bonds (proxied by the
JP Morgan EMBI Global Total Return Index), emerging market corporate bonds (proxied by the JP Morgan Corporate EMBI Broad Diversified Composite Index), and
commodities (proxied by the Bloomberg Commodity Index).
should these fragilities remain unaddressed, could leverage could contribute to asset price misalignments
potentially lead to larger-scale fire sales. and increase the risk of a sharp unwinding of posi-
•• Second, extremely low yields, compressed market vol- tions by leveraged investors during volatility spikes,
atility, and the apparent perception that central banks amplifying asset price declines.
will continue to backstop key markets are likely to •• Third, correlations across risk assets remain well
create incentives for financial releveraging. For exam- above the 2008–09 levels (Figure 1.15, panel 2).
ple, volatility-targeting investors that were reportedly These rising correlations may be partly driven by
forced to liquidate their positions during the March structural changes, including increased central
turmoil, thus amplifying the sell-off (see April 2020 bank presence in a number of markets. Higher
GFSR), may have already started to releverage as correlations tend to reduce portfolio diversification
equity and bond volatility normalized following cen- opportunities and could therefore increase contagion
tral bank interventions (see Figure 1.15, panel 1, for risk and propagate losses across investor portfolios
a theoretical portfolio).17 A rapid increase in financial during abrupt price corrections.
17Volatility-targeting To sum up, although swift policy actions have

strategies seek to keep expected portfolio vol-
atility to a specific target level. Lower market volatility then means mitigated risks to nonbank financial institutions during
that greater financial leverage is needed to meet volatility targets. the March sell-off, fragilities in the sector remain
Among these, variable annuity funds are the largest, at an estimated elevated and may lead to larger-scale distress and fire
$0.5 trillion in assets under management, and are more likely to
deleverage quickly when volatility spikes. See the April 2020 GFSR sales in a more prolonged episode of market stress.
for more details. In addition, increased linkages between nonbank

Figure 1.16. Sovereign Vulnerabilities and Interconnectedness
Sovereign debt has reached historically high levels in most jurisdictions ... with 6 out of S29 jurisdictions showing elevated vulnerabilities in all
with systemically important financial sectors ... three—corporate, banking, and sovereign—sectors.
1. Sovereign-Debt-to-GDP Ratios 2. Corporate, Bank, and Sovereign Vulnerabilities in the S29 Countries
(Bars = range over the past 30 years; dots = the latest value) (based on the data underlying Figure 1.9; red dots denote countries with
medium-high or high sovereign vulnerabilities)
300
Range over the last 30 years 2020E
medium-high — high
250
200
150
Higher bank vulnerabilities

low — medium-low
100
50
0
low — medium-low medium-high — high
JPN
ITA
SGP
USA
ESP
FRA
BEL
CAN
GBR
AUT
DEU
FIN
IRL
AUS
NLD
CHE
KOR
SWE
NOR
DNK
LUX
BRA
IND
MEX
CHN
POL
TUR
RUS
Higher corporate vulnerabilities
Sources: Bank for International Settlements; Haver Analytics; International Institute of Finance; IMF, October 2020 World Economic Outlook; and IMF staff estimates.
Note: Data labels in panel 1 use International Organization for Standardization (ISO) country codes. E = estimated; S29 = euro area economies (Austria, Belgium,
France, Germany, Ireland, Italy, Luxembourg, The Netherlands, Finland, Spain), other systemically important advanced economies (Australia, Canada, Denmark,
Hong Kong SAR, Japan, Korea, Norway, Singapore, Sweden, Switzerland, the United Kingdom, the United States), and systemically important emerging market
economies (Brazil, China, India, Mexico, Poland, Russia, Turkey).
financial institutions and banks imply that fragilities In 2020, headline fiscal deficits in advanced economies
could spread more easily through the financial system. are expected to be five times higher than in 2019 (see
Looking ahead, a prolonged period of low interest the October 2020 Fiscal Monitor). Emerging markets’
rates and high cross-asset correlations may pose further fiscal deficits have increased at a more modest pace,
challenges for institutional investors, whereas a widely largely reflecting financing constraints.
held belief that central banks will continue to suppress In the baseline scenario, public debt ratios are gener-
volatility may incentivize investors to take on more risk ally expected to stabilize in 2021, except in the United
and increase financial leverage to boost their returns. States and China. Unlike advanced economies, emerging
market economies will face greater fiscal challenges, as
their ratios of debt service to tax revenue are projected
Sovereign Debt Levels and Contingent to rise (see the October 2020 WEO). Although accom-
Liabilities Have Increased modative monetary policy could push interest rates
The COVID-19 crisis is expected to push global lower, hence potentially reducing sustainability concerns
public debt above 100 percent of GDP in 2020, the at higher debt-to-GDP levels, there could be a feedback
highest ever (see the October 2020 Fiscal Monitor). loop between high public debt and the risk premium
The large fiscal lifelines in response to the pandemic, (Alcidi and Gros 2019; Lian, Presbitero, and Wiriadi-
coupled with the sharp decline in output and higher nata 2020). Because private sector financing costs are
automatic stabilizers, have led to rapid expansion of linked to the sovereign risk premium, central banks in
sovereign debt. As a result, public debt reached historic emerging market economies where sovereign debt levels
highs in most systemically important economies at the are already high may face greater challenges in easing
end of the first quarter of 2020 (Figure 1.16, panel 1). financial conditions when they need to cushion the

impact of an adverse shock on the economy and the the second half of the year; some have increased their
financial system. This is because a sharp increase in the reliance on foreign currency debt,20 whereas elsewhere
sovereign risk premium could offset the central banks’ (Indonesia, Poland) central banks have purchased bonds
efforts to lower market interest rates. in the secondary market (see Chapter 2). Countries where
In addition, sovereigns may be facing a sharp the domestic investor base may not be deep enough to
rise in contingent liabilities. With the outbreak of absorb the additional supply could face some financing
the pandemic, vulnerabilities have increased across challenges.
multiple sectors (as shown in Figure 1.9), with 6 out The extraordinary level and speed of portfolio outflows
of S29 jurisdictions now showing elevated vulnerabil- from February to April 2020 created significant disrup-
ities in the corporate, banking, and sovereign sectors tions for emerging markets. Aggregate portfolio flows
(Figure 1.16, panel 2).18 Furthermore, bank holdings to emerging markets have recovered since then, driven
of government debt have increased in most countries, primarily by hard currency bond issuance, though more
again tightening sovereign-bank linkages. The simul- than half of emerging market economies have continued
taneous increase in vulnerabilities in the private and to experience outflows over the past three months, sug-
public sectors can also raise financial stability risks gesting that investors are differentiating across countries
through sovereign-corporate linkages at the local gov- based on economic fundamentals and policy frameworks.
ernment level, as is illustrated by the analysis presented IMF staff analysis based on the capital-flows-at-risk
for the case of China (see Box 1.3). methodology (see the April 2020 GFSR) points to an
improvement in the short- and medium-term outlook
on the back of easy global financial conditions, with the
Some Emerging and Frontier Markets May Face probability of outflows over the next three quarters fall-
External Financing Challenges ing from about 60 percent at the peak of market turmoil
Local currency government bond issuance—the (black line in Figure 1.17, panel 3) to about 25 percent
primary source of funding for many emerging market in September (red line in Figure 1.17, panel 3), though
sovereigns—picked up pace as the global backdrop still above the pre–COVID-19 level. Even before the
improved and domestic financial conditions in many pandemic, emerging market economies had elevated debt
economies eased. Several emerging market economies, vulnerabilities (see the October 2018 GFSR) and were
such as Chile, Colombia, and Thailand, have managed to dependent on portfolio flows (see the April 2020 GFSR).
fund large portions of their projected deficits for 2020–21 Increased fiscal deficits and external funding needs (rela-
(see Figure 1.17, panel 1), but many other economies still tive to exports) have made some emerging markets even
face significant financing requirements. Concerns about more vulnerable to shifts in external financing condi-
future debt supply and weak domestic fundamentals have tions, and these challenges are unlikely to moderate in
curtailed demand by nonresident investors, and portfo- the near term (see Figure 1.17, panel 4).
lio flows into local currency bond funds remain weak Frontier market economies face considerable financ-
since the COVID-19 sell-off (Figure 1.17, panel 2).19 ing challenges. Even before the global recession, the
As a result, many emerging markets (India and Mexico, share of frontier market economies in debt distress
among others) have delayed new local debt issuance to or at high risk of debt distress was relatively high (see
the October 2019 GFSR). The COVID-19 shock
18The sovereign vulnerability indicators behind Figures 1.9 and pushed borrowing costs for many of these economies
1.17 include standard balance-sheet indicators, such as government to prohibitive levels (Figure 1.18, panel 1). The Group
debt-to-GDP ratio, primary balance, maturity profile, etc. The of Twenty debt service suspension initiative sought to
assessment relies on the comparison of the latest values of these
help some 73 countries deal with financing pressures
indicators with those of a panel of peer countries (cross-section and
across time) (see annex to the April 2019 GFSR on the Indicator by allowing them to temporarily stop debt payments
Based Framework [IBF]). The objective of the IBF is to assess the to official creditors. The recent improvement in market
extent of financial vulnerabilities, which tend to contribute to
distress, in different countries and sectors. The forward-looking
assessments of the risk of distress (typically presented in the IMF 20Foreign-law foreign currency sovereign debt issuance has taken
debt sustainability assessments) are not part of the IBF. place at a record pace thus far in 2020. Some issuers have also relied on
19This is consistent with the findings of the April 2020 GFSR that increased local-law foreign currency debt issuance, such as Turkey reflect-
domestic fundamentals tend to influence local currency bond flows ing greater investor demand. Other countries with high foreign currency
more than hard currency bond flows. debt issuance in total government debt include Argentina and Ukraine.

Figure 1.17. Emerging Market Financing: Challenges, Options, and Risks
Government financing burdens remain steep in some countries with Investor flows into local currency bond funds remain weak.
issuance still lagging.
1. Local Currency Government Bond Gross Issuance Complete 2. EPFR Global Emerging Market Debt Dedicated Fund Flows and Returns
Relative to Estimated Total Issuance (Cumulative, year to date, billions of US dollars, left scale;
(Percent of total) percent, right scale)
Issuance pending for the year Hard currency fund flows
YTD gross issuance completed (through September) Local currency fund flows
67 percent line (elapsed time in year) Hard currency returns (right scale)
100 10 Local currency returns (right scale)
90 8
5
80 4
0
70 0
60 –5
50 –10 –4
40 –15 –8
30 –12
–20
20
10 –25 –16
0 –30 –20
Philippines
South Africa
Brazil
Malaysia
Turkey
India
Russia
Indonesia
Mexico
Poland
China
Hungary
Romania
Sri Lanka
Thailand
Colombia
Chile
Jan. 2020
Feb. 20
Mar. 20
Apr. 20
May 20
June 20
July 20
Aug. 20
Sep. 20
The outlook for portfolio flows remains challenging, with nearly The COVID-19 pandemic has exacerbated existing vulnerabilities,
25 percent probability of outflows next year. which are likely to remain elevated.
3. Capital Flows at Risk: Near-Term Portfolio Flow Forecast Densities 4. Evolution of Sovereign Debt and External Financing Requirements
(Probability Density) for EMs
March 23, 2020 September 29, 2020 (Percentile rank since 1990)
0.30 GFC 2019 2020 2023
Greater Current account
0.25 Aggregate 100
60% probability of vulnerabilities
portfolio outflows
an outflow 80
0.20 in 2020:Q1
60 Short-term debt to
0.15 25% probability of Sovereign debt 40 remaining
an outflow 20 maturity
0.10 5th percentile 0
0.05
0.00
–6 –5 –4 –3 –2 –1 0 1 2 3 4 5 6 7 8 9 10
Portfolio flows as a percent of GDP Fiscal balance External debt
Sources: Bloomberg Finance L.P.; Haver Analytics; HSBC analyst estimates; IMF, World Economic Outlook database; JP Morgan estimates; national sources; and
IMF staff estimates.
Note: In panel 1, data are not adjusted for inflation-linked debt. In panel 3, the analysis consists of portfolio flows (including both debt and equity components), based
on the model introduced in the April 2020 Global Financial Stability Report. The sample consists of 19 large and liquid emerging markets (Brazil, Bulgaria, Chile,
Colombia, Egypt, Hungary, India, Indonesia, Korea, Malaysia, Mexico, Peru, Philippines, Poland, Romania, Russia, South Africa, Thailand, Turkey). The capital flows at
risk (measured as the 5th percentile of the distribution) stands at –1.9 percent of GDP according to the latest assessment, which compares with –3.3 percent of GDP
on March 23 and realized portfolio outflows of almost 2 percent of GDP in 2020:Q1. In panel 4, the indicators are scaled by GDP. The figure plots the percentile rank
of the median value of the respective indicators across 71 major emerging markets in the corresponding year. The percentile rank is calculated since 1990. 2020
and 2023 estimates are based on World Economic Outlook database estimates. EMs = emerging markets; GFC = global financial crisis; YTD = year to date.

Figure 1.18. Emerging and Frontier Market Economy Spreads and Market Access
The COVID-19 pandemic pushed spreads of lower-rated economies to ... bringing into focus the large refinancing needs of several frontier
prohibitive levels ... market economies.
1. Hard Currency Bond Spreads 2. External Debt Service through the End of 2021
(Basis points) (Share of foreign reserves, percent, as of July 2020)
1,200 160
EMBI Loans: official Loans: non-official Bonds
Frontier market
economies 136 140
1,000 EMBI: Africa
EMBI IG
112 120
800 102
100
600 80
57 60
52
400
41
36 34 40
200 24
15 20
8
0 0
Jan. Feb. Mar. Apr. May June July Aug. Sep. Oct.
Zambia
Ethiopia
Pakistan
Angola
Mozambique
Kenya
Cameroon
Ghana
Senegal
Uganda
Nigeria
2020 20 20 20 20 20 20 20 20 20
Sources: Bloomberg Finance L.P.; World Bank Debtor Reporting System; and IMF staff calculations.
Note: EMBI = JP Morgan Emerging Markets Bond Index; IG = investment grade.
conditions has reduced these pressures, but many the policy focus will shift from dealing with liquidity
low-income countries with marketable debt have large pressures to managing a gradual reopening of the econ-
rollover needs (Figure 1.18, panel 2). This includes some omy and supporting the recovery. Table 1.1 provides a
that are eligible for the debt service suspension initiative road map for monetary and financial sector policies at
but are still unable to access international markets at different stages of the crisis.
pre–COVID-19 spreads (see Chapter 2 for discussion of
the role of creditor composition).
In late July and early August, Argentina and Ecua- Policy Priorities during Gradual Reopening
dor reached restructuring deals with bondholders. Under Uncertainty
These deals marked the end of protracted negotia- During this phase, which corresponds to the current
tions over both legal and financial terms and were a situation in a number of countries, lockdown measures
positive milestone for debt restructuring frameworks are eased, but uncertainty remains high, and contain-
going forward. ment measures may need to be reimposed if there
is a resurgence in cases. The priority for the gradual
reopening phase is to ensure that policy support is
Policies Need to Focus on Supporting a maintained for the recovery to take hold and become
Sustainable Recovery sustainable.
The pandemic has led to the worst global reces- •• Monetary accommodation should be maintained. After
sion since the Great Depression, and decisive and aggressively cutting policy rates early in the crisis,
timely policy actions have so far cushioned its impact most advanced economies are now facing effective
on households and firms, and managed to prevent lower bounds for conventional monetary policy,
economic stress from escalating into a full-fledged though there is still room for further policy cuts
financial crisis. As the economic recovery takes hold, in many emerging markets. Central bank balance

Table 1.1. Monetary and Financial Policy Road Map

Policy Areas Great Lockdown Gradual Reopening under Uncertainty Pandemic under Control
Monetary Policy Ease monetary policy, including Maintain monetary policy Maintain monetary policy
use of unconventional monetary accommodation accommodation until the policy
policy tools objectives (for example, inflation
target) are achieved
Liquidity Support to Provide support to maintain market Maintain support, but adjust pricing Gradually withdraw support, as
Core Funding Markets functioning and liquidity as appropriate to incentivize and warranted
prepare the ground for exit from use
of central bank facilities
Liquidity Support to Provide support to alleviate liquidity Maintain support, but adjust pricing Maintain liquidity support only
Financial Institutions stress and support monetary policy as appropriate to incentivize the return as required to support monetary
accommodation to normal market funding policy accommodation
Measures to Maintain Release macroprudential buffers, Continue allowing the use of capital Rebuild capital and liquidity buffers
the Flow of Credit allow the use of capital and liquidity and liquidity buffers gradually over time while ensuring
buffers, and apply regulatory continued financial institutions’
Suspend the distribution of banks’
flexibility as appropriate capacity to extend credit
profits (dividend payouts and share
Suspend the distribution of banks’ buybacks)
profits (dividend payouts and share
buybacks)
Provide financing support

to households and businesses
(see below)
Measures to Address Provide guidance on asset Maintain prudential standards Require banks to develop credible
Problem Assets classification and provisioning to incentivize the recognition and plans to reduce problem assets over
handling of problem assets an appropriate period of time
Handle weak banks that experience

significant credit losses
Foster the development of markets

for distressed assets
Financing Support Provide credit guarantees (or other Maintain financing support if containment Withdraw unwarranted support
to Business risk mitigation) and term funding to measures are reintroduced, but tighten
support new lending eligibility criteria to better target illiquid
but solvent firms
Debt Restructuring Introduce repayment moratoria Extend repayment moratoria only Facilitate debt restructuring that
for Businesses and if necessary to prevent widespread reduces debt overhang
Households insolvencies
Facilitate debt restructuring that

reduces debt overhang and/or adjust
repayment schedule
Provide solvency support to viable

systemic firms, grants for smaller firms
Ensure efficient out-of-court

agreements, with fast-track procedures
to support debt restructuring
Source: IMF staff.

sheets have also grown significantly since March sustainable, should be balanced against the need for
2020. Some emerging market central banks have banks to continue providing credit to the economy
launched asset purchase programs to stabilize local during both reopening and recovery phases.
markets and ease financial conditions, but in some •• Policymakers should develop effective strategies to
cases, these purchases have also facilitated financing deal with corporate and household solvency pressures.
of government deficits. In such cases, transparency Measures to alleviate liquidity stress can provide
and clear communication of the policy objectives are only temporary relief. Financing support will further
crucial to minimize risks to central bank credibility increase indebtedness, whereas firms and households
and the perception that these programs are used for may still face some financing difficulties after the
monetary financing—especially in countries with moratoria on debt repayments are lifted. Policymak-
weaker institutional and governance frameworks ers should shift their focus to solvency support. For
(see Chapter 2). instance, solvency support for firms deemed strategic
•• The necessary liquidity support to financial markets or systemic could mitigate adverse macro-financial
and institutions should be maintained. A number consequences. For SMEs, which account for a large
of backstops remain in place.21 Many central bank share of employment in some countries, govern-
programs were designed to provide support at prices ments could consider providing grants (see the
that were attractive in stressed markets but are at a October 2020 WEO).
premium in normal conditions. This feature creates •• Emerging and frontier market economies facing financ-
incentives for financial institutions to return to mar- ing difficulties may require official support. Financing
kets as funding conditions normalize. The presence widening fiscal deficits could be a challenge because
of these facilities still provides support to markets, of deteriorating public finances and shallow domestic
even if actual use is limited. markets.23 The IMF has proactively provided financing
•• Banks should be encouraged to continue lending. support to member countries during the COVID-19
Whereas banks should continue to make use of the crisis (80 countries to date).24 However, public debt
flexibility built into regulatory frameworks, pruden- may become unsustainable in some countries, and debt
tial and accounting standards for loan classification restructuring with international creditors would be
and provisioning should be maintained.22 Timely needed to safeguard macro-financial stability.
and reliable recognition of loan losses based on the
expected credit loss framework (under International
Financial Reporting Standard 9) is essential, but Policy Responses if Recovery is Delayed
country authorities may want to delay the impact •• In the event of a deterioration of the economic outlook
of additional provisions on regulatory capital, with (for example, due to new outbreaks), policymakers
adequate disclosure of fully loaded capital positions. should be prepared to scale up liquidity support but
Supervisors should provide guidance on how banks in a more targeted manner. Targeted fiscal measures
should deal with restructured loans, including would be an efficient way to help the most vulner-
those resulting from moratoria on loan repayments. able firms and individuals (see the October 2020
For example, in commercial real estate markets, Fiscal Monitor). Eligibility criteria would need to
extended forbearance and foreclosure moratoriums be gradually tightened to ensure that most of the
could help limit contagion across commercial prop- support goes to viable firms.25 This would help
erty markets (see Box 1.1). Guidance on the usabil- prevent a buildup of debt overhang further down
ity of bank buffers, including the optimal pace of the road, support necessary business adjustments
rebuilding these buffers once the recovery becomes and debt restructuring, and facilitate post-pandemic
reallocation of resources. Moratoria on repayments,
21For example, the Federal Reserve extended its support programs
until the end of 2020. 23For guidance on how sovereign debt managers handle financing
22According to the Financial Stability Board, there have been challenges, see IMF (2020c).
a few cases of measures that went beyond the flexibility of the 24For an overview of policy responses to maintain macro-financial
standards (reducing certain credit risk capital and leverage ratio stability in emerging market and developing economies, see
requirements, lowering liquidity requirements, and postponing the IMF (2020d).
application of the large exposure framework), but most of these 25For guidance on how to provide liquidity support to businesses,
measures are temporary and will be reversed as the crisis abates. see IMF (2020b).

which provide temporary relief, should be extended market liquidity as well as encourage excessive risk
only if necessary to prevent widespread insolvencies taking if it becomes embedded in investor expecta-
stemming from renewed lockdowns. tions. Systemwide liquidity support should be with-
•• Monetary policy may have to be eased further as needed drawn as market conditions normalize. Protracted
to support the flow of credit to the economy. Emer- liquidity support, including financing support to
gency lending and unconventional monetary policy businesses and moratoria on repayments, could keep
easing may have to be reactivated or expanded, nonviable borrowers afloat. This could delay the
depending on country circumstances, if the econ- business restructuring, balance sheet correction, and
omy slips into an adverse scenario in coming resource reallocation that are necessary to restore
months. macro-financial resilience.
•• Policymakers should provide solvency support to •• Banks should be encouraged to proactively clean up
mitigate systemic risk. Targeted transfers and tax nonperforming loans. Banks with high levels of non-
relief could be provided to hard-hit businesses and performing loans should be required to develop and
households. In addition, governments could scale up implement credible action plans to reduce nonper-
the solvency support to viable firms that are deemed forming loans within an appropriate time frame. To
strategic or systemic individually or collectively to underpin confidence, authorities should ensure that
mitigate adverse macro-financial consequences. banks maintain transparency on the performance
of their loan portfolios, the materiality of loan
restructuring, and any material adjustments made
Policy Priorities once Pandemic Is under Control to risk management and accounting policies. Some
Once the virus is fully under control, policymakers banks may face capital shortfalls as they recognize
should build on the policy actions taken during the credit losses. Supervisors may consider suspend-
gradual reopening phase, but with a greater focus on ing automatic triggers for corrective actions and
tackling solvency issues to ensure a sustainable recovery instead require banks to present credible plans to
and completing the structural transformation of the restore their capital.28 Exceptional measures taken to
economy to the new post-pandemic normal. support distressed borrowers should be phased once
•• Monetary policy accommodation should be maintained conditions allow.
until central bank objectives are achieved. Given •• Policymakers should develop effective strategies to
expectations of continued low inflation (see Online deal with private debt overhang. Well-functioning
Annex 1.1) and the likelihood of a pronounced insolvency frameworks can help ensure efficient
decline in real interest rates for many years, central exit of nonviable firms and facilitate the necessary
banks (including the US Federal Reserve and the structural transformation. Firms facing solvency
European Central Bank) are considering adjustments challenges should be recapitalized, restructured, or
to their monetary policy frameworks and commu- resolved:
nications to ensure policy efficacy, especially at the oo Recapitalization could be an option for firms
effective lower bounds.26 deemed viable (for example, with earnings suf-
•• Liquidity support should be withdrawn as warranted ficient to cover interest expenses). In such cases,
once conditions improve. Term funding provided equity-like support could prove more useful than
to banks may be maintained as needed to support liquidity support (as liquidity support leads firms
credit flows and ensure a sustainable recovery.27 to accumulate more debt). Modalities could vary
Prolonged central bank support in key financial depending on firms’ characteristics (SMEs, for
markets may distort price discovery and affect example, as discussed previously) and would need
to account for country-specific institutional and
26For example, Jordà, Singh, and Taylor (2020) found that past legal frameworks.
pandemics were followed by sustained periods of depressed invest- oo Restructuring of debt could be suitable for firms
ment opportunities and/or increased precautionary saving.
27Some central banks are beginning to withdraw support with facing structural challenges (because of the
no impact on market functioning. Examples include a reduction in
the size and frequency of open market operations in most advanced
economies and moderation of the pace of purchases of government 28For discussion of banking regulatory and supervisory issues in
securities in some advanced economies. response to the COVID-19 crisis, see IMF (2020a).

COVID-19 pandemic). In such cases, adjust- system to cut expenses (for example, physical
ments to firms’ business models would be branches) and extend services to underserved pop-
required to restore viability. Simplified, standard- ulations, thereby increasing financial inclusion. Digi-
ized procedures should be developed to facilitate tal currencies in particular could offer substantial
out-of-court agreements on debt restructuring. efficiency gains, especially in cross-border payments,
oo Resolution, or facilitation of an orderly exit, and reach unbanked populations. However, they
should be applied to unviable firms that can- need to be carefully regulated to ensure financial
not be saved through restructuring. Fostering stability and integrity, operational safety, market
the development of markets for distressed assets contestability, and consumer protection.
would facilitate their disposal.
•• Policymakers should prepare to deal with the implica- Post-Pandemic Financial Reform Agenda
tions of corporate and household insolvencies for banks To safeguard global financial stability and promote
and nonbank financial institutions, as well as for inclusive, sustainable growth in the post-pandemic era,
sovereigns. Bank and nonbank financial institutions the regulatory reform agenda should focus on strength-
will need to absorb credit losses, and some regu- ening the regulatory framework for nonbank financial
lated financial institutions may experience capital sector and stepping up prudential supervision to curb
shortfalls. Country authorities should ensure that excessive risk taking in the lower-for-longer interest
banks have credible recovery strategies in place and rate environment:
develop (or update) contingency plans for institu- •• Strengthening the regulatory framework for the
tions displaying substantial fragilities. Resolution nonbank financial sector: In light of lessons learned
tools, which have been strengthened since the during the COVID-19 crisis—including central
global financial crisis, should be used as necessary banks’ need to backstop essential segments of
to resolve failing banks in an orderly way. At the financial markets—policymakers should assess the
sovereign level, steps should be taken to develop a effectiveness of prudential tools that are currently
credible medium-term fiscal strategy to ensure debt available and consider strengthening the pruden-
sustainability in the medium term, considering that tial regulation as well as broadening the regulatory
prolonged policy support could translate into signif- perimeter of nonbank financial institutions.
icant fiscal costs. oo The operational frameworks for central counter-
•• Policymakers should adopt policies to encourage more party clearing houses (CCPs) have to be adjusted
proactive management of climate-change-related in light of the crisis experience (see April 2020
risks. The pandemic, despite substantial negative GFSR). While CCPs played an important role
effects on firms’ environmental performance (see in cushioning the impact of market stress during
Chapter 5), presents an opportunity to engineer the March sell-off, policymakers should exam-
a green recovery. Policymakers should encourage ine options for prudently limiting procyclicality
the appropriate pricing of climate-change-related in margin calls as well as ensuring derivatives
risks through gradual and well-communicated counterparties are able to anticipate and pre-
implementation of carbon taxes, better disclosure pare for them.
of climate-change-related risks, and increased use oo To enhance the global financial system’s resil-
of climate stress tests for financial institutions. ience, a more robust liquidity risk management
This could in turn generate the right incentives framework should be adopted for investment funds
to reduce exposures to physical risk and expedite (International Organization of Securities Com-
the transition. missions 2018), including a broad set of tools to
•• Policymakers should adopt policies to encourage greater better manage redemptions as well as to identify
digital investment to enhance financial sector efficiency related risks early (see the October 2019 GFSR).
and inclusion. The pandemic may have accelerated The usability of liquidity buffers in crisis times—
the transition of the economy toward digitalization. which has proven key in the banking sector this
Digital investment should enable the financial year—could be more actively considered. To the

extent the swing pricing has been successful in to resume and may lead to excessive risk taking.
helping to contain redemptions, a wider adop- Given the existing balance sheet weaknesses, a fur-
tion would be advisable, particularly in juris- ther buildup of leverage in the post-pandemic world
dictions with sizable asset management sectors. should be contained appropriately. The macropru-
Given jurisdiction-specific institutional and legal dential policy framework should be strengthened
arrangements, however, swing pricing will likely to ensure adequate capital and liquidity buffers in
have to be phased in over time, requiring modifi- banking systems, to contain excessive risk taking in
cations to the existing operational infrastructure. the nonbank financial sector and to create mac-
An internationally harmonized measurement of roprudential space that could be used to cushion
leverage in investment funds (International Orga- the impact of adverse shocks on the economy and
nization of Securities Commissions 2019) should financial system.29 Prudential authorities could
help with the timely recognition and mitigation implement measures such as loan-to-value ratio and
of respective financial stability risks. debt-to-income ratio to prevent excessive risk taking
that could inflate property prices, including in the
•• Implementing micro- and macroprudential measures commercial real estate segment (see Box 1.2).
to curb excessive risk taking in the lower-for-longer
29For instance, the ECB emphasized in its recent Financial Stabil-
interest rate environment: With market participants
ity Review the importance of creating the macroprudential space in
anticipating interest rates to remain very low for the the euro area in the form of releasable countercyclical capital buffers
foreseeable future, investor search for yield is likely (CCyBs) to help sustain credit in a downturn.

Box 1.1. Are Financial Stability Risks Rising in Commercial Real Estate Markets?
Market participants and policymakers have year over year, in the European Union and the United
increasingly pointed to the commercial real estate States, respectively (Figure 1.1.1, panel 4).
sector as a potential source of financial stability risks Stress in funding markets early this year rever-
because of its notable size, procyclicality, and systemic berated through the commercial real estate sector.
nature. In several economies, commercial real estate Funding costs increased sharply in mid-March, with
loans constitute a significant part of banks’ lending the spread on BBB-rated commercial mortgage-backed
portfolio, especially at local and regional banks.1 securities and commercial mortgage-backed security
Commercial mortgage-backed securities issuance has indices remaining much higher in June relative to the
also recovered since the global financial crisis, with pre-pandemic level (Figure 1.1.1, panel 5). Syndicated
the total volume exceeding $100 billion in 2019 commercial real estate lending dropped by about
(Figure 1.1.1, panel 1). Historically, volatility in the 50 percent in North America, 70 percent in Europe,
commercial property market has often been an ampli- and 40 percent in Asia in the second quarter of 2020,
fier of macro-financial instability—for example, in the year over year. Whereas the slowdown in lending may
United States in 2008. partly be a result of a drop in demand, increasing
In recent years, the riskiness of the commercial real delinquency rates and tightening of credit conditions
estate sector has increased globally. Over 2009–19, for bank loans, as is evident from the US Senior Loan
commercial property asset valuations rose, on aver- Officer Opinion Survey, may have also played a role
age, 4.5 percent a year to reach historical highs in (Figure 1.1.1, panel 6).3
several economies.2 Concurrently, capitalization Looking ahead, there is considerable uncertainty
rates—which measure rental income relative to the about the outlook for the commercial real estate sector.
value of the property—fell to their lowest levels (Fig- As economies open up, activity in the sector is likely
ure 1.1.1, panel 2). to pick up. However, based on current projections
The COVID-19 crisis has inflicted significant pain from rating agencies, the commercial mortgage-backed
on the sector. Worldwide commercial property trans- securities default rates are expected to more than dou-
actions slumped by about 50 percent in the second ble in the third quarter of 2020, suggesting that the
quarter of 2020 relative to last year, as containment sector may remain under pressure for a while. More-
measures imposed in response to the pandemic over, segments such as retail could continue to face
adversely affected economic activity and reduced the headwinds even after the pandemic is over because of
demand for commercial properties. Within the sector, the ongoing increased shift toward e-commerce. The
retail and hospitality businesses have been the most demand for office space may also drop as companies
affected, with sales down by 60 percent and 80 per- experiencing cost savings of work-from-home arrange-
cent, respectively (Figure 1.1.1, panel 3). Available ments consider extending them into the future.4 All
price data also point to a significant decline, especially in all, these shifts could induce significant volatility in
in the retail sector, with the retail sector price index commercial property markets and bear close monitor-
falling by about 18 percent and 23 percent in July, ing to limit broader macro-financial stability risks.
The authors of this box are Andrea Deghi and Salih Fendoglu. 3In the United States, 5.8 percent of commercial
1In the United States and the euro area, for example, com- mortgage-backed securities loans were delinquent in the second
mercial real estate loans constituted 50 percent and 23 percent, quarter of 2020, an increase of more than 200 basis points
respectively, of total bank lending to nonfinancial corpo- relative to the previous year.
rates in 2019. 4For example, a recent corporate survey by Green Street Advi-
2In some economies, for example Hong Kong SAR, Sweden, sors shows that the propensity of staff to work from home in
and the United States, commercial real estate valuations more the medium to long term has increased by about 30 percentage
than doubled between 2009 and 2019. points since the pandemic crisis.

Box 1.1 (continued)

Figure 1.1.1. Trends and Developments in Commercial Real Estate Markets
CMBS issuance has increased since the global financial ... whereas capitalization rates have continued to fall.
crisis ...
1. CMBS Issuance 2. Capitalization Rates for Selected Economies
(Billions of US dollars) (Percent)
350 Interquartile range Median US 10
US CMBS
280 Non-US CMBS US (spread over 10-year government bond yield) 8
210 6
140 4
70 2
0 0
2005 07 09 11 13 15 17 19 2001 03 05 07 09 11 13 15 17 19
Global commercial property transactions fell sharply in ... with prices also dropping, especially in the retail
2020:Q2 ... sector.
3. Change in CRE Transaction Volumes 4. Change in CRE Prices across Sectors
(Percent, 2020:Q2 versus 2019:Q2) (Percent, July 2020 versus July 2019)
Overall Office Industrial Retail Hotel
0 16
US Europe
–20 8
0
–40
–8
–60
–16
–80 –24
–100 –32
Global North and Europe and Asia and All Retail Office Industrial
Latin America other regions Pacific
Funding costs in the CMBS market have increased ... whereas lending standards have tightened, and
sharply ... delinquency rates have inched up in 2020:Q2.
5. CMBS Funding Conditions in the United States 6. Credit Standards and Delinquency Rates in the
(Basis points) US CMBS Market
(Percent)
1,000 3,200 100 20
CMBS OAS BBB CRE loans survey
800 CMBX S6 BBB– (right scale) 80 CMBS loans delinquency rates 16
2,400 (right scale)
CMBX S9 BBB– (right scale)
600 60 Forecast (right scale) 12
1,600
400 40 8
200 800
20 4
0 0 0 0
Aug. Nov. Feb. May Aug. Nov. Feb. May Aug.
2007:Q2
08:Q1
08:Q4
09:Q3
10:Q2
11:Q1
11:Q4
12:Q3
13:Q2
14:Q1
14:Q4
15:Q3
16:Q2
17:Q1
17:Q4
18:Q3
19:Q2
20:Q1
2018 18 19 19 19 19 20 20 20
Sources: Bloomberg L.P.; Commercial Mortgage Alert; Federal Reserve Bank; Green Street Advisors; Moody’s; MSCI Real Estate; Real
Capital Analytics; and IMF staff calculations.
Note: Panel 1 shows the total issuance of CMBS for the United States and other countries. Panel 2 shows the capitalization rate for the
United States and other selected economies and the spread of the US capitalization rate over the 10-year US government bond yield.
Selected economies are Australia, Austria, Belgium, Canada, China, the Czech Republic, Denmark, Finland, France, Hungary, Hong Kong
SAR, Indonesia, Ireland, Italy, Japan, Korea, Malaysia, The Netherlands, New Zealand, Norway, Poland, Portugal, Singapore, South
Africa, Spain, Sweden, Taiwan Province of China, Thailand, and the United Kingdom. Panel 3 shows the change in global real estate
sales (single asset, portfolio, entity) in 2020:Q2 relative to 2019:Q2. Panel 4 shows the change in the commercial property price index in
July 2020 relative to July 2019 for different CRE sectors and for the overall market. Panel 5 shows the spreads over the Treasury yield
curve for the Bloomberg Barclays Global Aggregate BBB index and the CMBX S6 and CMBX S9. Panel 6 shows the percent of
respondents in the US Senior Loan Officer Opinion Survey indicating a tightening in CRE lending standards and CMBS loan delinquency
rates (historical and projected to 2020:Q3). CMBS = commercial mortgage-backed security; CMBX = commercial mortgage-backed
security index; CRE = commercial real estate; OAS = option-adjusted spread; US = United States.

Box 1.2. The Behavior of Investment Funds during COVID-19 Market Turmoil
In March 2020 the global investment fund sector tion, swing pricing may have helped funds manage
and, in particular, fixed-income and nongovern- redemptions.4
ment money market funds experienced a short As a result, fixed-income funds that were forced to sell
period of intense withdrawals as investors redeemed assets in response to redemption pressures seem to have
shares following a sharp increase in valuation had some adverse effect on both asset prices and market
uncertainty in many asset classes, including debt liquidity. In March 2020 the bid-ask spreads of assets
securities (Figure 1.14, panel 2).1 The market sold most heavily by fixed-income funds facing large
liquidity of securities held by fixed-income funds redemptions increased more than the bid-ask spreads of
deteriorated substantially, as evidenced by the assets not facing such selling pressure. Similarly, during
near doubling in the average bid-ask spreads of March 2020 cumulative returns of assets under selling
securities held in their portfolios (Figure 1.2.1, pressure declined more than assets experiencing no pres-
panel 1).2 Though liquidity declined for almost all sure (Figure 1.2.1, panel 4). Hence, funds’ sales of liquid
fund portfolios, average bid-ask spreads more than assets are likely to have contributed to price pressures
tripled temporarily for the most affected portfolios, and liquidity strains observed in fixed-income markets.
indicating that a few funds bore the brunt of the Similarly, increased incentives for funds to sell corpo-
liquidity impact, while on average the industry rate bonds may have amplified the price dislocations
proved resilient. observed in risky credit markets in March 2020. Some
With only a handful of funds suspending redemp- funds, however—even some of those experiencing large
tions,3 most fixed-income funds resorted to a mix outflows—may have helped to mitigate price pressures,
of strategies to deal with outflows. First, the most as they were willing to absorb relatively illiquid assets
afflicted funds used their relatively ample liquidity even under uncertain market conditions (Figure 1.2.1,
buffers and shed liquid assets such as cash, cash panel 2, right side, and panel 4).
equivalents, and US Treasuries to cover redemptions, The behavior of fixed-income funds and their clients
whereas funds receiving inflows hoarded cash and during the March 2020 redemption stress episode
delayed investments, presumably because of uncertain highlight some fragilities in this industry. Selling rela-
market conditions (Figure 1.2.1, panel 2). Second, tively liquid assets first might have further intensified
despite large outflows, some funds were willing to funds’ liquidity mismatches, if liquidity conditions had
purchase assets at high bid-ask spreads, possibly not improved so rapidly. The weakening in the average
using cash reserves to take advantage of depressed liquidity profile of funds facing outflows may have also
prices of potentially illiquid assets (Figure 1.2.1, made them more susceptible to future redemption or
panel 2). Third, with their investors more sensi- valuation shocks. The sale of less liquid assets has con-
tive to performance and less amenable to increased tributed to price dislocations in the underlying asset
corporate exposures, fixed-income funds were less markets. In combination with fund investors’ increased
inclined to retain their relatively high exposures to sensitivity to fund performance, this could have gen-
corporate bonds, especially if they were anticipating erated feedback loops resulting in larger-scale fire sales
more redemptions (Figure 1.2.1, panel 3). In addi- had central banks not stepped in so quickly with asset
purchase programs and liquidity facilities.
Looking ahead, a comprehensive review of available
The authors of this box are Frank Hespeler and prudential tools in the investment fund sector, includ-
Felix Suntheim. ing considering a more widespread adoption of swing
1These outflows are still lower than those assumed under the
liquidity stress presented in Box 3.1 of the October 2019 Global

pricing, would help to mitigate vulnerabilities revealed
Financial Stability Report. during the COVID-19 market turmoil.
2Based on a sample of 323 fixed-income funds with available
information on individual securities held in their portfolios. 4Data limitations did not allow for an analysis of the effec-
3Fitch reported for 2020 that mutual funds suspended a total tiveness of swing pricing during the March 2020 turmoil period.
of $62 billion year to date, a mere 0.11 percent of the sector’s However, Jin and others (2019) provide respective evidence for
total assets (Fitch Ratings 2020). UK corporate bond funds during stress periods.

Box 1.2 (continued)

Figure 1.2.1. Vulnerabilities of Fixed-Income Funds Exposed during the March 2020 Market Turmoil
During March 2020, the liquidity of the fixed-income Funds facing redemptions reduced cash buffers and sold
funds’ portfolios deteriorated substantially. liquid assets, but in some cases also purchased illiquid
assets, taking advantage of illiquidity discounts.
1. Bid-Ask Spreads of Fixed-Income Funds’ Portfolios 2. Portfolio Shares of Cash and Fund Flows (left panels)
(Percent) and Bid-Ask Spreads of Assets Bought and Sold by
Funds (right panel), by Flow Quintile
(Percent)
2.5 Interdecile range Mean Cash share (left scale) Assets sold
Mean flow (right scale) Assets bought
2.0 14 1st flow 10 12 5th flow 8 1st flow 5th flow 140
quintile quintile 7 quintile quintile
12 5 10 120
1.5 10 6 100
0 8 5
8 80
1.0 –5 6 4
6 60
–10 4 3
0.5 4 2 40
2 –15 2 1 20
0 0 –20 0 0 0
July 2019
Oct. 19
Jan. 20
Apr. 20
July 2019
Oct. 19
Jan. 20
Apr. 20
May 2019
Aug. 19
Nov. 19
Feb. 20
May 2019
Aug. 19
Nov. 19
Feb. 20
Jan. 2020 Feb. 20 Mar. 20 Apr. 20
Funds facing outflows saw their investors become ... adding to asset sales as well as lower performance and
more sensitive to performance and were less keen to liquidity of assets under high selling pressure compared
hold on to corporate bonds ... with other assets.
3. Quantile Regression Coefficients of Fund Flows on 4. Bid-Ask Spreads and Cumulative Returns of Securities
Returns and Corporate Bond Exposures under Selling Pressure Held by Fixed-Income Funds
(Percent) (Percent)
Return X COVID Corporate bonds Cumulative returns Mean bid-ask spread
0.6 dummy exposure 0.02 1 (selling pressure) (selling pressure) 0.85
Return 0.01 0 0.80
0.5
0.00 –1 0.75
0.4 –0.01 –2 0.70
–3 0.65
0.3 –0.02
–4 0.60
0.2 –0.03 –5 0.55
Corporate bonds Cumulative Mean bid-ask
exposure X COVID –0.04 –6 0.50
0.1 returns spread
dummy –0.05 –7 (no pressure) (no pressure) 0.45
0.0 –0.06 –8 0.40
Quantile 0.1
Quantile 0.3
Quantile 0.5
Quantile 0.7
Quantile 0.9
Quantile 0.1
Quantile 0.3
Quantile 0.5
Quantile 0.7
Quantile 0.9
Mar. 2020
Apr. 20
May 20
June 20
July 20
Mar. 2020
Apr. 20
May 20
June 20
July 20
Sources: Bloomberg Finance L.P.; Morningstar; Refinitiv; and IMF staff calculations.
Note: Panel 1 is based on 323 fixed-income funds providing information on securities held in their portfolios. The graph on the left in
panel 2 reports average shares of cash and cash equivalents in fixed-income funds with assets over $0.5 billion in extreme flow
quintiles. The graph on the right in panel 2 shows the bid-ask spread of the assets bought and sold in a given month, relative to the
bid-ask spread of the fund’s portfolio. The bid-ask spread of assets sold and bought is the average bid-ask spread in the month the
assets were sold or bought. Panel 3 reports coefficients significant at the 5 percent level from unconditional panel quantile regressions
of fund flows on portfolio shares of cash, corporate bonds, and sovereign bonds and on returns, fund size, fund age, a quarter dummy,
and a coronavirus disease dummy, as well as interactions of the latter with cash, corporate bonds, sovereign bonds, and returns and a
set of macro-financial variables, including the Chicago Board Options Exchange Volatility Index, a term spread, a credit risk spread, a
proxy for US interest levels, and a basket of major exchange rates versus the US dollar. Fund fixed effects are included. Samples include
available monthly data for fixed-income funds with assets over $0.5 billion from January 2015 to May 2020. Panel 4 is based on
detailed portfolio holdings data of 390 fixed-income funds holding approximately 13,000 identifiable securities in March 2020. Prices
and bid-ask spreads are computed based on Refinitiv composite end-of-day bid and ask prices. Pressure of security in March 2020 is
defined similarly to the definition in Coval and Stafford (2007) as the fraction of flow-motivated trading in a security’s average monthly
trading volume. Flow-motivated trading is the difference between a security’s purchases by funds experiencing higher inflows than
90 percent of their peers and the sales by funds facing outflows higher than 90 percent of their peers. The mentioned fraction defines a
security as experiencing high selling pressure if it is in the bottom decile of the ratio’s distribution across all securities; it is considered
to experience no pressure if this ratio exceeds 0.

Box 1.3. Interlinkages among Local Government, Corporate, and Bank Vulnerabilities in China
In China, debt vulnerabilities at the local gov- sharply increased differentiation based on gov-
ernment level have increased in recent years. Direct ernment direct debt loads in 2019 (Figure 1.3.1,
borrowing by local governments was first permitted panel 2, bottom-right chart). Increased government
in 2015 but has risen quickly to 24 percent of GDP, debt may weaken backstops for local SOEs and
significantly outpacing growth in local government government-backed credit guarantee institutions,
tax revenues (Figure 1.3.1, panel 1). Direct borrowing indirectly tightening financial conditions for private
growth has accelerated during the COVID-19 crisis firms, which often rely on guarantees to access credit.
as it became a key funding source for macroeconomic Non-LGFVs may also be weakened by reduced LGFV
countercyclical measures, including for investment, activity given the significant linkages between them.
spending, and even bank recapitalization. This direct Investor concerns about local government debt
debt is considered low risk by investors, reflecting may have also limited the effectiveness of authorities’
perceptions of central government guarantees. COVID-19–related credit measures in financially
Local governments also remain exposed to debt weaker provinces. Net new credit to the household
owed by off-balance-sheet entities known as local gov- and corporate sectors in the first half of 2020 was
ernment financing vehicles (LGFVs) and, indirectly, equivalent to 18 percent of 2019 GDP, but 40 percent
to debt of local government-owned enterprises (local of that increase occurred in just three provinces. Prov-
state-owned enterprises [SOEs]). LGFVs are involved inces with worse debt-to-revenue ratios saw signifi-
primarily in quasi-fiscal projects such as infrastructure, cantly weaker credit impulses than the national average
but in recent years have expanded financial linkages to (Figure 1.3.1, panel 3).
local SOEs and in some cases to private firms, in the A large proportion of LGFV and local SOE debt
form of credit guarantees and capital injections. Enti- is likely unserviceable, implying significant further
ties identifying as LGFVs in bond prospectuses have deterioration in these local fiscal backstops. Roughly
outstanding debt equivalent to 39 percent of GDP 75 percent (RMB 26 trillion) of outstanding LGFV
(Figure 1.3.1, panel 1). debt is likely unserviceable, defined as owed by LGFVs
Local governments’ growing direct debt burdens with a net-debt-to-earnings ratio of more than 15
may affect financial stability by weakening the cred- or negative earnings. Local SOEs owe another RMB
ibility of their backstop for LGFV and other local 10 trillion in similarly defined debt. If local govern-
debt. This linkage can tighten financial conditions ments assume this unserviceable debt, it will more
for the corporate sector, transmitting risks from the than double existing debt loads and increase by tenfold
government to the corporate sector, and ultimately the debt owed by provinces with debt-to-revenue
to the banking sector, which is the lender for most ratios above 400 percent (Figure 1.3.1, panel 4).
corporate debt. The potential for spillovers to banks is also consid-
Bond market data show that borrowing condi- erable. Banks are the primary creditors to LGFVs and
tions for LGFVs and lower-rated non-LGFVs appear local SOEs. If these debts develop into nonperforming
sensitive to local governments’ direct indebtedness. loans, there will be a large negative spillover effect on
With weak revenue, LGFVs rely on implicit or explicit banks’ asset quality.
government guarantees to access credit. LGFVs in Linkages between local governments, firms, and
provinces with financially weaker local governments banks could pose significant financial stability risks
have seen bond market credit spreads widen notably and underscore the urgency of accelerating structural
relative to other provinces, whereas overall debt growth reforms in China, even as authorities seek to support
has slowed or contracted (Figure 1.3.1, panel 2). the recovery from COVID-19. Key priorities should
Lower-rated non-LGFV firms appear to be sim- be to strengthen the intergovernmental fiscal coor-
ilarly affected by government debt. Province-level dination framework, introduce bank and corporate
bond market credit spreads for this segment saw restructuring frameworks in line with international
best practices, and address remaining gaps in financial
This box was prepared by Henry Hoyle. supervision and regulation.

Box 1.3 (continued)

Figure 1.3.1. Interlinkages among Local Government, Corporate, and Bank Vulnerabilities in China
Direct local government debt has been rising faster than Bigger government debt loads may weaken backstops for
indirect debt incurred via local government financing local firms, resulting in increased credit risk premiums
vehicles, outpacing growth in local tax revenues. and deleveraging for firms with weaker stand-alone debt
servicing capacity.
1. China: Government Debt by Type: Local 2. China: Selected Measures of Corporate Borrowing
Government Debt to Total Revenue Conditions, by Province Quintile
(Percent of GDP; ratio)
Financial strength quintile:
Local government 5th (strongest) 4th 3rd
LGFV credit spread

LGFV 2nd 1st (weakest) 350
(basis points)
Central government 300
LG debt to revenue (right scale)
250
100 300
200
250 150
80 2014 2015 2016 2017 2018 2019
24 80
200
Debt/GDP (2019, percent)

60 15 18
11 21
150 60
R 2 = 0.3808
40 39 R 2 = 0.3473
35 36 36 36 100 40
20 50 20
16 16 16 17 18
0 0 0
2016 17 18 19 2020:H1 –20 –10 0 10 20 30 –200 –100 0 100 200
LGFV debt growth 2019 change in weighted average
(2019, percent) credit spread for low-rated
non-LGFVs (basis points)
Policy-driven credit growth acceleration in response to the Much of the LGFV and local SOE debt local governments
COVID-19 pandemic has disproportionately benefited are exposed to is unserviceable, implying significant
provinces with more manageable government debt loads. further deterioration in backstops.
3. China: Province-Level Household and Corporate Credit 4. China: Local Government Direct Borrowing and
Growth and Ratio of Government Debt to Revenue Unserviceable LGFV and Local SOE Debt, by Ratio of
(Percent) Debt to Revenue
(Percent of GDP)
35 0–200 percent 70
200–400 percent
2020:H1 new credit/2019 GDP
30 60
400–600 percent
Total 2020:H1 600–800 percent
25 50
new credit/2019 GDP 800–1000 percent
20 18.26% 40
15 y = –0.0219x + 19.893 30
R 2 = 0.1885
10 20
5 10
0 0
0 200 400 600 800 2016 2020:H1 2020:H1 +
2019 local government debt/revenues unserviceable LGFV
and local SOE debt
Sources: Bloomberg Finance L.P.; CEIC; and IMF staff calculations.
Note: In panel 1, LGFV debt is based on financial statements of 1,852 firms with bonds designated as urban investment vehicle bonds.
2020:H1 LGFV total borrowing is estimated as the 2020:Q1 level multiplied by the 2020:Q1 quarterly growth rate. In the top chart of
panel 2, each line is a quintile of provinces based on equally weighted ranking of fiscal deficit and debt-to-GDP ratio. In the bottom
charts of panel 2, each point represents a province. Borrowing cost measures are based on weighted average bond coupons. In the
bottom-right chart of panel 2, change is the 2019 average minus the 2018 average. In panel 4, unserviceable debt is defined as debt
held by firms with a net debt to EBIT ratio above 15 (or negative earnings). Consolidated firm earnings are added to local government
revenues. EBIT = earnings before interest and taxes; LG = local government; LGFV = local government financing vehicle; SOE =
state-owned enterprise.

References ———. 2020d. “Monetary and Financial Policy Responses for

Emerging Market and Developing Economies.” IMF Special
Alcidi, Cinzia, and Daniel Gros. 2019. “Public Debt and the
Series on COVID-19, International Monetary Fund, Wash-
Risk Premium: A Dangerous Doom Loop.” Centre for Euro-
ington, DC, June 8.
pean Policy Studies, Brussels.
International Organization of Securities Commissions (IOSCO).
Coval, Joshua, and Erik Stafford. 2007. “Asset Fire Sales (and
2018. “Recommendations for Liquidity Risk Management for
Purchases) in Equity Markets.” Journal of Financial Economics
Collective Investment Schemes.” Final Report FR01/2018,
86 (2): 479–512.
The Board of the International Organization of Securities
Durham, J. Benson. 2002. “The Extreme Bounds of the
Commissions, Madrid.
Cross-Section of Expected Stock Returns.” https://ssrn.com/
———. 2019. “Recommendations for a Framework Assessing
abstract=333362
Leverage in Investment Funds.” Final Report FR18/2019,
Fitch Ratings. 2020. “Global Mutual Fund Redemption Suspen-
The Board of the International Organization of Securities
sions Highlight Liquidity Mismatches: Application of Extraordi-
Commissions, Madrid.
nary Liquidity-Management Tools Becoming More Common.”
Jin, Dunhong, Marcin T. Kaperczyk, Bige Kahraman, and Felix
https://www.fitchratings.com/research/fund-asset-managers/
Suntheim, 2019. “Swing Pricing and Fragility in Open-end
global-mutual-fund-redemption-suspensions-highlight-liquidity
Mutual Funds.” IMF Working Paper 19/227, International
-mismatches-application-of-extraordinary-liquidity-management
Monetary Fund, Washington, DC.
-tools-becoming-more-common-21-06-2020
Jordà, Òscar, Sanjay Singh, and Alan Taylor. 2020. “Long-Run
International Monetary Fund (IMF). 2020a. “Banking Sector Reg-
Economic Consequences of Pandemics.” NBER Working
ulatory and Supervisory Response to Deal with Coronavirus
Paper 26934, National Bureau of Economic Research,
Impact (with Q and A).” IMF Special Series on COVID-19,
Cambridge, MA.
International Monetary Fund, Washington, DC, May 13.
Lian, W., Andrea Presbitero, and Ursula Wiriadinata. 2020.
———. 2020b. “Considerations for Designing Temporary Liquid-
“Public Debt and r-g at Risk.” IMF Working Paper 20/137,
ity Support to Businesses.” IMF Special Series on COVID-19,
International Monetary Fund, Washington, DC.
International Monetary Fund, Washington, DC, May 8.
Panigirtzoglou, Nikolaos, and Robert Scammell. 2002. “Analysts’
———. 2020c. “Debt Management Responses to the Pan-
Earnings Forecasts and Equity Valuations.” Bank of England
demic.” IMF Special Series on COVID-19, International
Quarterly Bulletin, Spring, London.
Monetary Fund, Washington, DC, May 6.

2
CHAPTER
EMERGING AND FRONTIER MARKETS
A GREATER SET OF POLICY OPTIONS TO RESTORE STABILITY
•• To mitigate stress in local bond and currency markets, many emerging market central banks used foreign
exchange (FX) interventions and, for the first time, asset purchases.
•• This Global Financial Stability Report (GFSR) presents a novel local stress index (LSI) to measure the stress
in local bond and currency markets.
•• Asset purchase programs (APPs) helped lower government bond yields, did not lead to FX depreciation,
and eventually reduced market stress. Asset purchases may have a role to play going forward, but ongoing
evaluation of the risks is also needed.
•• Strategies to address debt distress in frontier markets need to consider the impact of the expected treat-
ment of different creditors in future debt restructurings on investor perception of risk.
The Global Pandemic Has Required Bold Action

The pandemic has hit emerging and frontier market
economies hard, but the policy response has been equally Emerging market economies have responded
strong. Policymakers have taken steps to soften the hit to forcefully to the coronavirus disease (COVID-19)
economic activity, ease financial conditions, and reduce crisis. As a result of the sudden and unprecedented
stress in domestic markets. For the first time, many emerg- shock to economic activity, most governments have
ing market central banks have launched asset purchase increased spending for emergency measures and
programs to support the smooth functioning of financial transfers (Figure 2.1, panel 1). Over 90 percent of
markets and the overall economy. Asset purchases have been central banks have cut policy rates since March—
effective in reducing bond yields and have not contributed some to all-time lows—and many have taken
to currency depreciation, but they appear to have taken measures to provide liquidity to the banking system
longer to reduce broader domestic bond market stress. This (Figure 2.1, panels 2 and 3). As a result of these
chapter examines the effectiveness of these unconventional measures and buoyant global risk appetite, financial
policy measures and concludes that asset purchases with conditions have eased considerably (see Chapter 1).
credible monetary policy frameworks and good governance This chapter discusses the historic policy responses
may be a useful addition to the policy toolkit of central of emerging market policymakers to the global pan-
banks in emerging and frontier market economies, although demic and the financial stability implications of those
a careful ongoing evaluation of associated risks is needed, policies. The “FX Intervention by Emerging Market
especially for open-ended programs. In frontier market Central Banks” section considers the use and effec-
economies, the policy focus has been on addressing the tiveness of FX interventions during the peak of the
effect of the pandemic while dealing with high debt. This crisis and reviews central bank asset purchases—a new
chapter examines the potential impact on investor percep- policy tool for emerging market economies—including
tion of sovereign risk as a result of the expected treatment an examination of their effectiveness and lessons to
of different classes of creditors in future debt restructurings. evaluate their risks in the two sections that follow.
“The Role of the Official Sector in Frontier Market
Economy Debt Restructuring” section discusses many
Prepared by staff from the Monetary and Capital Markets Depart-
ment (in consultation with other departments): The authors of this frontier market economies’ loss of market access
chapter are Dimitris Drakopoulos, Rohit Goel, Evan Papageorgiou because of COVID-19 and the potential impact of
(team leader), Dmitri Petrov, Patrick Schneider, Can Sever, and Jeff different classes of creditors on debt restructurings
Williams, under the guidance of Fabio Natalucci and Anna Ilyina.
Magally Bernal and Andre Vasquez were responsible for word pro- and on investor perception of sovereign risk. Build-
cessing and the production of this report. ing on the findings of the chapter, the final section

Figure 2.1. Emerging Market Policy Response to the COVID-19 Pandemic
The need for emergency spending and the hit to revenues from the ... and most central banks have aggressively cut rates, some to
sharp economic shock of the COVID-19 crisis increased budget all-time lows.
deficits ...
1. General Government Deficit 2. Monetary Policy Rates
(Percent of GDP) (10th–90th percentile range, percent)
20 10th–90th percentile Policy rate as of end-2019 18
2020 deficit forecast in Oct. 2019 range from 2005–19 Latest 16
Change in forecast of 2020 deficit 14
15
2019 deficit
12
10
10
8
6
5 4
2
0 0
Brazil
Ghana
South Africa
India
China
Poland
Romania
Colombia
Peru
Chile
Hungary
Philippines
Turkey
Egypt
Indonesia
Thailand
TUR
MEX
RUS
IDN
IND
ZAF
PHL
CHN
BRA
COL
MYS
HUN
THA
CHL
PER
POL
In addition to rate cuts, central banks have responded forcefully to the COVID-19 crisis with an array of measures to boost market liquidity and
stabilize economic and financial conditions.
3. Central Bank Policy Actions
(Number of central banks on y-axis; percent of sample in brackets)
50
Latin America and the Caribbean (9)
45 MENA and Central Asia (9)
40 Emerging Europe (12)
35 Asia (11)
30 Sub-Saharan Africa (9)
25
92%
20
15 58% 52%
10 44%
34%
5
8%
0
Policy rate change FX intervention Reserve requirement Term repo Government bond Private security
ratio change purchase program purchase program
Sources: Bloomberg Finance L.P.; IMF, World Economic Outlook database; national authorities; and IMF staff calculations.
Note: In panel 3, countries are counted only once per action (for example, multiple policy rate cuts are counted once). The sample comprises 50 central banks and
does not include any advanced economies. The sample is defined in Online Annex 2.1 and is quantified by region in parentheses. Data labels in panel 2 use
International Organization for Standardization (ISO) country codes. FX = foreign exchange; MENA = Middle East and North Africa.
offers policy recommendations. The apparent absence economies from external movements in the pricing
to date of capital flow management measures during of risk. While many countries intervened, surpass-
the COVID-19 crisis and China’s policy challenges in ing recent stress episodes in absolute size (Figure 2.2,
maintaining supportive financial conditions are briefly panel 1), the use of reserves (as a share of total inter-
examined as well (Online Annex Boxes 2.1 and 2.2). national reserves) was about two-thirds the magnitude
observed during the global financial crisis for the
median country (Figure 2.2, panel 2). The limited and
FX Intervention by Emerging Market short-lived use of reserves can potentially be attributed
Central Banks to a relatively short duration of the stress episode due
FX interventions, including in some cases through to a quick turnaround in global risk sentiment, which
forward contracts, were widespread at the height of the has also likely reduced the need for the capital flow
crisis in March, as policymakers sought to insulate their management measures (see Online Annex Box 2.2).

CHAPTER 2 Emerging and Frontier Markets: A Greater Set of Policy Options to Restore Stability
IMF staff analysis shows that global factors, includ- ductive, primarily because of fears over portfolio
ing Federal Reserve rate cuts and global risk appetite outflows and ineffective policy transmission.
(proxied by the Chicago Board Options Exchange •• Central banks with policy rates closer to their lower
Volatility Index [VIX]1), played a significant role in bound (Chile, Hungary, Poland) have partially
driving currency surprises2 during the COVID-19 sought to use asset purchase programs for some-
sell-off (Figure 2.2, panel 3). Domestic policy rate what similar reasons as advanced economies, to ease
cuts and FX interventions, on the other hand, had a financial conditions, provide additional monetary
relatively insignificant impact. This contrasts with the stimulus, and exert greater influence on longer
2015 sell-off, which was more specific to emerging maturity bond yields. It is worth noting that in
markets and not driven by exogeneous global shocks, most cases market functioning and liquidity objec-
and during which emerging market currencies were tives were prominently featured.
significantly affected by domestic FX interventions and •• Some central banks explicitly stated that one of their
policy rate cuts (Figure 2.2, panel 4). objectives was to temporarily ease government financ-
ing pressure in the face of the once-in-a-generation
global pandemic (Ghana, Guatemala, Indonesia, and
The New Game in Town: Central Bank the Philippines through its repurchase agreement).
Asset Purchases
During the COVID-19 crisis, for the first time Central bank purchases of government securities
on a broad basis, at least 18 emerging market central played an important role in some domestic bond
banks adopted unconventional policies through the markets during the acute phase of the sell-off. Begin-
use of asset purchase programs3 targeting government ning in February 2020 (Figure 2.3, panel 2), almost all
or private sector bonds in local currency. In several economies faced sizable local currency bond outflows.
cases the purchases were sterilized, which alleviated Central bank asset purchases varied substantially in size,
downward pressure on exchange rates. The scope and but in most cases they helped the domestic investor base
motivation of these programs varied across economies absorb much of the outflow pressure and deal with the
(see Table 2.1 and Figure 2.3, panel 1), and the objec- government’s increased financing needs. For example,
tives were often multifaceted, but a view toward the in Poland between the end of February and June the
available conventional monetary policy space allows for central bank purchased more than 2 percent of GDP in
the identification of three broad groups: government bonds in the secondary market compared
•• Central banks with policy rates well above zero tended with outflows of 0.7 percent of GDP, alongside an
to use asset purchase programs as a tool to improve increase in net domestic issuance of 4.4 percent of GDP.
bond market functioning (India, Philippines, South In some countries that did not launch asset purchase
Africa) and provide liquidity to the financial sector. programs, debt management offices limited the local
In some cases, central banks may have seen nominal bond supply to avoid further deterioration of already
policy rates below a certain level as counterpro- stressed local bond markets. Instead, they relied on alter-
native sources of financing (for example, the use of cash
buffers in Brazil, increased external issuance in Mexico,
1Other policy variables, such as announcements by the Federal
and pension funds in some Latin American countries) or
Reserve of additional purchases, credit facilities, and swap lines,
must have also affected emerging market currencies indirectly, but back-loaded issuance to the second half of the year.
a significant part of that impact should be reflected through global
risk appetite.
2The results are broadly consistent even when simple currency
Local Market Stress Is Greater in Bonds than
changes are considered. For more details, see Online Annex 2.1.
All annexes are available at www.imf.org/en/Publications/GFSR.
in Currencies
3For the purpose of this GFSR, an APP is the expansion of the
This GFSR introduces a novel market conditions
central bank balance sheet via purchases of various type of securities.
APPs include quantitative easing programs that aim to ease financial
index designed to assess the level of stress in local
conditions and provide monetary stimulus, more limited programs bond and currency markets. The local stress index (LSI)
that aim to improve market functioning, and purchases in primary summarizes conditions into an indicator that can help
markets that aim to assist with government financing requirements.
guide central bank decisions regarding the need for
Some countries in the sample set up new purchase programs (for
example, Chile and Hungary); others adjusted their existing open interventions to support local market functioning.
market operations (for example, Malaysia and Turkey). Unlike financial conditions indices, which can loosen or

Figure 2.2. FX Interventions and Reserve Operations
Reserve operations were substantial and widespread in dollar terms ... ... though as a share of reserves they never reached the level of the
global financial crisis and receded quickly.
1. Reserve Operations by Region 2. Reserve Operations
(US dollars) (Share of reserve stock, three-month rolling sum)
50 12
IQR Median
25 8
4
0
0
–25
–4
–50
Renminbi –8
GFC, –4.4 Euro crisis, depreciation, COVID-19,
–75 Asia Latam EMEA –12
–2.3 –2.3 –2.9
–100 –16
2013 14 15 16 17 18 19 20 2006 07 08 09 10 11 12 13 14 15 16 17 18 19 20
Global factors played a significant role in driving emerging market ... in sharp contrast to the 2015 emerging market sell-off, when
currency surprises during the COVID-19 sell-off ... domestic factors played a significantly more important role.
3. Coefficients for the Drivers of the EM FX Surprise during the 4. Coefficients for the Drivers of EM FX Surprise during the
COVID-19 Sell-off (January 2020–May 2020) 2015 EM Sell-off (April 2015–February 2016)
2.0 10
8
1.0
6
0.0 4
2
–1.0 0
–2
–2.0
–4
–3.0 –6
FXI Policy VIX US FFR VIX US FFR VIX US FFR FXI Policy VIX US FFR VIX US FFR VIX US FFR
rates * FXI * FXI * PR * PR rates * FXI * FXI * PR * PR
Domestic Global Interaction Terms Domestic Global Interaction Terms
Sources: Data set from Adler and others (forthcoming); Bloomberg Finance L.P.; Haver Analytics; International Institute of Finance; and IMF staff calculations.
Note: In panel 1, data exclude China. In panels 1 and 2, data are as of end-August 2020. Data from May onwards include estimates for operations only in the spot
market, while data for April and earlier include estimates for operations in spot as well as derivatives markets. Operations in derivatives markets do not represent a
drag on the reserve stock but are included in the calculations to estimate the size of the intervention. These estimates do not adjust for foreign exchange bond
sales/purchases, so they may represent a partial picture in a few cases (for example, Mexico). In panels 3 and 4, the sample consists of 14 emerging markets with
panel data at monthly frequency (see Online Annex 2.1 for more details). The dependent variable is the forecast error between the spot currency value and the value
forecast by the previous month’s forward contracts. A positive value implies that the currency appreciated versus market expectations, assuming parity holds. In
reality, the forward values might vary from spot for an extended period of time, but the changes in this metric will still highlight currency pressures, albeit only
partially. The results hold broadly true even if the dependent variable is taken as foreign exchange appreciation. Foreign exchange intervention (FXI) is calculated as
valuation-adjusted changes in reserves and the intervention as taken in the derivative markets. A positive value means active intervention. Country fixed effects are
included. Coefficient estimates are shown with two standard error confidence intervals. In panels 3 and 4, blue bars are the statistically significant coefficients, while
gray bars are not statistically significant. EM = emerging market; EMEA = Europe, Middle East, and Africa; FFR = Federal funds rate (effective); GFC = global financial
crisis; IQR = interquartile range; Latam = Latin America; PR = policy rate; VIX = Chicago Board Options Exchange Volatility Index.

Table 2.1. Asset Purchase Programs in Response to COVID-19 in Emerging Market Economies
Total Program Significant General Government
Target or Limit Size Purchases Duration Program Government Debt
(local currency (percent of (observed Announcement 2020 Deficit (percent of
Country Primary Objectives Asset Type unless specified) Market GDP) or explicit) Dates (percent of GDP) GDP)
Colombia Provide liquidity to the financial sector Government, private 10 tn private, up to Secondary 1.1 Mar.–Apr. Mar. 23 –9.5 68.2
sector bonds 4 tn government
Chile Facilitate monetary policy transmission, Bank, central bank, and $16 bn Secondary 2.9* Mar.–present Mar. 16, Mar. 31, –8.7 32.8
ease financial conditions government bonds* Jun. 16, Aug. 12
Croatia Stabilize domestic bond market Government bonds Not specified Secondary 4.9 Mar.–Jun. Mar. 13 –8.1 87.7
Ghana Finance budget deficit Government bonds 5.5 bn (up to 10 bn) Primary 1.4 May May 15 –16.4 76.7
Guatemala Finance budget deficit Government bonds 11 bn Both 1.9 Apr.–Aug. Apr. 8 –5.6 32.2
Hungary Facilitate monetary policy transmission at Government, mortgage No upper limit, but Both 1.4 May–present Apr. 7, Apr. 28, –8.3 77.4
longer maturities, provide financial sector bonds (MBs) technical revision after (only MBs Jul. 21, Aug. 25
CHAPTER 2
liquidity 1 tn in government, in primary)

300 bn in MBs
India Stabilize domestic bond market Government bonds Not specified Secondary 1.0 Mar.–present Mar. 18 –13.1 89.3
Indonesia Stabilize domestic bond market, provide Government bonds Initially not specified, Both 3.8** Mar–present Mar. 31, Jul. 7 –6.3 38.5
liquidity to the financial sector, finance with direct “burden
budget deficit sharing” of 397.6 tn
later announced
Malaysia Provide liquidity to financial sector Government bonds Not specified Secondary 0.6 Mar.–Jun. Mar. 25 –6.5 67.6
Philippines Provide liquidity to financial sector, stabilize Government bonds, Secondary market Both 4.3 (7.3)*** Mar.–present Mar. 23, Apr. 10, –8.1 48.9
bond market, finance budget deficit including repurchase purchases not Oct. 2
(repurchase agreement) agreement specified, repurchase
amount limited to
about 850 bn
Poland Strengthen monetary policy transmission at Government, SOE Not specified Secondary 4.6 Mar.–present Mar. 17, Apr. 8 –10.5 60.0
longer maturities, stabilize domestic bond bonds
market, provide liquidity to financial sector
Romania Provide liquidity to financial sector Government bonds Not specified Secondary 0.5 Mar.–present Mar. 20 –9.6 44.8
South Stabilize domestic bond market Government. bonds Not specified Secondary 0.7 Mar.–present Mar. 25 –14.0 78.8
Africa
Thailand Stabilize domestic bond market Government, central Not specified Secondary 1.0 Mar.–Apr. Mar. 19, Mar. 22 –5.2 50.4
bank bonds
Turkey Provide liquidity to financial sector, Government bonds Not specified, but Secondary 1.6 Mar–present Mar. 31 –7.9 41.7
strengthen monetary policy transmission OMO portfolio
mechanism, secure credit conditions limited to 10 percent
of balance sheet
Sources: Local media; national authorities; and IMF staff estimates.
Note: Total purchase amounts are estimates of March through latest available as of publication process (late September). Program dates are not exhaustive, but generally reflect a significant program announcement or first purchase
date. Poland includes purchases of bonds from the State Development Bank (BGK) and State Development Fund (PFR). For Chile, only assets purchased under the Special Asset (June) and Bank Bond (March) Purchase Programs
that were in direct response to the COVID-19 crisis were included, and not the Nov. 2019 central bank debt buyback program through which the central bank bought back about 1.2% of GDP of its own debt in 2020. Bank of Thailand
International Monetary Fund | October 2020

also authorized a Corporate Stabilization Fund for short-term financing not included here. Papua New Guinea, Jamaica, Sri Lanka, and the Central African Economic and Monetary Community (through the Bank of Central African
States) are not included in the table but announced asset purchases of various forms. Brazil outlined plans to purchase corporate bonds in June, but had yet to do so. The BSP (Philippines) opened its purchase window in March
prior to written public announcement in April. bn = billion; OMO = open market operations; tn = trillion.
*Chile’s central bank did not gain the legal ability to purchase government bonds until August 12.
Emerging and Frontier Markets: A Greater Set of Policy Options to Restore Stability
39
**Indonesia includes staff estimates of secondary market purchases, primary market purchases prior to July, and the full 397.6 tn July burden sharing agreement, though only about 60 percent of the agreed purchases had been
completed through mid-September.
***Philippines includes staff estimates of secondary market purchases and the three-month repurchase agreement of 540 bn (3.0% of GDP) with the central government added in parentheses, and the BSP closed out a previous
300 bn repurchase agreement in September.
Figure 2.3. Central Bank Asset Purchases in Emerging Markets
Asset purchase programs in emerging markets differ in scope, size, Central bank purchases helped offset portfolio outflows during the
and duration from those in advanced economies and are often used crisis period in some economies.
with higher policy rates.
1. Central Bank Asset Purchases through August 2. Change in Local Currency Government Bond Holdings,
(Percent of GDP) end-February–June 2020
(Percent of 2020 GDP)
8
GHN Other domestic
< 15
TUR Non-residents
Policy rate at time of APP announcement (percent)
ZAF Domestic banks

Sovereign primary market Central bank 6
ROU
Sovereign secondary market Total change in LC debt outstanding
MYS
<6
ABS/Covered bonds
IND
Private/ETF 4
IDN Other
PHP
HUN
2
THA
≤1
COL
CHL
0
POL
HRV
EA Advanced
–2
< 0.5
JPN economies
US
UK
–4
0 2 4 6 8 10 12 14 POL ZAF IDN ROU HUN BRA COL MEX
Percent of GDP
Sources: Bloomberg Finance L.P.; Haver Analytics; national sources; World Bank; and IMF staff calculations.
Note: Data in panel 1 and panel 2 may in some cases have different sourcing related to definitional and availability reasons. Asset purchases in Hungary did not begin
until May. In panel 1, sovereign purchases for Poland include those from the state development bank (BGK) and the state development fund (PFR), which are
excluded in panel 2. Purchases for Chile include only those under Special Asset (June) and Bank Bond (March) Purchase Programs. Primary market purchases for the
Philippines refer to the 300 bn (~1.6% of GDP) repurchase agreement in April 2020, which was repaid in September. In panel 1, Indonesia primary market purchases
include only the share of the burden sharing agreement completed through August, not the entirety of the 397.6 trillion plan. In panel 2, total change for South Africa
differs slightly from aggregated holdings as it includes Treasury bills separately. Data are not adjusted for inflation-linked bonds. Indonesia central bank holdings are
defined as net of monetary operations. Data labels use International Organization for Standardization (ISO) country codes. ABS = asset-backed securities;
APP = asset purchase program; ETF = exchange-traded fund.
tighten as a result of changes in policy rates or external developments in the global financial markets. In line
spreads—as a reflection of the cost of funding—the LSI with past episodes of sharp tightening in global financial
focuses on local market liquidity and stress indicators conditions, the spillovers in FX markets emanating from
(such as bid-offer spreads, realized volatility, and other the United States and the European Union rose sharply
risk premium measures).4 (Figure 2.4, panel 3) as currencies played their role as
The level of stress in local markets during the shock absorbers.5 However, unlike what happened during
COVID-19 sell-off, as measured by the LSI, was com- past tightening episodes, the spillovers to local bond
parable to that of the global financial crisis, but the markets were more pronounced (Figure 2.4, panel 4).
period of stress was considerably shorter. In aggregate Most emerging markets have seen a large increase in non-
(Figure 2.4, panel 1), the level of stress was well above resident participation in their local bond markets since
that of previous episodes, such as the 2013 taper tan- the global financial crisis, which may have exacerbated
trum and 2014–15 stress episodes. However, markets increased volatility spillovers during the recent sell-off.
have been normalizing much faster than during previ- The stress in FX markets was lower than during
ous episodes (Figure 2.4, panel 2). 2008–09, with less noticeable demand for dollar liquidity.
A large part of the increase (and subsequent partial
reduction) in stress in local bond markets originated from 5Spillover indices in Figure 2.4, panel 1, are calculated using the
approach in Diebold and Yilmaz (2012), in which time-varying spill-

overs are constructed using rolling generalized forecast error decom-
4For details, see Online Annex 2.1, available at www.imf.org/en/ positions. The index is the contribution from a shock to market X to
Publications/GFSR. the overall variability in any other market Y.

Figure 2.4. Stress in Local Currency Bond and FX Markets
The COVID-19 shock led to significant market dysfunction comparable Stress dissipated faster than in previous episodes but remains
to that of the 2008 global financial crisis. elevated.
1. Emerging Market Local Stress Index 2. Emerging Market Local Stress Index
(Index) (Dates in parentheses correspond to day = 0 on x-axis)
0.6 FX LSI Bond LSI LSI LSI (Sep. 2008) LSI (May 2013) 0.6
0.5 LSI (July 2011) LSI (Feb. 2020)
Global COVID-19
0.4 financial crisis Euro area crisis BRA loss 0.4
Taper of IG
0.3 CNY
tantrum ARG/TRY FX
0.2 depreciation 0.2
pressures
0.1
0.0 0.0
2005 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20
–50
–40
–30
–20
–10
0
10
20
30
40
50
60
70
80
90
100
110
120
130
140
150
160
170
Number of days
The spillovers of tightening US/EU financial conditions to emerging ... while the spillover to emerging market bond market conditions is far
market currencies were of the same magnitude as in the past ... more pronounced now than in the past.
3. Bilateral Net Spillovers from the United States and the Euro Area to 4. Bilateral Net Spillovers from the United States and the Euro Area to
the FX LSI the Local Bond LSI
(Index) US FCI spillovers to FX LSI (Index) US FCI spillovers to Bond LSI
EU FCI spillovers to FX LSI EU FCI spillovers to Bond LSI 16
12 12
8 8
4 4
0 0
–4 –4
–8 –8
–12 –12
2008 09 10 11 12 13 14 15 16 17 18 19 20 2008 09 10 11 12 13 14 15 16 17 18 19 20
Policy actions in FX markets normalized conditions quickly, but ... ... local bond markets have remained more dysfunctional, triggering
asset purchase programs.
5. FX LSI and Components 6. Local Bond LSI and Components
(Index) (Index)
Correlation Realized volatility Risk reversals Bid-offer Curve term premium Realized volatility
Implied volatility FX basis Bid-Offer Asset Swap spreads Volumes (flows) Correlation
1.0 FX LSI Bond LSI 1.0
0.8 Global financial crisis 0.8
0.6 0.6
0.4 0.4
0.2 0.2
0.0 0.0
–0.2 –0.2
–0.4 Global financial crisis COVID-19 COVID-19 –0.4
–0.6 –0.6
Aug. Nov. Feb. May Dec. Mar. June Sep. Aug. Nov. Feb. May Dec. Mar. June Sep.
2008 08 09 09 2019 20 20 20 2008 08 09 09 2019 20 20 20
Sources: Bloomberg Finance L.P.; and IMF staff calculations.

Note: The local stress index (LSI) is calculated from the country LSIs of 16 countries. For more information see Online Annex 2.1. FCI = financial conditions index;
FX = foreign exchange; GFC = global financial crisis.

For example, increases in measures such as risk reversals, Domestic Asset Purchases Eventually Helped Reduce
which indicate the level of hedging demand for a sharp Market Stress
depreciation against the dollar, have been more muted.6 The announcement of asset purchase programs in
In addition, the wider cross-currency basis—a measure the second half of March did not have an immediate
of dollar funding liquidity stress (Figure 2.4, panel 5)— impact on local stress indices, given that global finan-
was more short-lived. These developments were likely cial conditions were very tight and market conditions
a result of: were hampered by illiquidity, strong risk aversion,
•• The rapid establishment of central bank swap line and fiscal concerns (Figure 2.5, panel 1).9 However,
facilities and bond repo facilities for foreign central as external conditions started to improve in April and
banks by the Federal Reserve and the European countries stepped up implementation of asset purchase
Central Bank.7 programs, country-level local stress indices showed
•• Structural shifts in the operation of FX markets some improvement and differentiation.10 A large part
since the global financial crisis (Schrimpf and of the improvement was seen in market liquidity
Sushko 2019),8 including increased turnover in measures, such as bid-offer spreads and a reduction in
emerging market currencies and electronic trading intraday volatility. Yet term premiums in some local
and a larger set of market-making institutions. bond markets remain elevated as investors are facing
bond supply risks over a longer horizon given the
Unlike FX markets, local bond markets became uncertainty of pandemic-related government financing
more stressed and triggered policy responses in the requirements.
form of asset purchase programs. A notable aspect is Evaluating the effectiveness of asset purchase
the increase in the risk premiums of long-end govern- programs with respect to their stated goal of improv-
ment bonds relative to short-end bonds and onshore ing market conditions is complex, and more work is
swap rates (Figure 2.4, panel 6). Despite the positive needed. Asset purchase programs helped reduce market
impact of asset purchase programs on market condi- stress, eventually, and several factors contributed to
tions (see next subsection), stress levels have been more this reduction. The size of announced asset purchase
elevated, likely as a result of: programs in emerging markets has been small over-
•• High local bond supply risks that weigh on yields all (except in Chile, Indonesia, the Philippines, and
through risk premiums. Poland) and short-lived, as is evident in the slow-
•• Weak foreign flows to local bond markets, which down of asset purchases since May for most countries
had a negative impact on liquidity. (Figure 2.5, panel 2). In addition, announcements and
•• Relatively limited depth of local currency government implementations of asset purchase programs can affect
bond markets. Unlike FX markets, local bonds are market conditions differently, and the lack of local
still traded largely domestically, and market depth currency bond inflows had a negative impact on market
has not matched higher foreign participation, which liquidity, especially in markets with a large foreign
could induce volatility (see Chapter 3 of the April presence. The introduction of asset purchase programs
2020 GFSR). In countries with a shallower domestic at the height of the crisis is likely to have served as a
investor base (see “Looking Ahead: Trade-offs of Asset useful circuit breaker, preventing further escalation of
Purchase Programs” section), domestic banks are the stress. Purchases of government bonds and other assets
sole liquidity providers in times of stress. signaled that emerging market central banks were ready
to stand as buyer of last resort (Arslan, Drehmann,
6In fact, during the early stages of the shock in February, the
depreciation pressures in emerging markets were more acute against

and Hofmann 2020). Moreover, the empirical analysis
the euro, likely because of unwinding of euro-funded carry trades presented in the following section suggests that asset
relative to high-yield currencies, such as the Russian ruble and the
Mexican peso.
7The IMF flexible credit lines for Chile and Peru in the second 9This is in line with developments in the United States, where the
quarter of 2020, and the renewal of the flexible credit line for Federal Reserve’s March 15 announcement of additional US Treasury
Colombia, also boosted confidence and provided insurance against purchases did not relieve market stress.
downside risks. 10Figure 2.5, panel 1, aggregates countries that have different
8Another structural shift worth noting is the shift toward more characteristics, which could be the main driver of the results rather
flexible exchange rate regimes since the 2008 global financial crisis than APPs. Online Annex 2.1 presents event studies around the asset
(for example, in Russia). purchase announcements that show country-level developments.

Figure 2.5. Bond Stress and Asset Purchase Programs in Emerging Market Economies
Stress has eased somewhat faster for countries with asset purchase Emerging market asset purchases rose significantly in March and April
programs than for those that do not have them. but moderated thereafter.
1. Local Stress Indices: APP versus Non-APP Economies 2. Asset Purchases by Major EM Central Banks
(Billions of US dollars)
0.7 APP sample: FX LCSI POL IND 45
APPs announced in most TUR IDN
APP sample: Bond LCSI
countries in sample ZAF COL 40
Non-APP sample: FX LCSI
0.6 HRV THA
Non-APP sample: Bond LCSI
HUN CHL 35
MYS PHP
0.5
ROU 30
0.4 25
0.3 20
15
0.2
10
0.1
5
0 0
Jan. Feb. Mar. Apr. May June July Aug. Sep. March April May June July August
2020 20 20 20 20 20 20 20 20
Sources: Bloomberg Finance L.P.; Haver Analytics; JPMorgan Chase and Co.; national authorities; and IMF calculations.
Note: Non-APP economies are Brazil, Chile, China, Mexico, Peru, and Russia. In panel 2, Indonesia uses change in gross holdings as proxy for asset purchases.
Monthly purchases are IMF staff estimates otherwise. Data labels in panel 2 use International Organization for Standardization (ISO) country codes. APP = asset
purchase program; EM = emerging market; FX = foreign exchange; LCSI = local currency stress index.
purchase program announcements had a positive impact relatively limited impact on currencies (Figure 2.6,
on yields on the announcement date and several days panel 3). The reaction seen in intraday data for selected
beyond, even after controlling for external factors. Nev- countries—to control for the effect of global and exog-
ertheless, large-scale APPs, especially when open-ended, enous factors on end-of-day levels—shows a similar
carry risks and may negate their initial effectiveness. trend, with declining government bond yields but rel-
atively less impact on currencies (Figure 2.6, panel 4;
Arslan, Drehmann, and Hofmann 2020).
Domestic Asset Purchases Lowered Bond Yields and Had This section discusses empirical analysis of the effect
Little Effect on Currencies of domestic asset purchase program announcements
Event studies show that asset purchase program on local currency government bond yields.13 The
announcements11 had a significant immediate impact model controls for policy rate cuts by emerging market
on asset prices and helped turn sentiment around.12 central banks and global factors, such as the VIX and
Financial conditions were tightening going into the the VIX rate of change and asset purchase program
announcements but were inflected following the announcements by the Federal Reserve. The analysis
announcements, with a corresponding sharp reduc- uses daily data from 13 emerging market economies
tion in government bond yields (Figure 2.6, panel 1) from January to mid-May 2020 and controls for
and term premiums (Figure 2.6, panel 2), but with unobserved country-specific factors using country
fixed effects (see Online Annex 2.1). The analysis is
based on the local projections method (Jordà 2005;
11The size of the announced programs may also have influenced
Teulings and Zubanov 2014), which capture the full
the market reaction, although it is not considered (in line with the
literature) given the lack of consistency across announcements and dynamics of sovereign bond yields in the aftermath of
divergent market expectations.
12Results in this section draw upon Drakopoulos and others 13Drakopoulos and others (forthcoming) discusses also the effect
(forthcoming). of APPs on equity markets.

Figure 2.6. Market Reaction to Domestic Asset Purchase Program Announcements
Event studies around emerging market asset purchase program announcements show a significant change following the event, including a decline
in sovereign bond yields and a decline in term premiums, but a relatively small and short-lived impact on EM currencies.
1. EM 10-Year Government Bond Yields 2. EM 5-Year ACM Term Premiums 3. EM Currencies
(Indexed at 0 on t = 0; basis points; (Indexed at 0 on t = 0; basis points; (Indexed at 1 on t = 0; days on x-axis)
days on x-axis) days on x-axis)
40 10 1.02
20 0 1.01
0 –10 1.00
–20 –20 0.99
–40 –30 0.98
–60 –40 0.97
APP APP APP
–80 announcements announcements –50 announcements 0.96
–100 –60 0.95
t – 10 t–5 t t+5 t + 10 t – 10 t–5 t t+5 t + 10 t – 10 t–5 t t+5 t + 10
Intraday price reaction showed a similar trend, with government yields reacting very sharply, but relatively limited impact on emerging market
currencies.
4. Intraday Asset Price Movement on the Days of Asset Purchase Program Announcements for Selected EMs (South Africa, India, and Poland)
Bond yields (left scale) Bond yields (left scale) Bond yields (left scale)
FX (right scale) FX (right scale) FX (right scale)
50 1.02 5 1.002 5 1.01
0
0 1.001 0
1.01 1.00
–50
–5 1.000 –5
–100
1.00 0.99
–10 0.999 –10
–150
–200 0.99 –15 0.998 –15 0.98

t–1 0 t+1 t+2 t–1 0 t+1 t+2 t–1 0 t+1 t+2
Sources: Bank for International Settlements; Bloomberg Finance L.P.; BNP Paribas; national authorities; and IMF staff calculations.
Note: In panel 2, term premium calculations are based on the methodology detailed in Adrian, Crump, and Moench (2013). In panels 3 and 4, a declining trend in the
foreign exchange lines implies an appreciation of the local currency versus the US dollar. In panels 1–3, the black line denotes the median across our sample, while
the blue range highlights the interquartile range across the events. The sample comprises Chile, Colombia, Hungary, India, Indonesia, Malaysia, the Philippines,
Poland, South Africa, and Turkey (across a total of 16 dates). ACM = Adrian, Crump, and Moench (2013); APP = asset purchase program; EM = emerging market;
FX = foreign exchange.
the announcements by central banks.14 The depen- are presented to account for the direct effect of the
dent variable is the cumulative change in bond yields, additional asset purchase announcement by the Federal
and the main variable of interest is the indicator for Reserve (Figure 2.7, panels 1, 3, and 5) and the VIX
the dates of asset purchase program announcements as a proxy for global risk appetite (Figure 2.7, panels 2,
(Figure 2.7). A challenge in this analysis is to isolate 4, and 6). Both specifications control for domestic
the impact of asset purchase program announcements policy rates.
on bond yields from the effect of policy rate cuts and Both specifications find that emerging market
announcements by the Federal Reserve around the central bank asset purchase program announcements
same time. To that end, two empirical specifications reduce long-end bond yields in a significant and
persistent way (Figure 2.7, panels 1 and 2), even con-
14Some evaluations of the effectiveness of asset purchases by the
trolling for the Federal Reserve asset purchase program
Federal Reserve use the surprise announcement of 10-year equiva-
lents on term premiums, but such an approach is beyond the scope announcement (Figure 2.7, panel 1) or the change
of the analysis here. in global risk appetite (Figure 2.7, panel 2). The size

Figure 2.7. Asset Purchase Program Announcements and Sovereign Bond Yields
Panels 1, 3, and 5 show the impulse response functions to APP Panels 2, 4, and 6 show the impulse response functions of APP
announcements by emerging market central banks, controlling for announcements by emerging market central banks, controlling for the
Federal Reserve actions and emerging market rate cuts. VIX as a proxy for global risk appetite and emerging market rate cuts.
Specification 1: Effect of Variable X on Bond Yields Specification 2 : Effect of Variable Y on Bond Yields
1. X = Domestic APP Announcements 2. Y = Domestic APP Announcements
(Percentage point change in yield) (Percentage point change in yield)
0.2 0
0 –0.2
–0.2 –0.4
–0.4 –0.6
–0.6 –0.8
0 1 2 3 4 5 6 0 1 2 3 4 5 6
Days since announcement Days since announcement
3. X = Federal Reserve Quantitative Easing Announcement 4. Y = Ten Point VIX Increase

0.2 0.2
0.1
0.1
0
–0.1
0
–0.2
–0.3 –0.1
0 1 2 3 4 5 6 0 1 2 3 4 5 6
Days since announcement Days since increase
5. X = 1 Percentage Point Domestic Policy Rate Cut 6. Y = 1 Percentage Point Domestic Policy Rate Cut
0.4 0.4
0.2 0.2
0 0
–0.2 –0.2
–0.4 –0.4
–0.6 –0.6
0 1 2 3 4 5 6 0 1 2 3 4 5 6
Days since cut Days since cut
Source: IMF staff calculations.

Note: Results are based on the local projections method (Jordà 2005; Teulings and Zubanov 2014) using panel data from 13 emerging markets at daily frequency
from the beginning of January to mid-May 2020. The dependent variable is the cumulative change (in percentage points) in local currency sovereign bond yields. The
first specification controls for the APP announcement by the Federal Reserve and domestic rate cuts (panels 1, 3, and 5). The second specification controls for the
Chicago Board Options Exchange Volatility Index (VIX) and domestic rate cuts (panels 2, 4, and 6). Country fixed effects are included in both specifications. Coefficient
estimates are reported with one standard error confidence interval. The x-axes represent the number of trading days following each episode. See Online Annex 2.1
for more details. APP = asset purchase program; VIX = Chicago Board Options Exchange Volatility Index.

Figure 2.8. Asset Purchase Program Announcements and (Figure 2.7, panel 4) had a positive effect on decreas-
Domestic Currencies
ing yields, reflecting the sensitivity of domestic bond
Announcements of asset purchase programs did not lead to a yields to global factors (April 2020 GFSR). This is also
significant depreciation of emerging market currencies. consistent with the finding by Beirne, Renzhi, and
Emerging Market Currencies Sugandi (2020) of evidence of spillovers to emerging
(Percent change)
2 market bond yields from quantitative easing by central
banks in advanced economies (see Chapter 1). The
magnitudes of the effect of the asset purchase program
announcements by emerging market central banks and
the Federal Reserve are broadly similar.
1 Announcements of asset purchase programs did not
lead to a significant depreciation of emerging market
currencies (Figure 2.8), in line with intraday event
studies (Figure 2.6, panel 4). This may reflect the rel-
atively small size of the programs and the fact that the
0
purchases were sterilized in many cases. Furthermore,
the restoration of stability and the decisive actions
taken by advanced and emerging market central banks
may have also contributed to investor confidence and
–1 reversal of the earlier considerable FX sell-off.
0 1 2 3 4 5 6
Days since announcement
Looking Ahead: Trade-offs of Asset
Note: Results are based on the local projections method (Jordà 2005; Teulings and
Purchase Programs
Zubanov 2014) using panel data from 13 emerging markets at daily frequency The experience with emerging market asset pur-
from the beginning of January to mid-May 2020. The dependent variable is the
cumulative change (in percent) in the value of domestic currencies vis-à-vis the chase programs has been largely positive so far, though
US dollar. The specification controls for the asset purchase program (APP) further expansion of duration or size could create
announcement by the Federal Reserve and domestic rate cuts, as well as country
fixed effects. Coefficient estimates are reported with a one standard error risks and thus warrant an ongoing evaluation of risks.
confidence interval. The x-axis shows the number of trading days following each APPs had a catalyzing effect on lowering local cur-
episode. See Online Annex 2.1 for more details.
rency government bond yields without indications of
immediate risks to financial stability. In some cases,
of the impact of domestic asset purchase program purchases may have intermediated an orderly exit of
announcements on yields ranges from 20 to 60 basis investors from local currency bond markets, but this
points and is statistically significant within one stan- was likely done in the interest of preserving investor
dard error confidence interval. The size of the effect confidence and avoiding more costly and widespread
is in the range of Arslan, Drehmann, and Hofmann market disruptions. Central bank communication
(2020) and Hartley and Rebucci (2020). By con- and benign market perception in terms of the scope,
trast, in both specifications, domestic rate cuts do not timing, and temporary nature of these programs were
appear to have a significant effect on yields, controlling essential in containing perceived risks of fiscal domi-
for other factors, such as asset purchase programs15 nance that would likely have led to higher bond yields
(Figure 2.7, panels 5 and 6). and weaker currencies.
The improvement in external conditions also had a Beyond the pandemic, this positive experience
significant and persistent impact on lowering long-end may motivate more emerging market central banks
yields. Both the Federal Reserve asset purchase to consider unconventional monetary policy as a
program announcement on March 23 (Figure 2.7, key additional part of their policy toolkit, especially
panel 3) and the improvement in global risk appetite where conventional policy space becomes limited.16
16For a deeper discussion of the use of unconventional monetary

15This
might also reflect that the rate cuts were already priced in policy in emerging market economies see Hofman and Kamber
or that risk premiums remained high. (forthcoming).

Figure 2.9. Considerations for Asset Purchase Programs
The depth of the domestic investor base and its ability to repatriate Credible monetary policy frameworks and sound governance are
foreign assets may affect the need for APPs. prerequisites for APPs.
1. Domestic Institutional Investor Base 2. Inflation: Volatility versus Deviation from Trend
(Assets, percent of GDP, latest data available) (Percent)
300 180 3.0
Insurers Pension funds Banks
Nonreserve external assets (right scale) 2.5
160
2.0
250 TUR
1.5
Inflation versus 10-year average

140 HUN
POL 1.0
200 120 IND 0.5
MEX
100 0.0
CHN ROM
150 CHL PHL –0.5
80 –1.0
PER
–1.5
100 60 ZAF THA
–2.0
COL
40 –2.5
50
RUS –3.0
20 MYS BRA
IDN –3.5
0 0 –4.0
0 1 2 3 4 5
South Africa
Malaysia
Chile
Thailand
Vietnam
Brazil
Ukraine
Poland
Colombia
Turkey
Saudi Arabia
India
Philippines
Russia
Hungary
Mexico
Romania
Indonesia
Nigeria
Volatility of inflation (last 10 years)
Sources: Bloomberg Finance L.P.; Haver Analytics; World Bank; and IMF staff calculations.
Note: In panel 1, data are as of latest vintage available, though gaps exist for select countries and series. Data labels in panel 2 use International Organization for
Standardization (ISO) country codes. APPs = asset purchase programs.
APPs may be suitable for countries constrained by their especially for open-ended programs (see Figure 2.9,
own effective lower bound, with inflation expectations panel 2, for select country characteristics to take into
steady, where the concern over capital outflows and FX consideration while deploying APPs, and Hofman and
depreciation is low or where the domestic absorption Kamber, forthcoming):
capacity of new bond supply is limited (Figure 2.9, •• Institutional and central bank credibility may be
panel 1). The goal of an APP in such cases is to exert weakened. Credible monetary policy frameworks
control over the medium- to long-end of the yield and sound governance are prerequisites for effective
curve (even when policy rates remain substantially unconventional policy actions such as APPs. Early
above zero) to lower government financing costs and to evidence suggests that APPs by central banks with
temporarily ease pressure on domestic investors when higher institutional quality tended to have a greater
there is increased issuance or foreign investor outflows. reduction of their bond local stress index, intro-
There are important caveats when it comes to this goal, duced earlier in this chapter. Increased balance sheet
however. Longer-term yields play a less central role exposure to long-term debt may raise concerns about
in most emerging market economies than they do in the central bank’s ability to raise interest rates when
advanced economies. Similarly, the fragilities behind conditions warrant or to achieve price stability.
higher short-term rates are likely to limit the scope for •• Asset purchases may invite concerns about fiscal
attempts to lower longer-term yields. dominance. When central banks become buyers of
Policymakers should consider both the benefits last resort, with large-scale and open-ended APPs
and potential significant costs of APPs with respect to in economies with weak monetary and fiscal policy
monetary policy and financial stability. If large-scale frameworks, it can lead to fiscal dominance, result-
APPs are used beyond the current pandemic-related ing in higher risk premiums and steeper government
extraordinary situation, the following risks could arise, bond yield curves.

•• APPs may intensify capital outflow pressure, already deemed to be at a high risk of debt distress
especially in countries with weaker fundamentals. (see the October 2019 GFSR) and with relatively
Expectations of large-scale APPs may put downward little policy space compared with major emerging
pressure on long-term yields and foreign exchange market economies. The postcrisis period of easy global
rates, putting capital flows at risk, especially during financial conditions allowed frontier market economies
risk-off periods, when emerging market assets are seen to raise unprecedented amounts of capital in private
as risky. Investors may decide to rebalance their port- markets (Figure 2.10, panel 1), all the while increasing
folios more decisively if APPs result in an excessive their rollover risk. Markets reflected these concerns, as
gap between domestic and peer-group risk premiums. bond spreads rose to their highest level since the global
•• The lasting presence of central banks as buyers financial crisis during the initial stages of the market
in the local currency bond market may distort sell-off, but spreads have since erased a significant
market dynamics. APPs can end up substantially amount of the widening (Figure 2.10, panel 2).
increasing the role of the central bank as a market To help alleviate the strains facing frontier econo-
maker, impairing the price discovery process, espe- mies, the Group of Twenty (G20) announced the Debt
cially in primary markets,17 and the development of Service Suspension Initiative (DSSI) to temporarily
the financial market. Considerations should also be ease the financing constraints of the poorest coun-
given to the effect of APPs on collateral availability tries by freeing up scarce money that they can use
in the banking system and its impact on the policy to mitigate the human and economic impact of the
rate transmission (Singh and Goel 2019) as well as COVID-19 crisis. While some countries have already
possible overvaluation of assets. begun to participate in the initiative, some have been
reluctant, in part because of fears of loss of market
The motivation, effectiveness, and associated risks of access (see also Chapter 1).
APPs vary considerably from country to country and Markets, however, are not pricing in a significant
depend on additional considerations, such as the struc- risk from DSSI participation, despite concerns about
ture and liquidity of capital markets, availability of possible negative actions by the credit rating agencies.
high-quality domestic assets, extent of foreign investor On average, spreads of countries eligible for the DSSI
participation, and level of development of the financial have outperformed those of other frontier countries,
sector (Hofman and Kamber, forthcoming). Focused even excluding countries eligible for the DSSI that do
use of APPs as part of the crisis toolkit of emerging not intend to participate (Figure 2.10, panel 3). This
and frontier market economy central banks with cred- outperformance could be a result of investor expec-
ible monetary policy frameworks and good governance tations that the initiative can allow these countries to
has a role to play. But continuing evaluation is needed better weather the outcome of the pandemic. For now,
as more data become available on the effectiveness of the initiative is providing relief primarily through a
unconventional monetary policy in emerging markets, moratorium on bilateral debt, whereas private sector
especially for open-ended programs. groups have begun assessing potential ways to assist.
Even though the DSSI helps free up scarce money
to mitigate the human and economic impact of
The Role of the Official Sector in Frontier COVID-19, once the impact of the pandemic becomes
Market Economy Debt Restructuring clearer, official sector relief may prove insufficient for
Frontier market economies18 entered the pandemic some countries. Overall, bilateral creditors represent
in a vulnerable position, with a number of countries about one-third of debt payments owed by coun-
tries eligible for the DSSI over the next few years
(Figure 2.10, panel 4). For many countries, private
17In markets that lack financial depth and where the government
sector debt represents a much larger proportion of
has large crisis-related short-term financing needs, there may be
scope for the central bank to provide, under conditions, temporary their external debt (Figure 2.10, panel 5).
support directly to the primary market to assist with the absorption For some countries, to achieve a necessary debt
of large issuance. reduction, it is impractical for only the official sector
18Frontier economies comprise 43 countries, defined in Online
Annex 2.1, the bulk of which are part of JP Morgan’s Next Genera- to proactively alleviate the debt burden. Putting off
tion Markets Index. debt relief by private sector creditors may eventually

Figure 2.10. Frontier Economies Have a Challenging Road Ahead
Frontier economies have become more dependent on private sector Market conditions have deteriorated substantially since the onset of the
debt in recent years. COVID-19 crisis.
1. Frontier Market Debt: Creditor Composition 2. Bond Spreads of Frontier Economies during the COVID-19 Crisis
(Percent of GDP and billions of US dollars) (Basis points)
25 Bilateral Multilateral Other private 250 2019–2020 precrisis average Crisis high Latest 2,000
Commercial loans Bonds 1,750
20 Private sector creditors (USD, right scale) 200
1,500
15 150 1,250
1,000
10 100 750
500
5 50
250
0 0 0
2006 07 08 09 10 11 12 13 14 15 16 17 18 Asia Africa Middle East Latin Europe
America
Countries eligible for the Debt Service Suspension Initiative have Bilateral creditors, primarily non–Paris Club creditors, represent about
outperformed somewhat since April. a third of debt payments over the next few years ...
3. Normalized Spreads of Frontier Market Economies 4. Debt Service Payments by Creditor for a Sample of Frontiers
(Index; January 1, 2020 = 100) (Share of total, percent)
Non-Paris Club Bilateral Paris Club Bilateral Multilateral/Other
DSSI non-DSSI frontier
Bonds Loans (nonofficial)
300 100
250 80
200 60
150 40
100 20
50 0
Jan. 2020 Mar. 20 May 20 July 20 Sep. 20 2020 2021 2022
... but for several countries, private creditor debt is significant.

5. Debt Outstanding: Private versus Official Creditors
(Public external debt outstanding, percent of GDP)
30
Bonds and loans (percent of GDP)
MNG
25
ZMB SEN
20 GHA
15 HND AGO
BEN CIV
10 NGA ETH COG
RWA TJK
KEN
5 PNG PAK
TZA UZB CMR
0
0 5 10 15 20 25 30
Total bilateral (percent of GDP)
Sources: Bloomberg Finance L.P.; Bond Radar; JPMorgan Chase and Co; World Bank; and IMF staff estimates.
Note: Panel 1 refers to public and publicly guaranteed debt. Panel 4 comprises a sample of 22 frontier economies that are DSSI-eligible. The broad frontier universe
comprises 43 countries defined in Online Annex 2.1. Panel 5 uses data from the World Bank as of 2018. Data labels in panel 5 use International Organization for
Standardization (ISO) country codes. DSSI = Debt Service Suspension Initiative; Latam = Latin America.

call for a larger debt write-down, which could dispro- the spike in global risk aversion and intervened to
portionately affect private sector debt. Markets appear smooth excess volatility of domestic currencies and
to perceive already that, in a default situation, they contain the spillovers of tighter global financial con-
would be forced to take a larger haircut than bilateral ditions to domestic financial conditions. Appropriate
creditors would. use of FX intervention, macroprudential policies,
Why this would drive higher spreads can be and capital flow management measures in the face
demonstrated in a hypothetical example. If a country of shocks, such as the global pandemic, can con-
requires a given overall debt reduction to make its tribute to financial stability and enhance monetary
debt sustainable, but one class of creditors is treated policy autonomy.
as senior, other creditors would need to take a greater This chapter finds that global factors played a more
burden (Figure 2.11, panel 1). Panel 2 of Figure 2.11 important role in driving currencies than FX interven-
demonstrates the impact that different levels of senior tion did, probably because of the global nature of the
debt would have on a bond’s spreads at given levels shock. The short-lived FX intervention is consistent
of expected probability of default.19 A country whose with using the currency as a key shock absorber when
debt is entirely “junior,” or private sector, would have a other vulnerabilities are in check. Countries with
much lower spread than one for which half of the debt shallow FX markets may experience macroeconomic
is considered senior. This spread impact increases as the destabilization after such shocks, and FX interventions
credit quality decreases (higher expected default prob- to lean against market illiquidity to mute excessive
ability). A model for sovereign bond spreads shows volatility can be appropriate (IMF 2020a).
that investors do expect a larger haircut than bilateral Most notably, many emerging and frontier market
creditors.20 The results of the model are consistent central banks for the first time embarked on APPs to
with investors expecting that bilateral creditors would ensure the smooth functioning of bond markets and
take a 30 percent haircut in the case of a country that provide accommodation in an environment of very low
requires an overall 40 percent haircut. This analysis policy rates. The apparent success in helping reduce bond
does not consider differences among groups of bilateral yields without risking financial stability so far prompts
creditors or whether the impact is less or more for the question of whether APPs should be part of the
Paris Club creditors. Considering that bilateral loans emerging and frontier market policy toolkit in the future.
are often extended at concessional levels, or at times For central banks with APPs in progress, transparency
when countries are not able to consistently borrow and clear communication21 are crucial to minimize
from private markets, it is not surprising that they risks to their credibility—especially in countries
would be expected to receive more favorable treatment with weaker institutional frameworks. In most cases,
in a restructuring scenario. APPs should be limited in time and scale and should
be linked to clear objectives. This chapter’s findings
suggest that APPs can be helpful, but that they are not
Policies for Recovery and Resilience a panacea to improve market conditions. They appear
Unprecedented policy measures put in place by to be more effective when used jointly as part of a
advanced and emerging market policymakers after broader macroeconomic policy package.
the onset of the COVID-19 pandemic averted the Central banks considering APPs for the first time or
worst outcome and helped stabilize domestic financial seeking to restart them should design programs that aim
conditions. Emerging market central banks actively to affect segments of the yield curve that are an effec-
used available and new tools to soften the blow from tive pricing benchmark to maximize transmission to
the real economy. Purchases should preferably be made
19This stylized exercise assumes a 10-year bond with an in secondary markets, as purchases in the primary
8 percent coupon. market or at below market rates can disrupt the price
20This is based on a variant of the emerging market hard currency
bond valuation model introduced in the October 2019 GFSR. The

domestic fundamentals include expectations for growth and inflation, 21Communication and transparency regarding the cost of ster-
current account balance, external debt, net issuance of foreign cur- ilization can also be crucial, especially in cases where central bank
rency government debt, and foreign currency reserves. External factors purchases are done below market rates. Large sterilization costs can
include global risk-appetite and growth expectations. The model was increase concerns about central bank losses and monetary policy
modified to also include the share of bilateral and multilateral debt. independence.

Figure 2.11. Large Shares of Senior Creditors Could Lead to Higher Spreads
If one class of creditors receives smaller haircuts, other creditors need Investors pricing a larger required haircut in case of default could
to bear a greater burden. meaningfully impact spreads.
1. Stylized Example of Issuer Requiring a Total 40 Percent 2. Bond Spread under Different Recovery Assumptions and
Haircut with Debt Evenly Split Expectations of Default
(Basis points)
3,000
50 percent senior share
0 percent senior share 2,500
50 percent
20 percent haircut
senior debt Aggregate 40%
haircut 2,000
Bond spread
1,500
1,000
50 percent
60 percent haircut 500
junior debt
0
30 40 50 60 70 80 90
Expected default probability (percent)

Note: Panel 2 assumes a bond with an 8 percent coupon and 10-year maturity. It assumes that an overall debt reduction of 40 percent is required, with senior debt
holders accepting only a 20 percent haircut.
discovery process and invite fiscal dominance. APPs maker in bond markets, impairing the price discovery
should take into consideration the efficacy of the port- process and financial market development. Specific
folio balance channel and whether investors have the measures for further local market development include
ability to allocate their investments in other domestic (1) developing efficient money market frameworks;
assets, such as corporate or covered bonds. In the (2) strengthening primary market practices to enhance
absence of such domestic alternatives, both foreign and transparency and predictability of issuance; (3) bolster-
domestic investors might choose to exit their country ing market liquidity, including the use of repo facilities
position altogether, which could increase the sensitivity for local dealers in times of stress; and (4) developing
of the exchange rate to APP policies. The resultant a robust market infrastructure, including local clear-
currency depreciation in countries with large currency ing and settlement and other services (as detailed in
mismatches in private sector balance sheets could at IMF 2020b). For countries with adequate preparation
least partly offset the stimulatory effect of APP policies in terms of legal barriers and market infrastructure,
by tightening overall financial conditions. The expe- authorities should work toward enabling settlement
rience of advanced economy central banks with exit and clearance of local currency debt in international
strategy plans may also be important for emerging capital markets so that domestic markets can benefit
market central banks to consider, particularly when the from access to wider liquidity pools.
size of the program is meaningful. Frontier market economies with unsustainable debt
APPs should be designed so as not to become bar- dynamics, limited market access, and high external
riers to the development of domestic capital markets financing requirements should preemptively and
or the growth of a stable and diversified local investor cooperatively seek debt resolution with their creditors.
base. In countries with relatively small bond markets, Countries that maintain market access at reasonable
large and prolonged APPs could end up substantially rates should decrease rollover risks as part of their debt
increasing the role of the central bank as a market management strategy.

References Hartley, Jonathan S., and Alessandro Rebucci. 2020. “An Event
Adler, Gustavo, Kyun Suk Chang, Rui Mano, and Yuting Shao. Study of COVID-19: Central Bank Quantitative Easing in
Forthcoming. “Foreign Exchange Intervention: A Data Set Advanced and Emerging Economies.” NBER Research Paper
of Public Data and Proxies.” International Monetary Fund, 27339, National Bureau of Economic Research, Cambridge, MA.
Washington, DC. Hofman, David, and Gunes Kamber. Forthcoming. “Unconven-
Adrian, Tobias, Richard K. Crump, and Emanuel Moench. tional Monetary Policy in Emerging Market and Developing
2013. “Pricing the Term Structure with Linear Regressions.” Economies.” Special Series on COVID-19, International
Federal Reserve Bank of New York Staff Report 340, revised Monetary Fund, Washington, DC.
April 2013, New York. International Monetary Fund (IMF). 2020a. “Toward an Inte-
Arslan, Yavuz, Mathias Drehmann, and Boris Hofmann. 2020. grated Policy Framework.” Washington, DC.
“Central Bank Bond Purchases in Emerging Market Econ- ———. 2020b. “Staff Note for the G20 IFAWG: Recent
omies.” BIS Bulletin 20, Bank for International Settle- Developments on Local Currency Bond Markets in Emerging
ments, Basel. Economies.” Washington, DC.
Beirne, John, Nuobu Renzhi, and Eric Alexander Sugandi. 2020. Jordà, Òscar. 2005. “Estimation and Inference of Impulse Responses
“Financial Market and Capital Flow Dynamics during the by Local Projections.” American Economic Review 95 (1): 161–82.
COVID-19 Pandemic.” Working Paper 1158, Asian Develop- Schrimpf, Andreas, and Vladyslav Sushko. 2019. “FX Trade
ment Bank Institute, Washington, DC. Execution: Complex and Highly Fragmented.” BIS Quarterly
Diebold, Francis X., and Kamil Yilmaz. 2012. “Better to Give Review (December).
than to Receive: Predictive Directional Measurement of Vola- Singh, Manmohan, and Rohit Goel. 2019. “Pledged Collateral
tility Spillovers.” Economic Journal 199: 157–71. Market’s Role in Transmission to Short-Term Market Rates.”
Drakopoulos, Dimitris, Rohit Goel, Evan Papageorgiou, and Can IMF Working Paper 19/106, International Monetary Fund,
Sever. Forthcoming. “Effects of Emerging Market COVID-19 Washington, DC.
Asset Purchase Program Announcements on Financial Mar- Teulings, Coen N., and Nikolay Zubanov. 2014. “Is Eco-
kets.” IMF Working Paper, International Monetary Fund, nomic Recovery a Myth? Robust Estimation of Impulse
Washington, DC. Responses.” Journal of Applied Econometrics 29 (3): 497–514.

3
CHAPTER
CORPORATE FUNDING
LIQUIDITY STRAINS CUSHIONED BY A POWERFUL SET OF POLICIES
•• In the Group of Seven (G7) economies, nonfinancial corporate borrowing surged in March and during
the second quarter of 2020, benefiting from unprecedented policy support as a consequence of the coro-
navirus disease (COVID-19) crisis.
•• Credit supply conditions across the G7 were generally favorable during the second quarter, yet the
buoyancy of the bond market in the United States stood in sharp contrast to tighter loan market lending
standards in that country.
•• Among listed firms, those vulnerable to liquidity shocks suffered relatively more financial stress in the
early stages of the COVID-19 crisis, and residual signs of strain remained as of the end of June.
•• Premature withdrawal of policy support could jeopardize the success achieved so far in broadly meeting
the nonfinancial corporate sector’s liquidity and funding needs.
The COVID-19 pandemic has adversely affected non- corporate sector, had a beneficial effect overall. Looking
financial corporate sector cash flows, generating liquidity ahead, premature withdrawal of policy support could
and solvency pressures. In the G7 economies—Canada, jeopardize the success achieved so far in broadly meet-
France, Germany, Italy, Japan, the United Kingdom, ing the nonfinancial corporate sector’s funding needs.
and the United States—corporate borrowing surged in
March and into the second quarter of 2020, thanks to
credit line drawdowns and unprecedented policy support. Introduction
This allowed firms to build cash buffers to cope with a The COVID-19 pandemic has triggered a deep global
period of reduced cash flow and high uncertainty. In the economic crisis. Closures and restrictions imposed by
United States, the bond market has been buoyant since governments to contain the spread of the virus, as well
the end of March, but credit supply conditions for bank as social distancing, have severely disrupted business
loans and the syndicated loan market have tightened. activity and clouded the economic outlook amid
In other G7 economies, credit supply conditions eased heightened uncertainty. Corporate cash flows have been
somewhat across markets during the second quarter. heavily impaired in many industries, with adverse impli-
Among listed firms, entities with weaker solvency or cations for corporate liquidity and solvency.
liquidity positions before the onset of COVID-19, as well In the major advanced economies, severe disruptions
as smaller firms, suffered relatively more financial stress in to corporate funding markets became apparent amid
some economies in the early stages of the crisis. However, a sharp tightening of financial conditions early in the
residual signs of strain remained as of the end of June, year following the onset of the COVID-19 crisis, as
when the stock market underperformance of French, UK, corporate bond funds, loan funds, and prime money
and US firms with pre–COVID-19 liquidity vulner- market funds faced large outflows. This led to a col-
abilities ranged between 4 and 10 percentage points. lapse in the issuance of nonfinancial corporate bonds,
Policy interventions, especially those directly targeting the syndicated loans, and commercial paper, and to a jump
in corporate spreads. Many firms turned to their exist-
The authors of this chapter are Andrea Deghi, Ken Zhi Gan, ing credit lines to secure funds in a “dash for cash.”
Tom Piontek, Dulani Seneviratne, Tomohiro Tsuruga, and Jérôme In response, policymakers in these economies
Vandenbussche (team leader), with contributions from Germán
Villegas Bauer, under the guidance of Fabio Natalucci and Mahvash quickly announced a wide range of powerful policy
Qureshi. Jeremy Stein served as an expert advisor. measures to support markets and address corporate

GLOBAL FINANCIAL STABILITY REPORT: BRIDGE TO RECOVERY
funding needs (see Online Box 3.1 for a brief descrip- The chapter seeks to address four broad sets of
tion of the key measures and their timing). Some of issues. First, it analyzes the impact of the COVID-19
these measures were unprecedented; one example is crisis on aggregate credit volumes in several segments
the new Federal Reserve facilities to support corporate of the corporate debt market as well as the effects of
credit. The combination of these fiscal, monetary, and the subsequent policy response on the debt financing
financial policy measures helped normalize financial choices of large firms. Second, it discusses the evo-
conditions during the second quarter, as discussed in lution of aggregate conditions in credit markets and
the June 2020 Global Financial Stability Report (GFSR) seeks to quantify the credit supply shocks in these
Update and Chapter 1 of this current report. However, markets. Third, it examines the extent to which ease of
corporate spreads remain wider than at the beginning access to external finance, or liquidity position, had an
of the year, especially in the high-yield segment, point- impact on firm-level financial performance in the early
ing to remaining concerns about default risk.1 stages of the crisis, potentially signaling the presence of
The degree of eventual economic scarring from the tighter credit conditions.5 Acknowledging that such an
COVID-19 crisis will depend a great deal on how well analysis is a very challenging task, the chapter turns to
the financial system—supported to an exceptionally an examination of the effect of key policy announce-
large extent by policies to date—is able to meet the ments and tries to gauge the impact of various types
corporate sector’s demand for liquidity during the of policy responses on the supply of corporate credit
crisis. This means preventing still-solvent firms facing during the containment phase of the pandemic.6
liquidity strains from turning into insolvent entities or The chapter finds that drawdowns of existing credit
being forced to significantly curtail their activities.2 lines and unprecedented policy support helped main-
Against this backdrop, this chapter assesses whether tain the flow of credit to firms, and that corporate
corporate liquidity needs were met for listed firms borrowing surged in March and the second quarter of
in the G7 economies during the first few months of 2020. As a result, firms managed to build cash buffers
the crisis (from the beginning of February to the end to cope with a period of reduced cash flow and high
of June).3 Given the rise in corporate sector leverage uncertainty. Since the end of March, the bond market
in several G7 economies during the period preced- has been buoyant in the United States, but credit sup-
ing COVID-19, as documented in recent issues of ply conditions for bank loans and syndicated loans have
the GFSR, the chapter also examines the impact of tightened. In Japan, bank lending standards have eased,
high corporate indebtedness on firms’ financial stress but bond market supply conditions have tightened
during the crisis. While the COVID-19 crisis has somewhat despite a solid year-on-year increase in issu-
severely hurt a very large number of unlisted small ance. In other G7 economies, credit supply conditions
and medium-sized enterprises, which traditionally have evolved in a more homogeneous manner across
face difficulties accessing external financing, lack of markets, with somewhat easier conditions prevailing, on
recent publicly available data for these firms prevents average, during the second quarter. Among listed firms,
a thorough analysis of their funding situation during entities with weaker solvency or liquidity positions
the pandemic.4 before COVID-19, as well as smaller firms, suffered rel-
atively more financial stress in some economies during
1Asof September 10, 2020, US investment-grade (high-yield) the early stages of the crisis, and residual signs of strain
credit spreads had widened 33 basis points (125 basis points) since
remained as of the end of June. Policy interventions,
the beginning of the year. In Europe, investment-grade (high-yield)
spreads had widened 9 basis points (101 basis points) on a net basis. especially those directly targeting the corporate sector,
Yet with US government bond yields having fallen significantly
during the crisis, junk bond yields were at, or close to, record lows.
2Several studies on the global financial crisis have documented reduc- 5The chapter does not aim to project liquidity gaps at the firm
tions in credit supply’s adverse consequences on employment, invest- level (see Banerjee and others 2020); rather, it aims to provide a
ment, and total factor productivity growth (Duchin, Ozbas, and Sensoy quantification of the challenges firms face in accessing debt financing
2010; Chodorow-Reich 2014; Duval, Hong, and Timmer 2020). during the containment phase of the COVID-19 crisis. Similarly, the
3The focus on G7 economies is dictated by these economies’ chapter does not aim to provide an account of differences in perfor-
global systemic relevance and their relatively better data availability. mance across industries but controls for the heterogeneous effect of
4Chapter 1 of the October 2020 World Economic Outlook discusses the crisis across industries in the empirical analysis.
a model-based analysis of the impact of the COVID-19 crisis on 6Data sources and variables used in this chapter are described
small and medium-sized enterprises, building on work by Gourin- in Online Annex 3.1. All annexes are available at www.imf.org/en/
chas and others (forthcoming). Publications/GFSR.

CHAPTER 3 Corporate Funding: Liquidity Strains Cushioned by a Powerful Set of Policies
had a beneficial effect, on average. These findings can concentrated in March, with a peak on the last day
help inform ongoing discussions about the appropriate of the month. Presumably, this reflects firms’ desire
level of policy support as the global economy moves to secure funds while they were still in compliance
toward the recovery phase. While trade-offs with other with their maintenance covenants and because they
policy objectives need to be considered, especially in a expected a sharp deterioration in cash flow during
context of limited fiscal space, premature withdrawal the second quarter. Gross drawdowns in the United
of policy support could jeopardize the success achieved States subsided at the beginning of April, resulting
so far in broadly meeting the nonfinancial corporate in a decline in utilization rates—that is, the share of
sector’s funding needs. credit line commitments used. The same reduction
can be observed in Canada; drawdown activity in
Japan, however, continued during the second quarter,
A Surge in Debt Financing and Cash Balances resulting in a utilization rate of 60 percent. Never-
This section discusses the provision of credit to firms theless, utilization rates across the seven economies
in key segments of the corporate credit market during remained well below 50 percent, on average, at the end
the containment phase of the crisis. Loans represent of June, suggesting that liquidity insurance remained
the major source of corporate debt funding in the significant, at least in the aggregate.10 Bank credit
G7 economies, ranging from 58 percent in the United developments during the second quarter also reflected
States to 90 percent in Germany, according to the latest the implementation of government programs (notably,
available financial accounts data. The remainder is com- off-budget credit guarantees) that transferred part—
posed of debt securities. In terms of issuance by large sometimes all—of the credit risk to the sovereign, as
firms, the ratio of syndicated loans (which are mostly well as government-sponsored loans with a significant
held by banks post syndication if they are investment grant component. These direct support programs
grade and by nonbanks if they are non-investment to corporate funding represented between 2.6 and
grade) to bonds ranges from two to three.7 34 percent of GDP as of June 12 (Figure 3.1, panel 4).
Despite a period of acute financial stress early in the They complemented other on-budget fiscal measures
year, outstanding amounts of bank credit to firms grew that directly supported corporate cash flows and
significantly in March and in the second quarter in all solvency, for example, through grants, employment
seven economies analyzed (Figure 3.1, panel 1). On support programs, and reductions in tax liabilities.11 As
a year-over-year basis, the rate of bank credit growth of early July, committed amounts appear to have been
during the first half of the year was clearly above significantly smaller than announced amounts in Euro-
trend.8 Part of this dynamic is clearly attributable to pean economies (Anderson, Papadia, and Véron 2020).
sizable credit line drawdowns, especially in the United Syndicated loan issuance during the first half of the
States (Figure 3.1, panel 2). Listed firms’ drawdowns year was somewhat more heterogeneous across econo-
increased more than 40 percent, on average, compared mies. It was generally stronger than in 2019 in Europe
with the first half of 2019. The increase was partic- and Japan, but weaker in the United States and Canada,
ularly spectacular in the United States, where net draw- especially during the second quarter. This appears to
downs at the end of March doubled, representing an have been driven by a surge in investment-grade loan
increase of $250 billion, which is of the same order of issuance in Europe and Japan (Figure 3.2, panel 1) and
magnitude as the increase in commercial and indus- a drop in leveraged loan issuance outside of Germany
trial loans by domestic banks over the same period.9 and Italy (Figure 3.2, panel 2).12 The weak recovery in
Panel 3 of Figure 3.1 shows that these drawdowns were the leveraged loan markets was to a large extent due
7Syndicated loans include both term loans and credit lines. 10Of course, there is substantial heterogeneity across firms and sec-
8Before the pandemic, the volume of nonfinancial corporate bank tors. In the United States, the utilization rate was significantly above
loans was on a declining trend in Italy. average in wholesale and retail trade at the end of June.
9Acharya and Steffen (2020) and Kapan and Minoiu (2020) 11See the IMF’s Fiscal Monitor Database of Country Fiscal
discuss credit line drawdowns in the United States in early 2020. Measures in Response to the COVID-19 Pandemic, https://www.imf
In contrast to the experience of the global financial crisis described .org/en/Topics/imf-and-covid19/Fiscal-Policies-Database-in-Response
in Ivashina and Scharfstein (2010), the increase in credit line -to-COVID-19.
drawdowns was related to immediate liquidity demand rather than 12It should be noted that the euro area leveraged loan market is
concerns about the health of the US banking sector. significantly smaller than the US market.

Figure 3.1. Bank Lending to Nonfinancial Firms and Government Liquidity Support
Corporate bank lending grew rapidly from March onward ... ... driven in part by credit line drawdowns ...
1. Bank Loans to Nonfinancial Firms, Amount Outstanding 2. Listed Nonfinancial Firms’ Net Credit Line Drawdowns
(NSA; corresponding period in 2019 = 100) (Percent change from same period in 2019)
140 120
Jan. 2020 Feb. 20 Mar. 20 2020:Q1 2020:Q2 2020:H1
Apr. 20 May 20 June 20
130 100
80
120
60
110
40
100 20
90 0
Canada France Germany Italy Japan United United Canada France Germany Italy Japan United United
Kingdom States Kingdom States
... especially in the United States in March. Liquidity support to firms by government was huge, especially in
Europe and Japan.
3. United States: Gross Credit Line Drawdowns 4. Governments’ Announced Liquidity Support Measures in Response
(Billions of US dollars; March 5–June 30, 2020) to COVID-19
(Percent of GDP)
80 40
70 35
60 30
50 25
40 20
30 15
20 10
10 5
0 0
5 Mar. 19 Mar. 2 Apr. 16 Apr. 30 Apr. 14 May 28 May 11 June 25 June Canada France Germany Italy Japan United United
Kingdom States
Sources: Federal Reserve; Haver Analytics; IMF, Fiscal Monitor Database of Country Fiscal Measures in Response to the COVID-19 Pandemic (June 2020); S&P
Capital IQ; S&P Leveraged Commentary & Data; and IMF staff calculations.
Note: Panel 2 is based on data available as of August 25, 2020. Half-yearly data are used instead of quarterly data for European economies because of scant
quarterly reporting (when first half data are not available, but first quarter data are, the latter are used). Panel 4 shows liquidity support (including equity injections,
loans, asset purchases or debt assumption, guarantees, and quasi-fiscal operations) per country as a percent of GDP. Amounts do not include above-the-line fiscal
measures, such as the US Paycheck Protection Program, which amounts to about 3 percent of US GDP. NSA = not seasonally adjusted.
to subdued demand from the traditional investor base. of downgrades and defaults (Figure 3.2, panel 4), which
Collateralized loan obligation (CLO) new issuance has may affect lower-rated tranches.
been slow to restart.13 While activity picked up mod- Corporate bond markets in the first quarter were
estly from March levels, new CLO supply ran at half generally more resilient despite coming under intense
of last year’s pace, while still accounting for more than pressure in mid-March. Policy responses by central
70 percent of new leveraged loan demand (Figure 3.2, banks announced in the second half of March, espe-
panel 3). CLO investors were concerned about the wave cially facilities aimed at directly supporting corporate
bond markets, appear to have boosted activity in these
markets and contributed to a reversal in corporate
13A collateralized loan obligation is a structured finance product
bond fund flows (including exchange-traded funds).
collateralized predominantly by broadly syndicated leveraged loans.
See Chapter 2 of the April 2020 GFSR for a discussion of risky During the second quarter, investment-grade issu-
corporate credit markets. ance surged to levels twice as large as those in 2019

Figure 3.2. Developments in Syndicated Loan Markets
During the second quarter, investment-grade loan issuance was much ... whereas activity in the leveraged loan market generally dropped
stronger in Europe and Japan than in North America ... sharply.
1. Investment-Grade Syndicated Loan Issuance, First Half of 2020 2. Leveraged Loan Issuance, First Half of 2020
(Corresponding period in 2019 = 100) (Corresponding period in 2019 = 100)
300 300
Jan.–Feb. March Q2 Jan.–Feb.
March
Q2
200 200
100 100
0 0
Weaker investor demand suppressed new leveraged loan issuance, ... as underlying asset quality deteriorated.
such as from slower CLO formation ...
3. US and EU CLO Issuance and Share of New Issue Leveraged Loans 4. CLO Credit Quality Composition
(Left scale = billions of US dollars; right scale = percent) (S&P CLO index, percent)
180 75 16 40
EU CLO issuance US CLO issuance CCC D
160 CLO share of new issue loans (right scale) 70 14 Negative 35
140 outlook
65 12 30
(right scale)
120
10 25
100 60
8 20
80 55
6 15
60
50 4 10
40
20 45 2 5
0 40 0 0
2002 04 06 08 10 12 14 16 18 20 Jan. 2020 Mar. 20 Apr. 20 May 20 June 20
Sources: Dealogic; S&P Capital IQ; S&P Global Ratings; S&P Leveraged Commentary & Data; and IMF staff calculations.
Note: For panel 3, 2020 data are annualized through end-June 2020. Data for individual European countries are not available, so the European Union aggregate is
shown. CLO = collateralized loan obligation; EU = European Union.
in France, Germany, the United Kingdom, and the (Figure 3.3, panel 3). By use of proceeds, more than
United States (Figure 3.3, panel 1). The response of 80 percent of year-to-date supply was for refinanc-
the high-yield segment was somewhat more muted ing existing debt as lower yields and strong investor
outside the United States, probably reflecting its rela- demand encouraged a range of issuers to tap into
tive underdevelopment and the focus of central banks’ the market to repay credit lines, or for short-term
purchases on the investment-grade segment. For its expenses such as working capital (Figure 3.3, panel 4).
part, the United States saw high-yield issuance during Issuances motivated by acquisition and dividends
the second quarter more than double compared with or share repurchases, however, were at their lowest
that in 2019 (Figure 3.3, panel 2). in a decade.
The characteristics of new debt in the high-yield Developments in bond and syndicated loan
bond market reveal a shift toward higher quality. issuance suggest that, for firms with access to these
In G7 economies, nearly 60 percent of high-yield markets, the bond market clearly was the preferred
new issues during the first half of the year were BB source of debt financing in the United States, but
rated, and more than 30 percent of the bonds were perhaps not in the other G7 economies. This hypoth-
secured, the highest levels for the past 15 years at least esis is confirmed by a granular investigation of the

Figure 3.3. Corporate Bond and Commercial Paper Issuance
Unlike for syndicated loans, bond issuance was buoyant during the ... as well as in the high-yield segment in the United States.
second quarter in the investment-grade segment ...
1. Investment Grade Bond Issuance, First Half of 2020 2. High-Yield Bond Issuance, First Half of 2020
(Corresponding period in 2019 = 100) (Corresponding period in 2019 = 100)
300 300
Jan.–Feb. March Q2 Jan.–Feb. March Q2
200 200
100 100
0 0
Canada France Germany Italy Japan United United G7 excluding United States United States
Kingdom States
High-yield bond supply shifted to higher quality with more security and The majority of high-yield bond supply was used for refinancing and for
stronger ratings. other purposes, such as repayment of credit lines.
3. Group of Seven High-Yield Bond Supply by Security and Rating 4. Group of Seven High-Yield Bond Issuance by Use of Proceeds
(Percent) (Percent)
BB rated Secured Other/corp purpose/working capital Refinancing M&A/LBO/recaps
70 100
60
80
50
40 60
30 40
20
20
10
0 0
2006 08 10 12 14 16 18 20 2006 08 10 12 14 16 18 20
The bond market was clearly more attractive to US firms during the ... both in the investment-grade and the high-yield segments.
second quarter ...
5. Change in Relative Attractiveness of Bond versus Loan Issuance 6. United States: Change in Relative Attractiveness of Bond versus
during the First Half of 2020 Loan Issuance
(Change in bond issuance probability, percentage points) (Change in bond issuance probability, percentage points)
40 Canada Euro area 40 United States 40 35 Investment grade High yield 35
30 30 30 30 30
20 20 20 25 25
10 10 10
20 20
0 0 0
15 15
–10 –10 –10
–20 –20 –20 10 10
–30 –30 –30 5 5

–40 –40 –40 0 0
2020:Q1 20:Q2 2020:Q1 20:Q2 2020:Q1 20:Q2 2020:Q1 20:Q2 2020:Q1 20:Q2

Figure 3.3. Corporate Bond and Commercial Paper Issuance (continued)
Volumes in the commercial paper market had opposite dynamics in the Nonfinancial corporate debt growth was strong overall.
United States and the euro area.
7. Total Value of Nonfinancial Commercial Paper Issuance 8. Total Debt Growth of Listed Firms
(Year-over-year growth rate, percent)
500 US commercial paper issuance (left scale) 20 Contribution from credit lines Contribution from other debt 15
Euro area short-term European paper issuance
400 (right scale)
15 10
Billions of US dollars
Billions of euros
300
10 5
200
5 0
100
2020:Q1
20:Q2
France 2020:H1
Germany 2020:H1
20:H1
2020:Q1
20:Q2
2020:H1
2020:Q1
20:Q2
0 0
Jan. 2015
May 15
Sep. 15
Jan. 16
May 16
Sep. 16
Jan. 17
May 17
Sep. 17
Jan. 18
May 18
Sep. 18
Jan. 19
May 19
Sep. 19
Jan. 20
May 20
Canada
Italy
Japan
United
Kingdom
United
States
Sources: Federal Reserve; Haver Analytics; S&P Capital IQ; S&P Leveraged Commentary & Data; and IMF staff calculations.
Note: For panels 3 and 4, 2020 data are through end-June. Euro area refers to three euro area economies (France, Germany, Italy). Panels 5 and 6 show the change
in the probability of issuing a bond (versus a loan) for a nonfinancial firm with characteristics equal to the sample mean during the first and second quarters of 2020
compared with before the COVID-19 crisis. Colored bars indicate significance at the 1 percent level. Empty bars indicate lack of statistical significance. See Online
Annex 3.2 for methodological details. Panel 8 is based on data available as of August 25, 2020. Data as of the first half of the year are used for European Group of
Seven economies to account for semiannual reporting of most firms (when first half data are not available, but first quarter data are, the latter are used).
LBO = leveraged buyout; M&A = mergers and acquisitions.
debt financing choice of these firms. Controlling for the United States confirms that the shift toward the
a large set of firm characteristics and macro-financial bond market happened in both the investment-grade
variables, the analysis documents a shift toward bond and high-yield segments, with the shift in the for-
financing in the United States but not in other juris- mer already visible in the first quarter, in line with
dictions (Figure 3.3, panel 5).14 This finding suggests record investment-grade issuance levels in March
that the Federal Reserve’s March 23 announcement (Figure 3.3, panel 6).18 These shifts in corporate
of its new corporate credit facilities had a stimula- financing choice during the first half of the year also
tive impact on domestic bond markets.15,16 That the varied, depending on firm characteristics such as
choice between bond versus loan financing was not leverage and investment opportunities, as discussed in
affected in other jurisdictions likely partially reflects Online Box 3.2.
the presence of central bank corporate bond purchase In contrast to the bond market, volumes in the
programs predating the pandemic in these economies commercial paper market in the United States have
(except in Canada).17 A more detailed analysis for not recovered since their sharp drop in March, when
investors shifted funds from prime to government
14See
Online Annex 3.2 for methodological details. money market funds (Figure 3.3, panel 7), despite
15Thus,
a key driver of the shift toward bond financing in the
the reintroduction of the Federal Reserve’s Commer-
United States appears to be related to policy rather than to the
weakness of banks’ balance sheets, as was the case at the time of cial Paper Funding Facility on March 17 and inflows
the global financial crisis (Adrian, Colla, and Shin 2013; Becker resuming into prime funds, especially from institu-
and Ivashina 2014). The Federal Reserve corporate credit facilities
cover the primary bond and loan markets as well as the secondary
bond market. As of August 31, no purchases had been made on the 18One factor contributing to the large volume of high-yield
primary markets. bond issuance in the United States in the second quarter was the
16The evidence for the US market is consistent with the findings announcement on April 9, 2020, by the Federal Reserve that the
of Acharya and Steffen (2020). scope of its new corporate credit facilities would be extended to
17The Bank of Canada announced its first corporate bond pur- high-yield exchange-traded funds and bonds and loans from firms
chase program on April 15, 2020. that lost their investment-grade status after March 22, 2020.

tional investors. It appears that the fall in bond market a further large expansion during the second quar-
yields has tempted firms to reduce their refinancing ter, especially in France and the United Kingdom
risk and substitute commercial paper with longer-term (Figure 3.4, panel 2).
debt.19 By contrast, commercial paper issuance in the
euro area, supported by the European Central Bank’s
expansion of its commercial paper purchases through Shifts in Aggregate Credit Supply Conditions
the Asset Purchase Programme and the Pandemic The large increase in borrowing (net of withdrawals
Emergency Purchase Programme, rebounded quickly from existing credit lines) in March and the second
from the March trough and hit a record high in quarter of 2020 was associated with credit spreads
June. Incentives to substitute commercial paper with that widened sharply in March and subsequently
longer-term bonds were weaker in the euro area, slowly declined (as discussed in the June 2020 GFSR
because the yield differential remained more stable Update and Chapter 1 of this report). A key reason
than in the United States.20 for the wider spreads is obviously the sharp dete-
All in all, the year-over-year growth rate of total rioration in corporate fundamentals and concerns
debt of listed firms was strong, generally exceeding about default risk in all seven economies (Figure 3.5,
10 percent, with notable contributions from credit line panel 1), but a tightening in credit supply may also
drawdowns in Canada and the United States during have played a role.
the first quarter (Figure 3.3, panel 8). To assess how much of the widening in spreads
Evidence suggests that this additional borrowing can be attributed to adverse credit supply con-
was used mostly to build cash reserves to cope with ditions, this section looks at evidence available
the uncertainty and the expected reduction in cash in different segments of credit markets. For the
flow triggered by the pandemic shock. In contrast commercial bank loan market, useful information
to Europe, all listed firms in Canada, Japan, and is obtained from central banks’ quarterly surveys
the United States are required to report quarterly, of bank lending officers, which measure officers’
and their cash flow statements for the first quar- perception of the strength of credit demand and of
ter reveal an accumulation of cash and short-term the evolution of their banks’ lending standards.21
investments of about 0.5 percent of assets in Japan For the European and US primary syndicated loan
and about 1.5 percent of assets in Canada and the markets, an empirical analysis to disentangle credit
United States. This behavior contrasts sharply with supply from demand factors is conducted by making
that observed a year earlier and during the peak use of publicly available transaction-level issuance
of the global financial crisis in the fourth quarter data. Specifically, the analysis relies on empirical
of 2008, when no cash accumulation took place estimation of a supply and demand system of equa-
(Figure 3.4, panel 1). The change in cash levels can tions that includes variables capturing lender and
be attributed mostly to an increase in financing in borrower characteristics and covers the mid- to late
Canada, a reduction in investment in Japan, and a 2000s through the second quarter of 2020.22 The
combination of both in the United States relative value of the credit supply shock in each quarter is
to 2019. During the second quarter, listed Japanese obtained by computing the time-varying “residual
and US firms built their cash buffers further, term” of the credit supply equation. For the second-
whereas listed Canadian firms reduced them some- ary corporate bond market, a measure of investor
what. The accumulation of cash is also visible from risk appetite—the so-called excess bond premium
nonfinancial corporate deposit data, which reveal
19Li and others (2020) suggest that liquidity rules introduced
at the time of the 2016 money market fund reform may not have 21An important caveat in interpreting results of bank lending
achieved the goal of making the system immune to runs. See also the officers’ surveys is that they do not always clearly distinguish between
discussion in Eren, Schrimpf, and Sushko (2020). changes in default risk and changes in credit supply in the definition
20The Bank of Canada and the Bank of England also introduced of lending standards.
commercial paper purchase programs, whereas the Bank of Japan 22The analysis addresses endogeneity concerns by using
stepped up its existing program. These countries are not shown on an identification-through-heteroscedasticity methodology
the chart for lack of data. (Rigobon 2003). See Online Annex 3.3 for details.

Figure 3.4. Change in Corporate Cash-to-Assets Ratio and Corporate Bank Deposits
Nonfinancial firms accumulated more cash during the first quarter of ... and this precautionary behavior continued during the second
2020 than during the same period of 2019, mostly because of quarter.
increased external financing in Canada and the United States ...
1. Change in Cash Holdings and Cash Flow Components 2. Nonfinancial Firm Deposits, Amount Outstanding
(Percent of lagged assets) (NSA; corresponding quarter in 2019 = 100)
4 From operations From investment 2020:Q1 2020:Q2 150
From financing Net change in cash
3
140
2
130
1
0 120
–1
110
–2
100
–3
–4 90
Canada France Germany Italy Japan United United
GFC
(2008:Q4)
Pre–COVID-19
(2019:Q1)
COVID-19
(2020:Q1)
GFC
(2008:Q4)
Pre–COVID-19
(2019:Q1)
COVID-19
(2020:Q1)
GFC
(2008:Q4)
Pre–COVID-19
(2019:Q1)
COVID-19
(2020:Q1)
Kingdom States
Canada Japan United States
Sources: Bank of Japan; Federal Reserve Board; Haver Analytics; S&P Capital IQ; and IMF staff calculations.
Note: Panel 1 shows the listed nonfinancial firms’ quarterly net change in cash as well as the contributions from the three cash flow components. European countries
are not shown because of insufficient data for the first quarter. Panel 2 shows the amount of nonfinancial firms’ deposits outstanding in the first and second quarters
of 2020 compared with the corresponding quarter of 2019. Data for the second quarter are not available for Japan and the United States. GFC = global financial
crisis; NSA = not seasonally adjusted.
proposed by Gilchrist and Zakrajšek (2012)—is other G7 economies.25 In particular, a large loosening
constructed to gauge shifts in supply.23,24 of credit conditions was observed in Japan and the
Survey-based evidence indicates that the commercial United Kingdom (Figure 3.5, panel 2).26 This stands
bank loan market in the United States was an outlier in sharp contrast to the experience during the global
across countries in the second quarter. Credit demand financial crisis, when surveys indicate that banks tight-
fell and lending standards tightened sharply, while the ened lending standards consistently across the board.
evolution was generally muted or the opposite in the The situation in the current crisis is likely related to
the fact that banks’ indicators of funding stress spiked
only briefly in late March before normalizing thanks
23This measure is constructed in two steps using detailed infor-
mation on many individual corporate bonds for the period from the
mid-2000s (or the first quarter of 2011 for the euro area) through 25The evolution of the index for the United States indicates only
the second quarter of 2020. First, for each bond, a spread to a that the tightening of lending standards was widespread, not that it
synthetic risk-free rate that considers information on the duration was intense. However, the text describing the survey results makes
of the bond is computed. Such a spread is more accurate than the it clear that lending standards were tight and explains that “banks,
more commonly used “naïve” spreads, whose construction ignores on balance, reported that their lending standards across all loan
bond duration. Second, the spread is purged of its credit risk compo- categories are currently at the tighter end of the range of standards
nent to obtain the excess bond premium, which can therefore be between 2005 and the present” (Board of Governors of the Federal
interpreted as an indicator of bond investor risk appetite. See Online Reserve System 2020).
Annex 3.4 for methodological details. The series for the United 26In the United Kingdom, the survey question refers to the “avail-
States is from the Federal Reserve Board. ability of credit” rather than to lending standards per se. The two
24The three euro area economies (France, Germany, Italy) are ana- notions are different in the presence of government loan guarantees,
lyzed as a group to improve sample size, and Canada is not included which may explain part of the difference between the United King-
in the analysis for data availability reasons. dom and the euro area economies.

Figure 3.5. Evolution of Credit Supply Conditions
As the risk of default increased ... ... bank lending standards tightened in the United States but eased in
Japan and the United Kingdom.
1. One-Year Expected Default Frequency of Nonfinancial Firms Rated 2. Change in Bank Lending Standards
between Baa1 and B3 at the End of 2019, End of Period, 75th Percentile (Index; see note for details)
(Difference from end-2019, percent)
4 80
Jan. 2020 Feb. 20 Mar. 20 2020:Q1 2020:Q2 2017:Q1–2019:Q4
Apr. 20 May 20 June 20 60
3 Tightening
40
2 20
0
1 –20
–40
0
–60
–1 –80
In the United States, credit conditions tightened somewhat in the In the United Kingdom, credit conditions also eased in the bond market
syndicated loan market, but eased in the bond market after a period of after the stress in March, while conditions in the syndicated loan
tension in March. market remained neutral.
3. Credit Supply Conditions in the United States 4. Credit Supply Conditions in the United Kingdom
(Top: syndicated loan market, spread residual, percent—quarterly; (Top: syndicated loan market, spread residual, percent—quarterly;
bottom: bond market, excess bond premium, percent—monthly) bottom: bond market, excess bond premium, percent—monthly)
0.5 Tightening Tightening 1.0
0.0 0.0
–0.5 Easing Easing –1.0
–1.0 –2.0
2007 08 09 10 11 12 13 14 15 16 17 18 19 20:Q1 20:Q2 2010 11 12 13 14 15 16 17 18 19 20:Q1 20:Q2
0.8 1.2
0.8
0.4
0.4
0.0 0
–0.4
–0.4
–0.8
–0.8 –1.2
2007 08 09 10 11 12 13 14 15 16 17 18 19 2007 09 11 13 15 17 19
20:M1
M2
M3
M4
M5
M6
20:M1
M2
M3
M4
M5
M6
to the speed of policy support to financial markets and syndicated loan market. Credit conditions were neutral
the economy, as well as to the effect of government in the first quarter, on average, and tightened during
programs to support lending to businesses (Bank of the second quarter, bringing the market into a tight
England 2020; European Central Bank 2020).27 position, though not as tight as in the aftermath of the
Turning to supply conditions in the syndicated global financial crisis. By contrast, the bottom part of
loan and bond markets, the divergence across the the same panel, which shows supply conditions in the
two markets during the second quarter in the United secondary bond market, reveals that a large part of the
States is striking. The top part of Figure 3.5, panel 3, March tightening was undone during the second quar-
shows the time series of the credit supply shock in the ter. Aside from the stimulative effect of the introduc-
tion of the Federal Reserve corporate credit facilities
27The total amount of credit line drawdowns could also be a factor
mentioned previously, two supply-side considerations
explaining the tightening of lending standards in the United States
because it reduced the amount of bank capital available for new may explain the buoyancy of the US bond market.
lending (Kapan and Minoiu 2020). First, with short-term rates near zero and Treasury

Figure 3.5. Evolution of Credit Supply Conditions (continued)
In the euro area, credit conditions eased in the syndicated loan market In Japan, conditions in the bond market tightened in March and
and remained broadly neutral in the bond market. remained slightly on the tight side in the second quarter.
5. Credit Supply Conditions in the Euro Area 6. Credit Supply Conditions in Japan
(Top: syndicated loan market, spread residual, percent—quarterly; (Bond market, excess bond premium, percent—monthly)
bottom: bond market, excess bond premium, percent—monthly)
3.0 2
Tightening Tightening
2.0
1.0
1
0.0
–1.0 Easing
–2.0 0
2010 11 12 13 14 15 16 17 18 19 20:Q1 20:Q2
1.2 –1
0.6
Easing
0 –2
–0.6
–1.2 –3
2011 13 15 17 19 2007 08 09 10 11 12 13 14 15 16 17 18 19
20:M1
M2
M3
M4
M5
M6
20:M1
M2
M3
M4
M5
M6
Sources: Bank of Japan; Bloomberg Finance L.P.; Dealogic; Federal Reserve Board; Haver Analytics; Moody’s Analytics; Refinitiv Datastream, Eikon; S&P Market
Intelligence; and IMF staff calculations.
Note: Panel 1 shows the change in the 75th percentile of the one-year end-of-period expected default frequency of nonfinancial firms rated between Baa1 and B3
(lower medium grade to highly speculative grade) at the end of 2019 in each Group of Seven country between the end of 2019 and each of the first six months of
2020. Panel 2 shows the quarter-on-quarter change in bank lending standards from the bank lending survey conducted by respective central bank; change is shown
in the form of an index ranging from –100 to 100. Canada, euro area economies, and the United Kingdom report a balance of opinions weighted by asset size with a
base value of 0; Japan reports a balance of opinion weighted by the level of easing or tightening; the United States reports an unweighted balance of opinion in two
categories by firm size (large versus small); and the figure shows the simple average of the two. See Online Annexes 3.3 and 3.4 for methodological details on the
construction of the series shown in panels 3–6. Credit conditions in Canada and in the Japanese syndicated loan market could not be computed because of
insufficient data. M = month.
purchases by the Federal Reserve bringing down term A yield curve that shifted toward zero, as in the United
premiums, investors’ search for yield pushed them States, may also have contributed to making the
toward yield-providing assets, especially those within corporate bond market attractive to investors. In the
the perimeter of central bank support. Second, expec- euro area, where key policy rates remained unchanged
tations of no rise in the policy rate for several years around zero, bond market conditions continued to be
reduced investors’ incentives to get exposure to floating broadly neutral, on average, during the first half of the
rates. As syndicated loan rates are floating and bond year, but a clear loosening of conditions took place in
rates are fixed, some investors may find bonds relatively the loan market during the second quarter (Figure 3.5,
more attractive in the current environment. A separate panel 5). In Japan, the March bond market tighten-
analysis for investment-grade syndicated loans and lev- ing persisted through the end of June, but overall risk
eraged loans indicates that conditions moved from easy aversion was within the normal range observed over
to tight during the second quarter in both segments.28 the past decade (Figure 3.5, panel 6).
The dynamics of credit conditions in the United All in all, the recent evolution of the excess bond
Kingdom’s bond market mirrored those in the United premium suggests that conditions in bond markets
States, but no tightening was observed in the syndi- were generally favorable during the second quarter,
cated loan market, on average (Figure 3.5, panel 4). especially in the United Kingdom and the United
States. In the United States, however, bank lending
28Loan covenant quality in North America appears to have
standards were tight, and the bank loan market was a
continued to weaken during the first quarter, reaching its all-time clear outlier compared with the other G7 economies,
worst level (according to Moody’s)—to the benefit of borrowers
who would need that flexibility during the crisis (Moody’s Investors where the change in lending standards ranged from
Service 2020). a small tightening to a large easing. These differences

across economies and markets likely reflect the analysis examines the effect of these three vulnera-
relative strengths of the different policy responses bilities over and above the effect of leverage-related
targeting the two markets, in particular the scope of vulnerabilities, which clearly amplified the effect of the
government-sponsored loan guarantee programs as negative cash flow shock related to COVID-19 in five
well as investors’ search for yield in an environment of of the seven economies (Figure 3.6, panels 1 and 2).32
ultra-low interest rates and shifting expectations about Evidence of relatively greater financial stress mea-
future policy rates.29 sured by cumulative abnormal returns—that is, the
cumulative difference between the actual returns and
the returns predicted by a simple one-factor asset
Greater Financial Stress Initially for pricing model—is pervasive for relatively smaller firms.
Some Vulnerable Firms Their underperformance during February–March in
Beyond aggregate indicators, changes in credit Germany, Japan, the United Kingdom, and the United
conditions are also likely to be visible through their States was close to, or greater than, 10 percentage
differential impact on firms with different charac- points (Figure 3.6, panel 2). Furthermore, firms that
teristics, as some firms may be more vulnerable to entered the COVID-19 crisis with relatively high
aggregate funding liquidity shocks than others. First, liquidity vulnerabilities also experienced relatively
firms that generally have more restricted access to greater financial stress than those with higher liquidity
credit markets—for example, because of their relatively buffers in some economies during late February and
smaller size—may be more exposed to a deterioration March. Panel 3 of Figure 3.6 shows the cumulative
in risk appetite than the rest of the corporate sector.30 abnormal returns of two groups of US firms: those
Second, firms with a worse liquidity position because with low and high relative cash. While the stock
of a lower stock of cash or higher short-term debt market performance of the two groups is indistinguish-
that needs to be rolled over are more sensitive to a able until late February, a wedge in favor of the latter
tightening of credit conditions. In addition, firms with group appears at that time and becomes wider during
higher leverage may also suffer more during episodes of the second half of March. A more formal econometric
financial stress. investigation, which controls for a number of firm
A comparison between the stock market perfor- characteristics (including the industrial sector) at
mance of firms most vulnerable to funding shocks and the end of 2019, as well as the expected size of the
that of other, less vulnerable firms can therefore be pandemic-related revenue shock, confirms that visual
a useful complement to the aggregate analysis pre- impression: firms with relatively less cash suffered
sented earlier in the chapter to better understand the more financial stress in the United Kingdom and
behavior of lenders with respect to credit to firms. In the United States, and those with a relatively higher
what follows, the analysis focuses on vulnerabilities to level of short-term debt (net of cash) suffered more in
funding liquidity shocks measured at the end of 2019 France, the United Kingdom, and the United States
along three dimensions: (1) small size (low total assets), (Figure 3.6, panel 4).33 In these five cases, the under-
(2) low cash and short-term financial investments rel- performance of firms with liquidity vulnerabilities
ative to industry peers (as a share of total assets), and between early February and end-March was about
(3) high short-term debt net of cash and short-term 5 percentage points.
financial investments (as a share of total assets).31 The
29It
is plausible that, in each country, the structure of the financial
Policies that Helped Relieve Funding Stress
sector (for example, market-based versus bank-based) played a role Precise measurement of the effects of policy
in the choice of policy instruments and calibration of the policy announcements and actions in the context of the
response across different markets, which in turn may explain the
relative dynamics of supply conditions in the various markets. COVID-19 crisis is an extremely challenging task.
30See Holmstrom and Tirole (1997) for a theoretical discussion.
Duchin, Ozbas, and Sensoy (2010) and Hadlock and Pierce (2010)
discuss various financial constraint indicators commonly used in the 32See Online Annex 3.5 for methodological details. For size, rel-
empirical corporate finance literature. ative cash, and liquidity gap (leverage), a firm is deemed vulnerable
31A high level of short-term debt net of cash exposes a firm to if it belongs to the weakest tercile (half ) of the distribution of the
rollover risk. A low level of cash reduces a firm’s room to maneu- characteristic at the end of 2019.
ver in case credit conditions tighten (see, for example, Joseph and 33The finding for the United Kingdom echoes that of Joseph and
others 2020). others (2020).

Figure 3.6. Firm-Level Stock Market Performance
High-leverage firms suffered more financial stress during late February ... and in four other Group of Seven economies, and small firms
and March in the United States ... underperformed in four economies.
1. Cumulative Abnormal Return of US Firms with Low and High 2. High-Leverage Firms’ and Small Firms’ Relative Equity Performance
Leverage during February–March 2020 during February–March 2020
(Percent) (Percentage points)
5 10
High leverage Small
0 5
–5
0
–10
–5
–15
Low leverage
–20 –10
High leverage
–25 –15
03 Feb. 10 Feb. 17 Feb. 24 Feb. 02 Mar. 09 Mar. 16 Mar. 23 Mar. 30 Mar. Canada France Germany Italy Japan United United
2020 20 20 20 20 20 20 20 20 Kingdom States
US firms with less cash than their industry peers suffered more ... as did UK firms with relatively less cash and French, UK, and US
financial stress during late February and March ... firms with a high liquidity gap.
3. Cumulative Abnormal Returns of US Firms with Low and High 4. Liquidity-Poor and Cash-Poor Firms’ Relative Equity Performance during
Relative Cash during February–March 2020 February–March 2020
(Percent) (Percentage points)
5 4
Low relative cash High liquidity gap
0 2
0
–5
–2
–10
–4
High relative cash
–15 –6
Low relative cash
–20 –8
03 Feb. 10 Feb. 17 Feb. 24 Feb. 02 Mar. 09 Mar. 16 Mar. 23 Mar. 30 Mar. Canada France Germany Italy Japan United United
2020 20 20 20 20 20 20 20 20 Kingdom States
Sources: Refinitiv Datastream; S&P Capital IQ; and IMF staff calculations.
Note: Firm characteristics are as of the end of the fourth quarter of 2019. Leverage in panels 1 and 2 is defined as the debt-to-asset ratio. A high-leverage
(low-leverage) firm is one in the top (bottom) half of the leverage distribution. In panels 2 and 4, equity performance is based on cumulative abnormal returns during
February 3–March 31, 2020, and firm-level characteristics are controlled for. “Relative cash” is defined as in Joseph and others (2020), and a low-relative-cash
(high-relative-cash) firm is one in the lowest (highest) tercile of the relative cash distribution. “Small” is defined as being in the lowest tercile of the distribution of
total assets. “Liquidity gap” is defined as total short-term financing minus cash and short-term investments as a ratio of total assets. A high-liquidity-gap firm is one
in the highest tercile of the distribution. Solid colored bars indicate statistical significance at the 5 percent level. Empty bars indicate lack of statistical significance at
conventional levels. See Online Annex 3.5 for methodological details.
A variety of policy measures—monetary, fiscal, and well as with the announcement of containment policy
financial—were announced over a short period of time, measures imposing restrictions on economic activity—
sometimes on the same day, making it difficult to iso- assessment of their impact is extremely difficult.34 In
late their effects. Important details of announced policy the face of these challenges, and with full acknowledg-
packages were sometimes released with a lag, and pol- ment of the associated limitations, this chapter follows
icy measures announced on different days could have
had strong complementarities. Furthermore, because 34For example, the March 12 announcement by the Federal
many of the economic policy measures announced Reserve Bank of New York of new large repo operations coincided
early on in the crisis were concurrent with negative with one of the worst declines in US stock market history. The
announcement, however, was a surprise and took place in the
news about the progression of the pandemic and its middle of the trading day, at a time when the intraday decline was
effect on the real economy and financial markets—as already very large.

two simple approaches to try to gauge the impact as macroprudential measures or changes in financial
of key policy announcements on corporate funding sector regulation, have only an indirect impact. Com-
liquidity stress. First, it examines the effect of policy paring announcement days when at least one policy
announcements on the relative stock market perfor- with a direct impact was announced with those when
mance of the most vulnerable firms over a horizon policies with only an indirect impact were announced,
of two trading days, taking into account the negative it appears that policies with a direct impact benefited
impact of global financial market volatility during days firms with liquidity vulnerabilities relatively more.38
when it was extreme.35,36 Second, it assesses the overall The effect amounts to 0.2 percentage point of overper-
impact of the policy response by extending the window formance a day over two days for liquidity-poor firms
of the analysis (to the end of June) of the relative stock and to 0.13 percentage point a day over two days for
market performance of the groups of vulnerable firms cash-poor firms (Figure 3.7, panel 2). No difference
that have underperformed during February–March, across types of policies is observed for high-leverage
as identified in the previous section. In both cases, firms and small firms.39
several firm characteristics are controlled for.37 As in The analysis of the stock market performance of
the previous section, the relative performance of firms vulnerable firms through the end of June confirms
most vulnerable to adverse funding liquidity shocks that stress at smaller firms had generally disappeared
(controlling for solvency and other firm characteristics) by then—except in the United Kingdom, where it
is interpreted as a symptom of changing credit supply remained significant—while strains in high-leverage
conditions. The focus on those firms does not suggest firms remained in Germany and Japan (Figure 3.7,
that policies explicitly targeted them but that policies panel 3). Stress at firms with liquidity vulnerabilities,
to support the economy (and credit provision in par- however, persisted in France, the United Kingdom,
ticular) may benefit them relatively more. and the United States (Figure 3.7, panel 4), echoing
Policy announcements appear to have had a positive findings from the aggregate analysis of the loan mar-
effect on the relative stock market performance of kets in the US economy.
smaller firms (relative to larger firms) as well as on
those with high leverage (relative to those with low
leverage). Pooling all 85 announcement days in the Conclusion and Policy Considerations
sample, this effect amounts to about 0.3 percentage The tightening of credit conditions that took place
point of overperformance a day over two days for across G7 economies in March as the COVID-19
smaller firms and about 0.1 percentage point a day pandemic gathered momentum was quelled to a very
over two days for high-leverage firms. By contrast, no large extent thanks to an unprecedented set of powerful
significant effect can be found for firms with liquidity
vulnerabilities (Figure 3.7, panel 1). Given the small 38When estimated separately, the effect of measures with an
number of announcement days, identifying significant indirect impact is not statistically significant. It is plausible that
effects at the country level is challenging. Yet the data such measures, including changes in financial sector regulation or
macroprudential policy, take longer to have an effect on financ-
suggest a positive effect for small firms in Canada and ing conditions for nonfinancial firms than measures with a direct
for small firms and high-leverage firms in Japan. impact. Among measures with a direct impact, the announcements
It is plausible that some types of vulnerable of on-budget fiscal measures supporting firm solvency appear to
have been the most powerful: excluding announcement days when
firms were more affected by certain types of policy such measures were announced, the difference between the effect
announcements than others. Some policies, such as of measures with a direct impact and those with an indirect impact
government guarantees or purchases of corporate loses significance. Among the other four types of measures with a
direct impact, corporate asset purchase programs appear to have been
securities by central banks, have a direct impact on
relatively more powerful.
corporate funding and solvency, whereas others, such 39While it is very plausible that major policy announcements in
the United States had positive spillover effects on other G7 econo-

mies, spillover analysis is impeded by the occasional concurrence of
35The analysis does not try to assess whether program eligibility major announcements in the United States with those in the other
mattered for firms’ financial performance. countries. Focusing on days when an announcement was made in
36Global financial market volatility is defined as extreme when the the United States only, no evidence can be found that the announce-
Chicago Board Options Exchange Volatility Index (VIX) is above the ment had a positive effect on the relative performance of vulnerable
80th percentile of its distribution during February–June 2020. firms in other G7 economies. Spillovers to emerging markets are
37See Online Annex 3.6 for methodological details. discussed in Chapter 2 of this report.

Figure 3.7. The Effect of Policies on Vulnerable Firms
Policy announcements helped relieve financial stress on average in ... and policies targeting the corporate sector directly had a stronger
small firms and high-leverage firms ... effect on cash-poor and liquidity-poor firms than policies with an
indirect impact.
1. Effect of Policy Announcements on the Relative Equity Performance 2. Relative Effect of Announcements of Policies Targeting the Corporate
of Vulnerable Firms Sector Directly on the Relative Equity Performance of Vulnerable Firms
(Percentage points, average effect over two days) (Percentage points, average effect over two days)
0.4 0.3
0.3 0.2
0.2 0.1
0.1 0.0
0.0 –0.1
–0.1 –0.2
High leverage Small Low relative High liquidity High leverage Small Low relative High liquidity
cash gap cash gap
The relative performance of small firms improved during the second ... but strains remained for liquidity-poor and cash-poor firms at the
quarter ... end of June.
3. High-Leverage and Small Firms’ Relative Equity Performance during 4. Cash-Poor and Liquidity-Poor Firms’ Relative Equity Performance during
February–June 2020 February–June 2020
(Percentage points) (Percentage points)
High leverage (Feb.–June) Small (Feb.–June) High liquidity gap (Feb.–June) Low relative cash (Feb.–June)
High leverage (Feb.–Mar.) Small (Feb.–Mar.) High liquidity gap (Feb.–Mar.) Low relative cash (Feb.–Mar.)
0 0
–4
–4
–8
–8
–12
–16 –12
Germany Japan United Kingdom United States France United Kingdom United States
Sources: IMF, COVID Policy Tracker; press releases and press reports; Refinitiv Datastream; S&P Capital IQ; Yale Program on Financial Stability; and IMF staff
calculations.
Note: In panels 1 and 2, the effect of policy announcements is calculated net of the effect of extreme volatility, and equity performance is based on cumulative
abnormal returns on the day of the policy announcement and the following day. Leverage is defined as the debt-to-asset ratio. A high-leverage (low-leverage) firm is
one in the top (bottom) half of the leverage distribution. “Relative cash” is defined as in Joseph and others (2020), and a low-relative-cash (high-relative-cash) firm is
one in the lowest (highest) tercile of the relative cash distribution. “Small” is defined as being in the lowest tercile of the distribution of total assets. “Liquidity gap” is
defined as total short-term financing minus cash and short-term investments as a ratio of total assets. A high-liquidity-gap firm is one in the highest tercile of the
distribution. In panels 3 and 4, equity performance is based on cumulative abnormal returns during February 3–June 30, 2020. Solid colored bars indicate statistical
significance at the 5 percent level. Empty bars indicate lack of statistical significance at conventional levels. See Online Annex 3.6 for methodological details.

policy interventions. Despite the deterioration in its strains in smaller firms. It is thus critical to carefully
solvency, the nonfinancial corporate sector, as a whole, calibrate any withdrawal of fiscal policy support to
was generally able to obtain the funding it needed to funding markets.
continue operating during the second quarter.40 Yet Beyond the calibration of funding and liquidity
signs of tighter credit conditions also surfaced during support by fiscal and monetary policymakers, a key
the second quarter in some segments of the credit issue for financial stability in the near to medium term
market or did not fully dissipate for some types of firms will be the deterioration in corporate solvency as a
with a viable business model but vulnerable to adverse result of the pandemic-induced decline in profitability
liquidity shocks. In particular, while US bond markets and increased corporate indebtedness. This deteriora-
have been buoyant, bank-dependent firms, as well as tion will have a severe impact on banks’ asset quality
those with pre–COVID-19 liquidity vulnerabilities, con- and capital adequacy (see Chapter 4), which in turn
tinue to face a more difficult environment. Firms with could limit the credit supply to firms over the next
pre–COVID-19 liquidity vulnerabilities in the United several quarters.
Kingdom also appear to have been left behind, despite Chapter 1 of this report provides a policy road map
overall favorable credit conditions. An interesting topic to navigate the gradual reopening and the recovery
for future analysis would be further exploration of the phases of the COVID-19 crisis (see Table 1.2 in that
reasons for the cross-country differences in the evolution chapter) and discusses policy trade-offs relevant to
of credit supply conditions documented in the chapter. corporate funding issues documented in this chapter,
While most G7 central banks have already signaled including the impact on fiscal space and sovereign
their intention to leave their pandemic-related facilities contingent liabilities as well as the risk of capital misal-
in place for the foreseeable future, it may be increas- location. Once the recovery is well entrenched, the expe-
ingly difficult for governments to maintain the same rience of the COVID-19 shock on corporate funding
level of fiscal support because of fiscal space concerns markets must also be examined to determine the reasons
or other political economy considerations. The latest for the fragility they experienced in March. The regula-
bank lending survey of the euro area suggests that tion of nonbank financial institutions must be revisited
tighter bank lending standards may be around the and mechanisms to enhance their resilience to large
corner, as government guarantee programs are set liquidity shocks devised, as discussed in recent GFSRs.
to end soon (European Central Bank 2020). Yet the The evidence provided in this chapter also indicates
evidence analyzed in this chapter suggests that it is the that liquidity and leverage-related vulnerabilities have
policies supporting firms directly that have had the amplified the impact of the COVID-19 shock. The
most beneficial effect on firms with liquidity vulnera- experience of the current crisis, therefore, is a reminder
bilities. Policies also appear to have cushioned financial to supervisory authorities to continue to monitor
corporate vulnerabilities closely and offers an oppor-
40Because
tunity for them to consider the benefits of macro-
of lack of firm-level data for unlisted small and
medium-sized enterprises in 2020, the analysis could not establish prudential policy tools for the nonfinancial corporate
the degree to which this conclusion carries over to those firms. sector (IMF 2020).

References Eren, Egemen, Andreas Schrimpf, and Vladyslav Sushko.

Acharya, Viral V., and Sascha Steffen. 2020. “The Risk of 2020. “US Dollar Funding Markets during the COVID-19
Being a Fallen Angel and the Corporate Dash for Cash in Crisis—The Money Market Fund Turmoil.” BIS Bulletin 14,
the Midst of COVID.” CEPR COVID Economics Vetted Bank for International Settlements, Basel.
and Real-Time Papers 10, Centre for Economic Policy European Central Bank. 2020. “The Euro Area Bank Lend-
Research, London. ing Survey—Second Quarter.” https://www.ecb.europa
Adrian, Tobias, Paolo Colla, and Hyun Song Shin. 2013. .eu/stats/ecb_surveys/bank_lending_survey/html/ecb
“Which Financial Frictions? Parsing the Evidence from the .blssurvey2020q2~d8de5b89f0.en.html#toc1
Financial Crisis of 2007 to 2009.” NBER Macroeconomics Gilchrist, Simon, and Egon Zakrajšek. 2012. “Credit Spreads
Annual 27 (1): 159–214. and Business Cycle Fluctuations.” American Economic Review
Anderson, Julia, Francesco Papadia, and Nicolas Véron. 2020. 102 (4): 1692–720.
“Government-Guaranteed Bank Lending in Europe: Beyond Gourinchas, Pierre-Olivier, Sebnem Kalemli-Ozcan, Veronika
the Headline Numbers.” Peterson Institute for International Penciakova, and Nick Sander. Forthcoming. “COVID-19
Economics. https://www.piie.com/blogs/realtime-economic and Business Failures.” IMF Working Paper, International
-issues-watch/government-guaranteed-bank-lending-europe Monetary Fund, Washington, DC.
-beyond-headline Hadlock, Charles J., and Joshua R. Pierce. 2010. “New Evidence
Banerjee, Ryan, Anamaria Illes, Enisse Kharroubi, and on Measuring Financial Constraints: Moving Beyond the KZ
José-Maria Serena. 2020. “COVID-19 and Corporate Sector Index.” Review of Financial Studies 23 (5): 1909–40.
Liquidity.” BIS Bulletin 10, Bank for International Settle- Holmstrom, Bengt, and Jean Tirole. 1997. “Financial Inter-
ments, Basel. mediation, Loanable Funds, and the Real Sector.” Quarterly
Bank of England. 2020. “Credit Conditions Survey—Quarter Journal of Economics 112 (3): 663–91.
Two.” https://www.bankofengland.co.uk/credit-conditions International Monetary Fund (IMF). 2020. “United States:
-survey/2020/2020-q2 Financial System Stability Assessment.” IMF Country Report
Becker, Bo, and Victoria Ivashina. 2014. “Cyclicality of 20/242, International Monetary Fund, Washington, DC.
Credit Supply: Firm Level Evidence.” Journal of Monetary Ivashina, Victoria, and David Scharfstein. 2010. “Bank Lending
Economics 62: 76–93. during the Financial Crisis of 2008.” Journal of Financial
Board of Governors of the Federal Reserve System. 2020. Economics 97 (3): 319–38.
“The July 2020 Senior Loan Officer Opinion Survey on Joseph, Andreas, Christiane Kneer, Neeltje Van Horen, and
Bank Lending Practices.” Washington, DC. https://www Jumana Saleheen. 2020. “All You Need Is Cash: Corporate
.federalreserve.gov/data/sloos/sloos-202007.htm Cash Holdings and Investment after the Financial Crisis.”
Chodorow-Reich, Gabriel. 2014. “The Employment Effects of Bank of England Working Paper 843, London.
Credit Market Disruptions: Firm-Level Evidence from the Kapan, Tumer, and Camelia Minoiu. 2020. “Liquidity Insurance
2008–09 Financial Crisis.” Quarterly Journal of Economics versus Credit Provision: Evidence from the COVID-19
129 (1): 1–59. Crisis.” Unpublished.
Duchin, Ran, Oguzhan Ozbas, and Berk A. Sensoy. 2010. Li, Lei, Yi Li, Marco Macchiavelli, and Xing Alex Zhou. 2020.
“Costly External Finance, Corporate Investment, and the “Runs and Interventions in the Time of COVID-19: Evi-
Subprime Mortgage Credit Crisis.” Journal of Financial Eco- dence from Money Funds.” Social Research Science Network.
nomics 97 (3): 418–35. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3607593
Duval, Romain, Gee Hee Hong, and Yannick Timmer. 2020. Moody’s Investors Service. 2020. “Default Report.” April.
“Financial Frictions and the Great Productivity Slowdown.” Rigobon, Roberto. 2003. “Identification through Heteroskedas-
Review of Financial Studies 33 (2): 475–503. ticity.” Review of Economics and Statistics 85 (4): 777–92.

4
CHAPTER
BANK CAPITAL
COVID-19 CHALLENGES AND POLICY RESPONSES
•• The coronavirus disease (COVID-19) crisis may pose challenges to the capital of banks, even though they
entered the crisis with higher capital ratios than before the global financial crisis and despite the large
policy interventions aimed at containing the economic fallout from the current crisis.
•• Forward-looking simulations based on a new global stress test tool show that in a baseline scenario
consistent with the October 2020 World Economic Outlook (WEO) bank capital falls sharply but recovers
quickly, while an adverse scenario suggests sustained damage to average capital ratios.
•• In the adverse scenario, a weak tail of banks, corresponding to 8.3 percent of banking system assets, would
fail to meet minimum regulatory requirements, and the capital shortfall relative to broad statutory regula-
tory thresholds reaches $220 billion.
•• In absence of the bank-specific mitigation policies already implemented, the weak tail of banks would
reach 14 percent of banking system assets, and the global capital shortfall would be $420 billion.
•• Bank-specific mitigation policies would help reduce financial stability risks if the crisis recedes promptly
but may pose risks to banks’ capital adequacy if the crisis proves to be longer lasting.
Will Banks Remain Adequately Capitalized? sample, might fail to meet minimum regulatory capital
Banks entered the current COVID-19 crisis with higher requirements in an adverse scenario. Government loan
levels of capital than before the global financial crisis, and guarantees and other bank-specific policies that adjust
policymakers have quickly deployed an array of policies the calculation of capital ratios help relieve the decline of
to support economic activity and the ability of banks to reported capital ratios and reduce the incidence of bank
lend. However, the sheer size of the shock and the likely capital shortfalls. In considering the duration of these and
increase in defaults from firms and households may pose other measures, policymakers should pay attention to the
challenges to banks’ profitability and capital positions. intertemporal trade-off they pose, as policies that reduce
A forward-looking simulation of the trajectory of capital the financial stability risks of a transitory shock may
ratios in a sample of about 350 banks from 29 jurisdic- increase vulnerabilities related to banks’ loss-absorbing
tions, accounting for 73 percent of global banking assets, capacity and overall indebtedness if the crisis proves to be
shows that such ratios would decline as a result of the persistent. Policies aimed at limiting capital distributions
COVID-19 crisis, but remain, on average, comfortably and ensuring adequate funding for deposit guarantee
above regulatory minimums. However, there is hetero- programs, as well as contingency plans that lay out how
geneity across and within regions, and a weak tail of to respond to possible pressures, would help deal with
banks, accounting for 8.3 percent of banking assets in the the consequences of a potentially adverse scenario.
Prepared by staff from the Monetary and Capital Markets

Department (in consultation with other departments): The authors Introduction
of this chapter are John Caparusso and Claudio Raddatz (lead
In many respects, the COVID-19 crisis presents
authors), Yingyuan Chen, Dan Cheng, Xiaodan Ding, Ibrahim
Ergen, Marco Gross, Ivo Krznar, Dimitrios Laliotis, Fabian the largest shock that banks have experienced since
Lipinsky, Pavel Lukyantsau, Elizabeth Mahoney, Nicola Pierri, and the Great Depression (see the October 2020 WEO).
Tomohiro Tsuruga with contributions from Hee Kyong Chon, Authorities have adopted unprecedented policy
Caio Ferreira, Alejandro Lopez, and Luc Riedweg under the guid-
ance of Fabio Natalucci (Deputy Director). Magally Bernal was measures to blunt the impact of this shock. Govern-
responsible for word processing and the production of this report. ments have introduced substantial fiscal support to

households and businesses (see the October 2020 The chapter also discusses policy options to deal
Fiscal Monitor), monetary policy rates have been with the potential challenges that banks could face
cut worldwide, and many central banks have imple- in the baseline and adverse scenarios, and highlights
mented large asset purchase programs to support the intertemporal trade-off that arises from targeted
markets and to maintain the credit flow to the real policies that encourage banks to use the flexibility
economy (see the April 2020 Global Financial Stabil- embedded in the regulatory regime to sustain the flow
ity Report [GFSR]). of credit to borrowers facing liquidity problems in
Importantly, policymakers have taken steps to avoid response to a transitory shock.
the procyclical credit crunch that was evident during
the global financial crisis, encouraging banks to use the
flexibility embedded in the global regulatory frame- Initial Impact of COVID-19 on the Global
work to deal with the temporary consequences of the Banking Industry
COVID-19 shock and thus stifle negative feedback After spending the past decade building capital and
loops that could amplify the impact of the crisis. liquidity buffers following the regulatory reforms put
Following a decade during which banks aggressively in place after the global financial crisis, banks came
built their capital positions, standard setting bodies into the COVID-19 crisis in much better shape than
have issued guidance to support national authorities they did before previous crises (Figure 4.1, panel 1).
in their policy response to the pandemic. Policymak- However, bank profitability was already challenged
ers have released capital buffers to sustain the flow of in many jurisdictions amid the prolonged period
credit to households and firms. Banks have also been of low interest rates and low term spreads in recent
allowed, for loans whose deterioration is attributed to years (Figure 4.1, panel 2). This low-interest-rate
the shock, to defer the recognition of bad debts and environment is likely to persist for several years, as
the reporting of loan loss provisions and to waive the policymakers have engaged in further expansive mon-
increase in risk-asset weightings and the deduction of etary policies to support the flow of credit to the real
provision charges from capital. Banks have also been economy (see the April 2020 GFSR).
compelled (by regulation or strong administrative Despite the stronger initial position of banks and
guidance) to cancel capital distributions. the aggressive response of policymakers, the initial
Despite the large negative impact of the pandemic stage of the COVID-19 crisis has confronted banks
on the global economy during recent quarters, banking with significant challenges. The initial contrac-
systems have so far been able to weather these eco- tionary shock triggered a scramble for liquidity. In
nomic difficulties, due in part to aggressive policy the United States, corporate borrowers aggressively
support. Following an initial plunge, bank equity drew on committed credit lines, causing a sudden
prices have partially recovered. While banks’ assessment increase in loans that drove down bank capital
of borrower credit quality has naturally deteriorated, ratios.1 Since then, bank credit in the United States
bank credit expanded in March as corporate bor- and Europe has remained largely flat. Crucial
rowers drew on committed credit lines and has since elements of financial system plumbing (for exam-
remained stable. Nonetheless, credit conditions have ple, repo and US Treasury markets) encountered
remained tight. Despite significantly increased loan loss liquidity challenges, as did emerging market banks
provisions in virtually all systems, most banks con- in US funding markets, and financial markets were
tinue to report positive earnings, and capital positions severely stressed for several weeks. Increased loan
have declined only modestly over the initial quarters loss provisioning—particularly among US banks,
of the crisis. for which the onset of the crisis coincided with
This chapter addresses two central questions.
•• How prepared are banks to withstand continued chal- 1Risk weights for undrawn credit lines are in the range of
lenging economic conditions in the coming years? 20–50 percent, whereas those for drawn credit lines are 100 percent.
Therefore, the large drawdown of committed credit lines has an
•• How much would bank-specific regulatory policies immediate material impact on risk-weighted assets, the denominator
recently implemented help them face these scenarios? of bank capital ratios.

CHAPTER 4 Bank Capital: Covid-19 Challenges and Policy Responses
Figure 4.1. Historical Context: Magnitude of the Current Crisis and the Ex Ante Position of Banks
Banks, particularly in Europe and in emerging market economies, ... despite low profitability challenging capital accretion in some
massively improved their capital positions in the last decade ... regions.
1. Average Tier 1 Ratio, by Region 2. Average Return on Equity, by Region
(Percent) (Percent)
NA EU Other AEs EMs Total NA EU Other AEs EMs Total
17 20
16
15
15
14 10
13 5
12
0
11
10 –5
2010 11 12 13 14 15 16 17 18 19 2010 11 12 13 14 15 16 17 18 19
Bank lending standards tightened sharply—to near the 2008 peak in Banks attribute tightening to deteriorating borrower conditions, not to
the United States. capital or liquidity constraints.
3. Bank Lending Standards: Net Tightness 4. Causes of Bank Credit Tightening
(Percentage points) (Percentage points)
100
United States Europe Japan
Worsening industry problems
80
60 Uncertain economic outlook
40 Reduced risk tolerance
20 Decreased secondary market risk

0 Deterioration of bank liquidity
–20
Bank capital deterioration
–40
2008 09 10 11 12 13 14 15 16 17 18 19 20 –100 –50 0 50 100
UNIMPORTANT IMPORTANT
Source: Haver Analytics.

Note: Bank lending standards for Europe are based on the European Central Bank’s one-quarter forward expectations, while both the U.S. and Japan are based on
the most recent quarter. Other AEs = other advanced economies, including Japan, Australia, Hong Kong SAR, and Singapore; EMs = emerging markets; EU = Europe,
including the United Kingdom and continental Europe; NA = North America, including United States and Canada.
a transition to “expected credit loss” accounting financial market stress subsided, but most banks
standards—weighed on bank financial results in took sharply higher loan loss provisions and tight-
the first quarter of 2020.2 In the second quarter, ened lending standards as the economic outlook
continued to deteriorate (Figure 4.1, panel 3), with
2The transition to expected credit losses in the United States
became effective on January 1, 2020, and virtually all US banks

chose to book large provisions for “transitional” increases in loan bank regulations mitigate the impact of this transition charge on
loss reserves. In one extreme example, Citi took a $4.2 billion bank capital. Before the COVID-19 outbreak, the Federal Reserve
current expected credit losses transitional charge, more than half announced a rule allowing banks to phase in the impact of current
of the $7 billion total 2020 first-quarter loan loss provision. The expected credit losses transition provisions over three years. During
Federal Reserve promulgated a regulation allowing banks to defer the first quarter of 2020, the regulator lengthened the phase-in
transition-related provisions, but most large banks chose to retain path to zero capital charges over two years, followed by a three-year
the transition charges recognized on January 1. However, US phase-in path.

Figure 4.2. Mitigation Policies Announced since February 1, 2020, by Category and Jurisdiction
Among the wide range of policy responses to the COVID-19 shock and slowdown, this chapter focuses on three that relate most directly.
Number of policy announcements: > 25 11–25 6–10 2–5 1 0
USA CAN GBR CHE DEU FRA ITA ESP PRT BEL NLD IRL SWE NOR FIN AUS JPN KOR HKG SGP IND IDN BRA MEX ZAF
Borrower Support
New loans
Loan guarantees
Repayment relief
Loss Recognition
IFRS/CECL relaxation
Recognition deferral
Capital Adequacy
Lower buffers
Lower RWA and leverage exposures
Lower capital deductions
Sources: Financial Stability Board; KBW; Yale School of Management; and IMF staff estimates.
Note: The intensity of the colors in the figure denotes only the number of measures announced but has no bearing on the absolute or relative economic magnitude of
those policies. For instance, a single large policy announcement in one jurisdiction could surpass in economic relevance many announcements by a different
jurisdiction. The figure includes policy announcements up to July 10, 2020. Austria, Denmark, Greece, and Luxembourg are not included in the analysis due to
incomplete data. See Online Annex 4.1, www.imf.org/en/Publications/GFSR, for an explanation of the data and methodology on which this policy taxonomy is based.
The row labeled “Lower buffers” also includes public announcements by authorities explicitly encouraging banks to use the flexibility embedded in the regulatory
framework to use the capital conservation buffer to support lending, although these statements do not entail a formal change in the rulebook. Data labels use
International Organization for Standardization (ISO) country codes. CECL = current expected credit loss; IFRS = International Financial Reporting Standards;
RWA = risk-weighted assets.
loan officers in the United States reporting the This chapter focuses specifically on the impact of
tightest credit standards since 2005. government loan guarantee programs and capital ade-
As improved liquidity conditions relieved bor- quacy policies that can be directly quantified (henceforth,
rowers’ appetite for precautionary borrowing, the “bank-specific” policies). Other policies have an indirect
first-quarter spurt of loan growth slowed or reversed effect on banks’ capital adequacy. For example, fiscal stimu-
for most banks. This relieved risk-weighted asset lus and monetary policy indirectly support banks’ finan-
pressure on capital ratios (Figure 4.1, panel 4). During cial results through macroeconomic channels. Policies to
the second quarter of 2020, some major banks support bank funding could affect bank capital by lowering
(particularly in the United States) also reported large costs and allowing banks to sustain their level of activity.
capital-market-driven gains. Policies intended to support borrowers’ repayment ability,
including repayment moratoria, may reduce banks’ need
to set aside provisions for loan losses—and thus bolster
The Reactions of Financial Sector Authorities to capital—by lowering the probability that a borrower will
the COVID-19 Crisis enter default (probability of default). Nonetheless, some of
Governments around the world have responded to these policies may also simply postpone loss recognition.
the economic disruption of the COVID-19 crisis with Within the risk-based capital framework, the poli-
policies of unprecedented scope and magnitude to cies analyzed in this chapter can alter the capital space
support the real economy, prevent permanent damage through three channels.
to the balance sheets of firms and households, and •• Increasing capital levels: This has been promoted
maintain the flow of credit to the real economy. These mainly through restrictions (often “voluntary” guid-
policies extend from broad macroeconomic policies ance) on distribution of profits through dividends
to specific measures that directly address bank balance and share buybacks. Most of these come with specific
sheet management (Figure 4.2). end dates (typically not later than the end of 2020).

Figure 4.3. Magnitude of Announced Mitigation Policies
The magnitude of loan guarantees varies widely across countries. Many jurisdictions have relaxed statutory capital buffer requirements to
support banks’ credit underwriting.
1. Loan Guarantees 2. Change in Statutory Bank Capital Buffers since February 1, 2020
(Percent of GDP) (Percent of risk-weighted assets)
40 1
CCyB CCB D-SIB/Systemic Risk Buffers
35
30 0
25
20 –1
15
10 –2
5
0 –3
US
GB
CH
SSM
DE
FR
IT
ES
PT
BE
NL
LU
IE
AT
SE
NO
FI
BR
AU
JP
KR
SG
IN
ID
BR
MX
US
GB
CH
SSM
DE
FR
IT
ES
PT
BE
NL
LU
IE
AT
SE
NO
FI
GR
AU
JP
KR
SG
IN
ID
BR
MX
CA
ZA
CA
ZA
DK
HK
DK
HK
Some jurisdictions have also taken steps to improve reported capital A few countries highly sensitive to capital market depth have also
ratios or lower required capital buffers. taken steps to improve leverage ratios.
3. Estimated Pro Forma Increase in CET1 Capital Ratio and Buffer 4. Increase in Leverage Ratio from Announced Policies
from Announced Policies (Percentage points)
(Percentage points)
7 CET1 Buffer - Statutory Buffer - Guidance 1.8
6 1.6
1.4
5
1.2
4 1.0
3 0.8
0.6
2
0.4
1 0.2
0 0.0
US
GB
CH
SSM
DE
FR
IT
ES
PT
BE
NL
LU
IE
AT
SE
NO
FI
GR
AU
JP
KR
SG
IN
ID
BR
MX
US
GB
CH
SSM
DE
FR
IT
ES
PT
BE
NL
LU
IE
AT
SE
NO
FI
GR
AU
JP
KR
SG
IN
ID
BR
MX
CA
ZA
CA
ZA
DK
HK
DK
HK
Sources: Bloomberg Finance L.P.; Financial Stability Board; IMF (2020b); KBW; SNL Financial; Yale School of Management; and IMF staff estimates.
Note: Figures include the 29 countries captured in the bank stress test, plus data on the SSM as a supervisory jurisdiction. “Loan guarantees” is based on the
announced programs, not actual take-up of guaranteed loans. Loan guarantee data are not captured for Austria, Finland, Greece, Hong Kong SAR, Ireland,
Luxembourg, and Portugal. D-SIB surcharges are not captured as a separate buffer in several jurisdictions, mainly because D-SIB requirements are often expressed
in terms of the overall CET1 ratio. Countries are identified by two-digit International Organization for Standardization (ISO) code and indicate policies pronounced by
the European Central Bank and the European Banking Authority. Figures for individual European countries indicate local policies distinct from those announced by
European authorities. CCB = capital conservation buffer; CET1 = common equity Tier 1; CCyB = countercyclical capital buffer; D-SIB = domestic systemically
important bank; SSM = Single Supervisory Mechanism.
Policymakers have issued such guidance for the large panel 1).3 In some instances, policymakers have also
European banks and for all banks in Brazil, Italy, Spain, reduced risk weights on banks’ exposures to targeted
Switzerland, the United Kingdom, and other countries. borrowers, often small businesses, to encourage
Government loan guarantees can also boost capital lev- credit to this segment. A few countries—Japan,
els by reducing the loss that a bank experiences when the United Kingdom, and the United States—have
a borrower defaults and the need to set aside loan loss exempted central bank reserves, and the latter
provisions for this event (loss given default).
•• Lowering risk-weighted assets or “leverage expo- 3This is distinct from the effect of government guarantees on
sure”—the capital ratio denominators: National the borrowers’ “point-in-time” probability of default resulting from
improved access to funding—which is captured in the analysis of the
regulators have typically waived risk-asset weights for corporate sector—and from their effect on the “loss given default,”
loans covered by government guarantees (Figure 4.3, previously discussed and quantified in the next section.

two have exempted holdings of government bond banking systems in a forward-looking manner using
holdings, from banks’ leverage exposure measures the latest baseline projection of the economic outlook
(the denominator of the leverage ratio). These and the adverse scenario outlined in the October 2020
policies are intended to facilitate large asset purchase WEO (Figure 4.4). These two scenarios provide a
programs and to encourage banks to continue to broad assessment of the potential paths of the pan-
intermediate in government bond markets. demic; however, given the unprecedented nature of the
•• Releasing some capital buffers: In many jurisdic- shock, uncertainty remains.
tions, policymakers have increased banks’ overall These macro scenarios implicitly incorporate the
space between reported and regulatory capital levels effects of broad macroeconomic and monetary policy
by releasing the countercyclical capital buffer that is interventions, including interest rate cuts, unconven-
designed to be used during downturns (Figure 4.3, tional monetary policies, fiscal measures, social safety
panel 2). In some instances, policymakers have net packages, and other policies that support the real
formally released required capital buffers, effecting a economy. By improving the liquidity of borrowers,
reduction in statutory capital buffers. In other cases, these policies indirectly affect the condition of banks.
policymakers have publicly reminded banks that However, the consequences of bank-specific policies
some buffers—typically the capital conservation buf- for the distribution of banks’ capital may not be fully
fer of 2.5 percent of total capital aimed at preventing captured in macro aggregates. The chapter also assumes
banks from breaching the minimum regulatory capi- that the accounting impact of bank-specific policies
tal adequacy ratio—could be used to support lending on bank balance sheets is not fully captured in macro
and be gradually rebuilt through retained earnings trajectories.
as conditions improve. This chapter characterizes The assessment relies on a recently developed global
the latter as reductions in the “guidance buffer” that stress test (see Online Annex 4.1) that uses publicly
determines de facto minimum capital levels. available data on the financial statements of about
350 banks in 29 major banking systems—accounting
These policies combined are estimated to have already for 73 percent of global banking sector assets—to
improved banks’ reported common equity Tier 1 estimate how key components of banks’ financial state-
(CET1) ratios and, either by statute or by guidance ments react to macroeconomic variables.5 The future
releasing some capital buffer requirements, regulators paths of these variables are embedded in the scenar-
have further expanded the capital space between banks’ ios used to conduct a forward-looking simulation of
current positions and broad regulatory capital levels the evolution of the profitability and capital position
(Figure 4.3, panel 3).4 In addition, although this section of each of the banks in the sample, which is then
focuses on the CET1 capital position because that is aggregated across different regions and across global
the binding constraint for most banking systems where systemically important banks.
bank market-making activity is not large, policymakers The stress test exercise relies on publicly available
in a few jurisdictions (Japan, Switzerland, United States) data. While this allows for a global assessment of the
have also eased constraints on banks’ leverage ratios, prospective health of the banking system, it comes
typically by excluding government bonds, central bank at the cost of lower data granularity and higher
reserves, or other low-risk assets from the leverage expo- reliance on statistical methods than in supervisory
sure denominator (Figure 4.3, panel 4). stress tests. This narrows the types of policies that can
be analyzed in this context and also requires several
assumptions to map the impact of those policies to
Bank Capital Ratios in the Wake of COVID-19
and the Role of Policies
5Online Annex 4.1 is available at www.imf.org/en/Publications/
This chapter assesses the consequences of the GFSR. The jurisdictions included are Australia, Austria, Belgium,
COVID-19 crisis for the future capital ratios of global Brazil, Canada, Denmark, Finland, France, Germany, Greece,
Hong Kong SAR, India, Indonesia, Ireland, Italy, Japan, Korea,
Luxembourg, Mexico, The Netherlands, Norway, Portugal,
4Capital requirements that include all statutory buffers (but Singapore, South Africa, Spain, Sweden, Switzerland, the United
exclude recent statutory reductions) are defined in this chapter as Kingdom, and the United States. In each jurisdiction, the largest
“statutory broad capital requirements.” Capital requirements that banks covering up to 80 percent of banking assets are included.
exclude buffers released by recent informal guidance statements are Therefore, the simulation does not include the consequences of the
defined as “guidance capital requirements.” scenarios for the solvency of small banks.

Figure 4.4. Scenarios for Stress Test Simulation

1. Real GDP Growth 2. Unemployment Rate
(Year over year, percent) (Percent)
6 9
4 8
2 7
6
0
5
–2
4
–4 AE: Baseline AE: Baseline 3
–6 AE: Adverse AE: Adverse
EM: Baseline EM: Baseline 2
–8 EM: Adverse EM: Adverse 1
–10 0
2019 20 21 22 2019 20 21 22
3. Short-term Interest Rates 4. Term Spread

(Percent) (Percent)
7 4
AE: Baseline
6 AE: Adverse
EM: Baseline
5 3
EM: Adverse
4
2
3 AE: Baseline
2 AE: Adverse
EM: Baseline 1
1 EM: Adverse
0 0
2019 20 21 22 2019 20 21 22
Source: IMF, October 2020 World Economic Outlook.

Note: Median across sample countries in each group. AE = advanced economy; EM = emerging market.
banks’ financial statements.6 The base model is aug- Consequences of COVID-19 for Bank Capital
mented by a satellite model that explicitly considers before Bank-Specific Mitigation
the contribution of corporate and consumer risk to The consequences of each scenario for banking
banks’ loan loss provisions and is used to estimate the systems’ future capital ratios are first simulated without
impact of government guarantees (see Box 4.1).7 adjusting for how the bank-specific mitigation policies
discussed earlier alter the recognition of provisions, cal-
6Given the lower granularity of the data, the global stress test
culation of risk-weighted assets, or flexibility in using
also relies more heavily on econometric methods than standard
supervisory stress tests and is simpler than models that would existing capital buffers.
typically be used by authorities. It is a stand-alone solvency stress test The results of the stress test show a significant
that does not consider interaction with other risks, such as liquidity decline in CET1 of the global banking system, reaching
and contagion risks or macro-feedback effects, such as between the
banking sector and the sovereign, which might amplify the impact minimum levels of 9.6 percent in the baseline scenario
of initial shocks, nor does it take into consideration spillovers across and 9.3 percent in the adverse scenario—a drop of
interconnected banking systems. Also, the exercise does not allow for 3.6 percentage points and 3.9 percentage points, respec-
behavioral responses by banks that may change their balance sheets.
The model also assumes that bank balance sheets remain static
tively, below the CET1 level in 2019. The trajectory
during the simulation period, which does not allow banks to reach of aggregate CET1 recovery also varies importantly
lower levels of capital by deleveraging (see Online Annex 4.1). across scenarios. In the baseline scenario, CET1 steadily
7The COVID-19 crisis has had a heterogenous impact across sec-
recovers after reaching a trough in 2020, but is still
tors beyond nonfinancial corporations and households. For instance,
the transportation and entertainment industries have suffered dis- 0.7 percentage points below its initial level at the end of
proportionately from the social distancing measures implemented to the simulation in 2022. In contrast, the capital position
mitigate the spread of the disease. For this reason, it would be desir- decline is much more persistent in the adverse scenario,
able to incorporate further sectoral disaggregation in the analysis, but
more granular decompositions of banks loan portfolios are typically with CET1 levels remaining 2.4 percentage points
available only for a small subset of banks. below their initial levels by 2022 (Figure 4.5, panel 1).

Figure 4.5. Bank Solvency under COVID-19 without Policy Mitigation
Banks’ capital ratios fall significantly ... ... driven by large provision costs.
1. CET1 Ratio 2. Drivers of Changes in the CET1 Ratio between 2019 and 2022
(Percent) (Percent)
16 25
Baseline Adverse Baseline
20
13.2 –5 12.5 15
14
10
5
12
0
25
10 Adverse
20
13.2 –6 15
10.7
8 10
5
6 0
19 20 21 22 19 20 21 22 19 20 21 22 Provision NFCI OCI Tax RWA
Global Advanced economies Emerging markets CET1 NII NTI Other Dividend CET1
Near fifteen percent of the global banking system will fall below 4.5% The maximum capital shortfall against a broad statutory capital
CET1 ratio. requirement could reach over $400 billion.
3. Distribution of Bank Assets by CET1 Ratio under Adverse Scenario 4. Maximum Broad Capital Shortfall under Adverse Scenario
(Percent; T = trough year) (Billions of US dollars)
80 400
60 300
40 200
20 100
0 0
Sources: Haver Analytics; SNL Financial; and IMF staff estimates.

Note: In panel 2, green and red bars denote increases and decreases in capital, respectively. AE = advanced economies, which comprise euro area, low-rate AEs,
North Atlantic, and other AEs; CET1 = common equity Tier 1; EM = emerging markets; GSIB = global systemically important bank; NFCI = net fee and commission
income; NII = net interest income; NTI = net trading income; OCI = other comprehensive income; Other = several financial accounts, including operating expenses
and non-operating items; RWA = risk-weighted assets.
The decline in the CET1 ratio over the simulation The sizes of the aggregate decline and the contribu-
horizon stems mainly from an increase in loan loss provi- tion of different components vary across regions. The
sions (Figure 4.5, panel 2). In the baseline scenario, higher maximum decline in CET1 in the baseline scenario
loan loss provision expenses contribute to a 5 percentage is much larger in advanced economies (Figure 4.5,
point decline in CET1, whereas in the adverse scenario panel 1). The situation reverses, however, in the
their contribution is 6 percentage points. This is directly adverse scenario, where advanced economies see a
related to the different trajectories of economic activity maximum decline in CET1 of about 4.0 percentage
in the two scenarios, where the rebound projected in the points, compared with 4.9 percentage points for
baseline scenario for 2021 results in lower provisioning emerging markets. This difference is a result mainly
expenses. In contrast, the increase in risk-weighted assets of higher provision costs in emerging markets due
plays only a minor role in driving the changes in CET1. to the relative economic underperformance of this

group of countries in the adverse scenario and the against two benchmarks: the regulatory minimum for
varying sensitivity of banks in these economies to CET1—corresponding to a ratio of 4.5 percent plus
macro-financial conditions. the bank-specific capital surcharge for each global
The trajectory of aggregate capital ratios masks systemically important bank—and a broad regulatory
significant heterogeneity across banks. Even at their threshold that also includes the current statutory levels
trough, and in the adverse scenario, more than half of of the capital conservation buffer and the countercycli-
the banks in the sample (by assets) have CET1 ratios cal buffer in place as of June 2020.8 The first threshold
above 10 percent—much higher than the minimum defines a “barebones capital shortfall” with respect to a
requirement of 4.5 percent. But banks accounting level of capital at which supervisory action would take
for 13 percent of assets in the sample fall below place. The second threshold defines a “broad capital
4.5 percent in the adverse scenario, with an addi- shortfall” relative to a capital ratio that includes the
tional 3 percent of assets below 6 percent (Figure 4.5, statutory buffers currently in effect.9 Banks facing a
panel 3). The weak tail of banks—defined as those shortfall relative to this broad statutory threshold have
with CET1 ratio below 4.5 percent plus their GSIB the capital space to provide credit by using remaining
buffer—amounts to 14 percent by assets. In the base- statutory buffers as envisioned by the international reg-
line scenario, the weak tail is 5 percent. ulatory framework, particularly where regulators have
In the adverse scenario, there is also heterogene- issued guidance announcements making those buffers
ity across regions and between global systemically available. However, they may feel less willing to expand
important banks and other banks. Global systemically lending activity for precautionary reasons or because of
important banks fare better than the average bank, market pressure.
in part because of their stronger initial capital ratios The two measures of capital shortfall in the adverse
resulting from their mandatory systemic buffers. scenario show important variation across groups of
However, 8 percent of these banks’ assets end the sim- banks (Figure 4.5, panel 4). At the global level, the
ulation period with capital ratios below 4.5 percent. barebones capital shortfall is about $200 billion, and
Among non–global systemically important banks, the broad capital shortfall reaches about $420 bil-
16 percent of bank assets fail to maintain a 4.5 per- lion (0.6 percent of sample banking assets). In both
cent CET1 ratio. Banks from emerging markets are cases, global systemically important banks capture
the most severely affected, with almost 40 percent of an important part of the shortfall, which is largely
total banking assets ending the simulation period with explained by the size of these institutions. The differ-
CET1 ratios below 4.5 percent. Banks from advanced ences across regions are driven by differences in the
economies fare better, although there is still a 12 per- size of their banking systems, with the level of capital
cent of banks’ assets below 4.5 percent by 2022. shortfalls being much larger for advanced economies.
Across regions and types of banks, the main dif- When considering the broad measure, the global
ference between banks that fail to meet regulatory shortfall represents 0.8 percent of the GDP of coun-
minimums and the rest of banks is the initial level tries where at least one bank has a capital shortfall.
of CET1. Banks that fall below 4.5 percent CET1 Across those countries, the average broad shortfall is
ratio plus GSIB buffer during the simulation period 1.1 percent of GDP.
are mainly distinguished by their lower initial capi-
tal levels—about 0.8 percentage point below those
that maintain their ratios above regulatory minimum 8For large US banks this includes the stressed capital ratio levels
levels. Also, banks with a high propensity to fall below recently defined by the Federal Reserve instead of the countercycli-
minimum capital standards generate meaningfully cal capital buffer and the capital conservation buffer. While many
jurisdictions have recently released the countercyclical capital buffer,
lower returns than peers that maintain adequate capital
the buffer is above zero in a few. The calculation does not include
throughout adverse conditions. the effect of “guidance” statements regarding banks’ ability to use
The importance of the weak tail of banks can also remaining statutory buffers.
9The calculation assumes that countercyclical capital buffers will
be assessed by estimating the capital shortfall, which
remain at current levels—0 percent in almost all countries—and
is the difference between simulated CET1 ratios and does not assume that this buffer will revert to a pre-pandemic or
those set by regulation. The shortfall is measured “normalized” level that is difficult to determine a priori.

Effect of Bank-Specific Policies on Capital Ratios applied on a bank-specific basis.11 These incre-
As discussed, authorities have implemented policies ments are integrated into each bank’s balance sheet
aimed at giving banks flexibility to maintain the flow positions at the end of each period.
of credit to the real economy. These policies, which
include government loan guarantees and capital ade- In quantifying the impact of these policies, it is
quacy policies, affect the need to set aside provisions assumed that they are maintained over the three-year
and the way in which capital ratios are computed and horizon of the scenario, unless an explicit expiration
should therefore also improve measured bank capital date was mentioned when the policy was announced.
ratios over the next three years. Although this assumption avoids speculating about
The mitigating impact of some of these policies can the timing of withdrawal of some of these policies, it
be quantified in the stress testing exercise as follows: may be too benign, especially in the baseline scenario,
•• Government guarantees: The impact of government in which authorities might decide to withdraw them
guarantees on banks’ provisions is captured by their as the economy recovers during the latter part of the
impact on banks’ expected losses. These losses are simulation window.
the product of banks’ exposure to firms, the proba- Bank-specific mitigation policies improve average
bility of default of those firms, and the loss expe- capital ratios across countries and scenarios. In the
rienced by banks when firms default. Government adverse scenario, the CET1 ratio for advanced econo-
guarantees can be understood as reducing the latter mies is about 110 basis points higher at the end of the
term—known as the “loss given default”—because, simulation when both government loan guarantees and
under these conditions, the guarantee would be capital adequacy policies are considered. In the sim-
executed. Because of lack of data on the extent to ulations, the improvement in capital ratios is a result
which banks originate guaranteed loans, all banks in largely of the decline in provision expenses because of
a country are assumed to benefit equally from the government loan guarantees; capital adequacy policies
guarantee in a proportion equal to the ratio of gov- explain about a third of the overall improvement in
ernment guarantees to total corporate loans. Because CET1 at the end of the simulation period in advanced
announced guarantee programs apply mostly to economies (Figure 4.6, panels 1 and 2). In the sample
new loans, this assumption likely overestimates their of emerging market economies, capital adequacy
initial impact. It is also assumed that guarantees policies do not play a meaningful role, as these policies
are used to the full extent of announced amounts are largely absent in this sample. Given the estimated
(full uptake).10 In the model, a lower uptake of impact of loan guarantees, the final uptake of these
government guarantees would lead to a proportional policies—the extent to which the announced guarantee
increase in provision expenses and therefore a pro- programs are used—could be an important driver of
portionally lower impact of the policy on loan loss the final solvency position of the banking system. As
provision expenses. discussed, an ultimate uptake of half the announced
•• Capital adequacy policies: The three categories of amount would reduce the mitigating effect of the
capital adequacy policies are quantified from the policy roughly by half.
estimated impact of each announced policy on Government loan guarantees and capital mitigation
each bank. For example, the effect of canceling policies reduce the share of bank assets with CET1
dividends is quantified from stress test model ratios below 4.5 percent in the adverse scenario from
forecasts. The release of capital buffers is estimated 13 percent without mitigation policies to 8 percent
by multiplying the percentage reduction by forecast when those policies are in place (Figure 4.6, panel 3,
risk-weighted assets. Changes to the calculation compared with Figure 4.5, panel 3). Among global
of risk-weighted assets similarly apply to the systemically important banks, these policies reduce
announced change to the relevant exposure class. the share of assets with CET1 below 4.5 percent from
In a very few instances, bank-specific policies are 8 percent to 3 percent. This decline is also important
for non–global systemically important banks, going
10Many of these programs were announced only a few months
ago, so the extent to which the guarantees will be used by banks to 11Online Annex 4.1 describes the estimation of policy mitigation
originate loans is still unclear. effects in greater detail.

Figure 4.6. Bank Solvency under COVID-19 with Policy Mitigation
Policy mitigations would cushion some of the capital depletion ... ... especially provision policies.
1. CET1 Ratio under Adverse Scenario 2. Impact on CET1 from Policy Mitigations under Adverse Scenario
(Percent) (Basis points)
16 200
No mitigation Advanced economies
With provision mitigation 150
14 With provision and capital mitigation 100
50
12
0
120
10 Emerging markets 100
80
8 60
40
20
6 0
Y19 20 21 22 Y19 20 21 22 Y19 20 21 22 Provision NFCI OCI Tax RWA
Global Advanced economies Emerging markets CET1 NII NTI Other Dividend CET1
Policy support would reduce the weak tail of banks by 5 percent ... ... and the capital shortfall by over $200 billion.
3. Distribution of Bank Assets by CET1 Ratio under Adverse Scenario 4. Maximum Broad Capital Shortfall under Adverse Scenario
(Percent; T = trough year) (Billions of US dollars)
Barebone: 4.5% + GSIB buffer, with mitigation
Broad: fully loaded, with mitigation
80 400
60 300
40 200
20 100
0 0
Source: Haver Analytics.

Note: Provision mitigation policies include guarantees only. Estimation of the impact of capital mitigation is explained in Online Annex 4.1. AE = advanced economies;
CET1 = common equity Tier 1; EM = emerging markets; GSIB = global systemically important bank; NFCI = net fee and commission income; NII = net interest
income; NTI = net trading income; OCI = other comprehensive income; Other = several financial accounts, including trading and investment income, operating
expenses, and non-operating items; RWA = risk-weighted assets.
from 16 percent to 12 percent. In advanced econ- In economies where banks with shortfalls are head-
omies, the policies analyzed shrink this segment of quartered, the broad shortfall represents about 0.4 per-
banks from 12 percent to 6 percent, and in emerging cent of their combined GDP, and, across countries, the
markets, the consideration of these policies in the average shortfall is about 0.7 percent of GDP. In terms
simulation has only a small effect on the troubled tail of the initial CET1 ratios of those banks that experi-
of banks. Overall, the weak tail of banks, whose CET1 ence a shortfall during the simulation, in the adverse
ratio fall below 4.5 percent plus GSIB buffers, declines scenario the global shortfall reaches 6.5 percent and
from 14 percent to 8.3 percent of bank assets. the average is 7.7 percent. All in all, the bank-specific
The mitigating role of bank-specific policies also policies quantified in this chapter mitigate the impact
maps into lower barebones and broad capital shortfalls of the adverse scenario on bank capital ratios, but the
(Figure 4.6, panel 4), with an especially remarkable impact is still sizable, and a share of global systemically
decline for global systemically important banks. Across important bank assets would still be part of the weak
banks, the broad capital shortfall is about $220 billion, tail of banks, even when maximizing the impact of
half of which corresponds to the barebones shortfall. these policies on capital ratios. The capital shortfall

relative to a minimum capital standard that treats insolvency, thus reducing lending procyclicality and
all guidance statements as reducing capital buffers is supporting banks’ profitability and solvency. Simi-
lower—about $110 billion, or about 0.2 percent of larly, the use of capital buffers creates lending space to
global GDP. However, reduction of capital levels to the support the real economy. Hence, these policies can
extent of these informal capital releases would likely be help bridge the impact of the COVID-19 shock and
unsustainable. reduce the chances that a transitory shock will have
Some policies that are more challenging to quantify permanent consequences for financial stability and
would also lead to an improvement in bank capital the global economy. However, if the pandemic and
ratios. Most important, several countries have provided the containment measures last longer than initially
guidance on loan classification, provisioning, and expected, ultimately affecting the solvency of bor-
disclosure, and have revised the automatic reclassifica- rowers despite the mitigating role of these policies,
tion for restructured loans. Others have gone further banks will need larger future provisions and will have
and changed the criteria for the reclassification of loans lower buffers against future shocks, including from a
or frozen those classifications. The effects of these meaningful second wave of the virus. Maintenance of
policies on loan loss provisions, in principle, are cap- generous guarantee programs over an extended period
tured through GDP effects of continued credit flow. of time could also jeopardize fiscal solvency if defaults
However, the changes in reclassification criteria for eventually materialize and could lead to further bank
credit also spare it from increased risk-asset weighting. losses related to their sovereign exposures. Further-
Because the quantity of loans that would have been more, given the unusual degree of uncertainty around
reclassified in the absence of these measures cannot be the depth and duration of the COVID-19 recession,
quantified in advance and is generally not reported, the a severely adverse scenario with stronger consequences
stress test model cannot capture the risk-weighted asset for the banking sector cannot be ruled out.
savings associated with these policies.
Overall, while the bank-specific policies quantified
in this section help improve banks’ capital ratios over Summary and Policy Discussion
the simulation period, the main contribution of the COVID-19 has had important consequences for the
broad policy packages implemented by authorities global banking sector and will pose further challenges.
likely comes from the support they provide to the Should a quick rebound in economic activity not
macroeconomy. This is because the increase in loan materialize, corporate and household solvency prob-
loss provision expenses in response to the macroeco- lems will likely deteriorate further and collateral values
nomic scenario is the main driver of the simulated may decline, resulting in greater credit losses and
decline in capital ratios, even after accounting for posing challenges for banks globally. These challenges
the bank-specific mitigation policies. A more adverse could interact with other, more structural challenges,
macroeconomic scenario, as would be the case in the such as the low profitability observed in some regions
absence of the broad support measures implemented, in an environment of persistently low interest rates and
would have likely resulted in significantly lower term spreads, a scenario that has become increasingly
capital ratios. Although counterfactual forecasts for likely in the wake of the pandemic.
the trajectory of the global economy in the absence of The simulations presented in this chapter show that,
broad support policies are not available, the import- on aggregate, the banking systems analyzed would
ant difference in simulated capital ratios between the remain solvent in coming years, although there is
baseline and adverse scenarios suggests how broad heterogeneity across and within regions. The aggregate
macroeconomic support has likely helped banks’ capi- solvency is partly due to the buffers accumulated as a
tal adequacy. result of the regulatory reforms introduced after the
The policies discussed in this section support the global financial crisis. In fact, banks analyzed in this
solvency of banks, but they also pose intertemporal chapter had a median CET1 ratio of 11.9 in 2007,
trade-offs that could become relevant in the future. compared with 16.2 percent in 2019. This improve-
Delaying provision expenses because of temporary ment in the initial solvency conditions carries over to
liquidity shocks to borrowers can help prevent borrow- the minimum CET1 ratios achieved in response to the
ers’ liquidity challenges from immediately turning into COVID-19 crisis.

Nonetheless, while aggregate capital ratios remain affect the reliability of financial statements and capital
above regulatory minimums, at a global level and ratios, and risk undermining the confidence in the
within regions there is a weak tail of banks that could banking system. Moreover, they may lead to lending to
see their solvency challenged. The size of this tail insolvent borrowers while not recognizing loan losses,
depends largely on the depth and persistence of the which may not only jeopardize the financial soundness
crisis, becoming sizable across almost all regions and of banks but also the recovery as credit is diverted from
groups of banks in an adverse scenario with a persistent productive uses.
decline in economic activity. Some global systemically Looking ahead, the benefits of these policies in
important banks are also part of this weak tail, which easing banks’ capital constraints and maintaining the
could have broader repercussions for financial stability flow of credit to the real economy should be carefully
in an adverse scenario. balanced against their potential medium-term risks
Policies adopted by governments, central banks, to financial stability. Although using the flexibility
and bank regulators have helped ease banks’ challenges embedded in the prudential framework in accordance
amid the COVID-19 crisis. Direct support to bor- with recommendations made by standard setters could
rowers (both firms and households)—and liquidity help reduce procyclicality and negative feedback loops
provision to key markets, banks, and other financial in response to temporary liquidity shocks, relaxing
intermediaries—have had a marked effect on bank loan classification and provisioning rules undermines
capital ratios through the resultant improvement in transparency and data reliability as financial statements
macroeconomic conditions. On top of this support, and prudential ratios may no longer adequately reflect
government loan guarantees and capital adequacy the true strength of banks. A decline in the quality of
policies have provided a second line of defense that information could lead to a loss of confidence in the
has eased and will likely continue to ease pressures, as banking system, with adverse implications for stability.
shown in the quantitative forward-looking analysis of It is thus important that some of these measures be
this chapter. carefully phased out as the economy recovers, especially
The majority of regulatory responses taken so far in the baseline scenario. It is also essential that, in any
are consistent with the core standards implemented scenario, banks promptly recognize losses for borrow-
after the global financial crisis and with internationally ers that become insolvent as evidence of impairment
agreed guiding principles. National authorities have becomes available. More broadly, phasing out govern-
taken capital and liquidity measures using the flexibility ment support, including government guarantees, too
embedded in the prudential framework to help support quickly would lead to lasting damage to the economy,
lending to the real economy. Authorities have clarified but phasing it out too late could risk damaging public
the usability of capital and liquidity buffers, encouraged finances or unduly keeping insolvent borrowers afloat.
banks to use these buffers to absorb losses and sustain Despite the mitigating effect of government policies,
credit, and restricted capital distributions to preserve in the adverse scenario simulated in this chapter,
capital. However, in several cases, regulatory easing was there is a weak tail of banks that fail (or nearly fail) to
achieved by lowering minimum requirements below meet minimum regulatory requirements. This finding
Basel framework levels. Such deviations risk undermin- highlights the usefulness of forward-looking stress
ing the credibility of the internationally agreed stan- tests to assess the health of banking systems and to
dards, could contribute to market segmentation, and guide prospective policy responses to the current crisis.
may increase the risks to bank safety and soundness. When conducted by regulators or supervisors, this type
Standard setting bodies (like the Basel Committee) of assessment would rely on more granular data than
and national authorities have also encouraged banks to used in this global exercise, and thus would provide
work constructively and prudently with borrowers and additional richness.
have issued guidance on how to treat restructured loans Once the assessment is done, however, what should
and public and private moratoria for prudential asset authorities do about banks that could become trou-
classification and provision. Nonetheless, some mea- bled? The answer to this question should take into
sures that run contrary to these recommendations have consideration country-specific circumstances. Acting
been observed, such as the freezing of asset classification now to strengthen the financial safety net, including
status and provisioning requirements. These measures deposit guarantee programs, resolution regimes, and

central bank liquidity facilities, is key. Capital preserva- plans to rebuild capital gradually for the most vulner-
tion measures will help, including temporarily limiting able entities to ensure confidence, avoid procyclicality,
the distribution of dividends, as some countries have and preserve financial stability.12 Preparing contingency
already done. For countries that allowed banks to draw plans that detail how the authorities will respond to
down capital buffers, the stress test results will help possible future pressures is critical to support effective
guide the timing and pace at which these exceptional policy responses if the adverse scenario materializes.
measures can be unwound. Supervisors could use this
information to reassess forward-looking capital plans 12For a broader discussion of the banking regulatory and supervi-
and take measures aimed at preserving and supporting sory actions to deal with COVID-19, see IMF (2020a).

Box 4.1. The Role of Corporate and Consumer Risk in the Evolution of Banks’ Loan Loss Provisions
The COVID-19 crisis is likely to impact the credit A forward-looking simulation of the evolution of
risk of both firms and households. Households and loan loss provisions (as a share of total loans) in the
firms may have different effects on bank provisioning baseline scenario of the World Economic Outlook and
and capital, according to the severity of the shock and the share of them explained by corporate and con-
the composition of the lending portfolios. Disentan- sumer risk shows that the crisis generates a strong but
gling the impact of these two sources of credit risk is gradual response that peaks during the first half of
important to evaluate the policy response to the crisis 2021 (Figure 4.1.1). At its peak, the increase in the
as both the magnitude and type of support measures loan loss provision ratio is about 1 percentage point in
differ across these two sectors. advanced economies and about 0.4 percentage point
A satellite model of loan loss provisions that considers in emerging market economies.
the mix of bank loans across corporate (firms) and con- Most of the increase is due to heightened corporate
sumer (households) loans was developed to complement risk, although households play a significant role in
the core global stress test model. This model relies on the advanced economies because of their larger share on
local projection method to decompose bank loan loss advanced economy banks’ portfolios. These results show
provisions into a component related to household risk that the level and composition of total provisions depends
(captured by the unemployment rate or changes in house on the mix of bank loan portfolios and on the relative
prices) and another related to corporate loans risk (cap- size of the shocks to firms and households. The analysis
tured by a measure of the probability of default of the highlights the importance of considering the loan mix
corporate sector). It provides a starting point for a more for the assessment of the impact of the crisis and the
nuanced discussion of the implications of bank business analysis of policy responses. In the chapter, these insights
models for future financial performance and for tackling are carried to the global stress testing model to assess the
the impact of mitigation policies that target specific sec- impact of policies that affect a specific sector, such as the
tors (see Online Annex 4.1 for additional details). government loan guarantees that tend to be focused on
corporate loans. If data were available, this type of analy-
This box has been prepared by Nicola Pierri and sis could also be used to further disaggregate the impact
Tomohiro Tsuruga. of the crisis on different productive sectors.
Figure 4.1.1. Additional Quarterly Provisioning

(As share of loans)
1. Advanced Economies 2. Emerging Markets and Developing Economies

(Percentage points; share of total loans) (Percentage points; share of total loans)
1.2 0.5
Household Household
Corporate Corporate
1.0
0.4
0.8
0.3
0.6
0.2
0.4
0.1
0.2
0.0 0.0
2020:Q1 20:Q3 21:Q1 21:Q3 22:Q1 22:Q3 2020:Q1 20:Q3 21:Q1 21:Q3 22:Q1 22:Q3
Sources: Fitch Connect; S&P Global Market Intelligence; and IMF staff estimates.

References
International Monetary Fund (IMF). 2020a. “Banking Sector ———. 2020b. “Fiscal Monitor Database of Country Fiscal
Regulatory and Supervisory Response to Deal with Coro- Measures in Response to the COVID-19 Pandemic.” https://
navirus Impact.” Special Series on COVID-19. https://www www.imf.org/en/Topics/imf-and-covid19/Fiscal-Policies
.imf.org/~/media/Files/Publications/covid19-special-notes/ -Database-in-Response-to-COVID-19.
enspecial-series-on-covid19banking-sector-regulatory-and
-supervisory-response-to-deal-with-coronavir.ashx

5
CHAPTER
CORPORATE SUSTAINABILITY
FIRMS’ ENVIRONMENTAL PERFORMANCE AND THE COVID-19 CRISIS
•• Tighter financial constraints and weaker economic conditions can act as a drag on firms’ environmental
performance.
•• The coronavirus disease (COVID-19) crisis could substantially reduce firms’ green investments, reversing
gains in their environmental performance made in past years.
•• Climate policies and green investment packages are therefore warranted to support a green recovery and
the transition to a low-carbon economy.
•• Policies aimed at fostering sustainable finance such as better disclosure standards and product standardiza-
tion could further help mobilize green investments and alleviate firms’ financial constraints.
The shutdown in economic activity as a result of the Introduction

COVID-19 crisis has resulted in a temporary decline
The shutdown in economic activity as a result of the
in global carbon emissions, but the long-term impact of
COVID-19 crisis resulted in a sharp decline in global
the pandemic on the transition to a low-carbon econ-
carbon emissions (Figure 5.1, panel 1).1 Daily emis-
omy remains uncertain. While the economic fallout from
sions in early April 2020 fell by about 17 percent com-
the crisis may constrain firms’ ability to invest in green
pared with 2019 levels, though most of this decline has
projects, thus slowing down the transition, the COVID-
reversed since then as economic activity has picked up
19 crisis could also induce a structural shift in consumer
across countries. Such a reversal in emissions is in line
and investor preferences toward environmentally friendly
with what turned out to be only a temporary decline
products, providing an opportunity to introduce mitigation
in the price of carbon emission allowances in March
policies that help diversify away from fossil fuel produc-
2020 (Figure 5.1, panel 2). Overall, recent studies
tion. Looking back at previous episodes of financial and
forecast a temporary reduction in emissions of about 4
economic stress, this chapter finds that tighter financial
to 7 percent in 2020, far from the large and sustained
constraints and adverse economic conditions are generally
decrease in emissions required under the Paris Agree-
detrimental to firms’ environmental performance, reduc-
ment to limit the increase in global temperature to well
ing green investments, and setting back their progress
below 2°C (Le Quéré and others 2020).2
by several years. This suggests that the COVID-19 crisis
There is also a possibility that the transition to a
could potentially slow down the transition to a low-carbon
low-carbon economy could be delayed should the
economy. In light of the urgent need to reduce global
economic scarring from the pandemic crisis run
greenhouse gas emissions, it also underlines the importance
deep, inducing economic agents and policymakers
of climate policies and green investment packages to support
to sideline or postpone environmental objectives.
a green recovery and the energy transition. Policies aimed
Heightened economic uncertainty, a sharp drop in
at fostering sustainable finance, such as improved trans-
energy prices, and corporate balance sheet vulner-
parency and standardization, could further help mobilize
abilities may result in a reduction in investments
green investments and alleviate firms’ financial constraints.
and research in long-horizon, capital-intensive green
The authors of this chapter are Zhi Ken Gan, Pierpaolo Grippa,
Pierre Guérin, Oksana Khadarina, Samuel Mann, Felix Suntheim 1In the short term, there is an almost one-to-one relationship
(team lead), and Yizhi Xu, with contributions from Alan Feng, between economic growth and emissions (Hale and Leduc 2020).
Germán Villegas Bauer, and Julia Xueliang Wang, under the guidance 2The UN Environment Programme (2019) estimates that emis-
of Fabio Natalucci, Mahvash Qureshi, and Jérôme Vandenbussche. sions need to decline by 2.7 percent annually in order to reach the
Harrison Hong served as an expert advisor. 2°C goal by 2030.

Figure 5.1. The Energy Transition during the COVID-19 Crisis
Carbon emissions declined rapidly as COVID-19 became a global ... but, unlike during the global financial crisis, the decline has been
pandemic ... short-lived, with a rebound in emissions.
1. Change in Daily CO2 Emissions in 2020 Compared with 2019 Mean 2. Price of European Union Emissions Trading System CO2 Emission
Daily Emissions Allowance
(Percent) (Euros per metric ton)
0 35
Global financial crisis COVID-19
–2
30
–4
–6 25
–8
20
–10
15
–12
–14 10
–16
5
–18
–20 0
Jan. 2020 Feb. Mar. Apr. May June 2008 10 12 14 16 18 20
Sources: Global Carbon Project; Refinitiv Datastream; and IMF staff calculations.
Note: Panel 1 shows the reduction in daily CO2 emissions in 2020 compared with 2019 mean levels. Panel 2 shows the price of futures contracts on carbon emission
allowances traded on the Intercontinental Exchange. The European Union Emissions Trading System was subject to several changes in regulation over the sample
period that may have affected the price level.
projects. In addition, subsidies or economic rescue implementation of green policy measures such as
packages aimed at softening the impact of the crisis carbon taxes.4
may slow the transition—for example, by supporting Against this backdrop, this chapter aims to address the
firms or activities not compatible with long-term following two key questions: (1) How has the COVID-19
climate mitigation goals. crisis affected green financing so far? (2) What can be
At the same time, the current crisis could also learned from past economic crises about the likely behav-
present an opportunity to accelerate the transition to ior of the corporate sector in the near and medium terms
a low-carbon economy by inducing structural shifts with respect to the greening of the economy?
in consumer and investor preferences toward envi-
ronmentally friendly products in the event economic
agents change their beliefs about the likelihood of The COVID-19 Crisis and Financing the
other catastrophic events, such as those linked to Energy Transition
climate change.3 In the corporate sector, for example, The COVID-19 crisis has not led to a sustained
climate change has become an increasingly important decline in green financing so far. Issuance of green
topic since the onset of the pandemic, as is evident corporate bonds, which has trended up over the past
from firms’ earnings calls transcripts (see Box 5.1). decade, declined in March 2020 in the midst of the
More generally, an increased awareness of the bene- financial market turmoil, but it has picked up since,
fits of long-term disaster prevention could facilitate
4Calls for implementing “green recovery” packages in the after-
math of the COVID-19 crisis have come from different quarters,

including the private sector in some cases. For example, in June
3Survey evidence suggests that voters have become more worried 2020 more than 100 global investors called for a green European
about other global threats, such as climate change, after experiencing Union recovery plan. The EU coronavirus recovery package earmarks
the COVID-19 pandemic (Geman 2020). about 37 percent of the funds for climate protection.

CHAPTER 5 Corporate Sustainability: Firms’ Environmental Performance and the COVID-19 Crisis
Figure 5.2. The COVID-19 Crisis and Green Investments
Green bond issuance dropped in the first quarter of 2020 before Bank lending has shifted to green firms over the past decade.
picking up again beginning in April 2020.
1. Green Corporate Bond to Total Corporate Bond Issuance and 2. Total Amount of Syndicated Loans to Firms with Environmental Scores
Total Green Corporate Bond Issuance, January 2014–June 2020 Higher than Median and Firms with Environmental Scores Lower than
Median, 2009:Q1–2020:Q1
Total issuance (billions of US dollars, right scale) (Billions of US dollars)
3.0 Ratio of green issuance to total issuance (percent, left scale) 30 400
Firms with high environmental performance
Firms with poor environmental performance 350
2.5 25
300
2.0 20
250
1.5 15 200
150
1.0 10
100
0.5 5
50
0 0 0
2009 10 11 12 13 14 15 16 17 18 19 20
Jan. 2014
June 14
Nov. 14
Apr. 15
Sep. 15
Feb. 16
July 16
Dec. 16
May 17
Oct. 17
Mar. 18
Aug. 18
Jan. 19
June 19
Nov. 19
Apr. 20
Flows into sustainable and environmental equity funds slowed in the Equity indices with a focus on environmental issues performed at least
first quarter of 2020 but remained positive. as well as the overall market.
3. Sustainable and Environmental Fund Flows as a Share of Fund Size, 4. Cumulative Returns of Green and Conventional Equity Market Indices
2003:Q1–2020:Q1 (Percent)
(Moving averages; percent)
MSCI ACWI
10 140
Environmental fixed income funds MSCI ACWI excluding fossil fuels USD price index
8 Environmental equity funds MSCI ACWI low carbon target USD price index
Sustainable fixed income funds MSCI Global Environment Index
6 120
Sustainable equity funds
4
2 100
0
–2 80
–4
–6 60
2003 05 07 09 11 13 15 17 19 Jan. 2020 Feb. Mar. Apr. May June July
Sources: Bloomberg Finance L.P.; Dealogic; Morningstar; Refinitiv Datastream; and IMF staff calculations.
Note: Panel 1 shows global green corporate bond issues. Panel 3 shows quarterly flows into sustainable or environmental fixed-income or equity funds.
MSCI ACWI = Morgan Stanley Capital International All Country World Index.
with the share of green bonds in total corporate bond types of firms dropped slightly in the first quarter of
issuance returning to 2019 levels (Figure 5.2, panel 1). 2020 (Figure 5.2, panel 2).
In the syndicated loan market, loans to firms with an Investment funds with a focus on sustainable or
above-median score in environmental performance have environmental investments have continued to attract
increased over the past decade compared with loans investors throughout the crisis, especially fixed-income
to firms with a below-median score.5 Lending to both funds, with only a small drop in aggregate inflows in
5Firm-level environmental, social, and governance data come
with several caveats. First, the data cover only publicly listed firms, may capture different features of environmental performance. Third,
so the results do not necessarily carry over to the entire economy, as some scores are self-reported by firms, accuracy may vary across
which includes unlisted small- and medium-sized enterprises. the sample. See Online Annex 5.1 for a description of the variables
Second, there is a lack of standardization and transparency across used in this chapter. All annexes are available at www.imf.org/en/
data providers, so environmental scores from different providers Publications/GFSR.

some asset classes (Figure 5.2, panel 3).6 A possible constraints is likely to reduce their ability to invest in
driver of the good performance of sustainable and green projects and cut greenhouse gas emissions.
environmental funds may have been the relatively high Extending this analysis to a global sample and
returns that green investments have experienced during specifically analyzing firms’ environmental performance
this crisis in general (Figure 5.2, panel 4). shows that tighter financial constraints are indeed
Overall, the impact of the COVID-19 crisis on the associated with worse environmental performance
financing of green investments so far seems to have (Figure 5.3, panel 1). Proxying firms’ financial con-
been modest and short-lived. However, given the sever- straints by firm size (logarithm of total assets), rating
ity and possible persistence of the shock—in terms of status, interest coverage ratio, ability to pay dividends,
output decline, the extent of potential scarring, and and the commonly used Kaplan-Zingales index, the
the heightened economic uncertainty—there could environmental performance of financially constrained
be significant strains on corporate balance sheets. It is firms is in each case significantly weaker than that
therefore challenging to forecast whether such trends of unconstrained firms. Specifically, environmental
will continue and ultimately what the overall impact of performance falls by 10 points when firm size drops
the crisis will be on firms’ environmental performance from the median to the 25th percentile of the firm
and on their ability to contribute to global climate size distribution. When a firm does not pay dividends
change mitigation efforts. In view of this concern, the or when it is not rated, its environmental score is 4
analysis in the next section examines firms’ environ- points and 3 points lower, respectively, than the score
mental performance during previous episodes of finan- of dividend-paying and rated firms. The environmental
cial and economic stress to draw possible implications score is 1 point lower when an aggregate measure of
for the current episode. financial constraints (the Kaplan-Zingales index) is
above the median of the sample distribution. Similar
results are obtained when considering firms’ carbon
Lessons from Past Economic Crises for Firms’ intensity instead of their environmental performance.
Environmental Performance during the A key channel through which financial constraints
COVID-19 Crisis can affect firms’ environmental performance is a
Existing research focusing on the United States decline in investments in green technologies. Con-
suggests that the environmental, social, and governance strained firms may postpone or reduce such invest-
(ESG) performance of financially constrained firms— ments if they do not directly contribute to revenue
that is, firms that face difficulties in raising external generation. Moreover, financially constrained firms
capital—is generally weaker relative to unconstrained may face difficulties in borrowing against future profits
firms (Hong, Kubik, and Scheinkman 2012).7 There- to invest in research and development, consequently
fore, a deterioration in financial or economic condi- postponing investments in intangibles that could
tions that results in a tightening of firms’ financial potentially improve their environmental performance.
Regression analyses support these hypotheses and
suggest that financially constrained firms are less likely
6Sustainable funds explicitly indicate all kinds of sustainabil-
to make investments that reduce future environmental
ity; impact; and environmental, social, and governance (ESG) risks, such as treatment of emissions or installation of
strategies in their prospectus. Environmental funds invest in cleaner technologies (Figure 5.3, panel 2). For exam-
environmentally oriented industries. See the October 2019 Global ple, the probability that a firm will make an environ-
Financial Stability Report for a discussion of sustainable finance
and financial stability.
mental investment falls by 6 percentage points when
7Because financial constraints are not directly observable, different firm size drops from the median to the 25th percentile
proxies are used in the literature (see Online Annex 5.2): firm size of the firm size distribution.
(large firms are expected to be less financially constrained than
These results have important implications
small firms), rating status (firms with a rating may have easier access
to capital markets than those without), the interest coverage ratio in the current COVID-19 context. An adverse
(defined as earnings before interest and taxes divided by interest macro-financial shock that increases uncertainty and
expenses, reflecting a firm’s debt repayment capacity with higher amplifies firms’ financial constraints is likely to affect
values indicating less financially constrained firms), the ability to pay
dividends, and the Kaplan-Zingales index (an aggregate measure of firms’ environmental performance and has the poten-
financial constraints). tial to significantly impede their ability to invest in

Figure 5.3. Financial Constraints, Financial Stress, and Environmental Performance
Financially constrained firms have weaker environmental ... and are less likely to make environmental investments.
performance ...
1. Effects of Financial Constraints on Environmental Score 2. Marginal Effects on the Probability of a Firm Making Environmental
(Index) Investments
(Percent)
0 0
–2 –1
–2
–4
–3
–6
–4
–8
–5
–10 –6
–12 –7
Dividends ICR Ratings Size KZ score Dividends ICR Ratings Size KZ score
Severe financial stress leads to poorer corporate environmental ... and the effects of financial stress are stronger for financially
performance ... constrained firms.
3. Response of Environmental Score to a VIX Shock 4. Coefficient of the Interaction Term between Firm-Level Financial
(Index) Constraints and a VIX Shock
–3.0 0
–3.5 –0.02
–0.04
–4.0
–0.06
–4.5
–0.08
–5.0 –0.10
–5.5 –0.12
0 1 2 3 Dividends ICR Ratings Size KZ score
Years
Sources: Refinitiv Datastream; Standard & Poor’s; and IMF staff calculations.
Note: “Dividends” refers to firms that do not pay dividends, “ICR” to firms with earnings below interest expenses, “Ratings” to firms that do not have a rating from
Standard & Poor’s, “Size” to the log of total assets (the sign of this variable is reversed so that higher values indicate smaller firms), and “KZ score” to firms above
the median of the Kaplan-Zingales index score distribution (more financially constrained firms have higher KZ scores). Panel 1 shows regression estimates of
environmental scores on financial constraints. Regressions include firm-level controls as well as industry, country, and time fixed effects. Firm-level controls are the
log of total assets and earnings, except when using “Size” as a measure of financial constraint, when only earnings are used as a firm-level control. Panel 2 shows
the marginal effects of a given financial constraint measure on the probability of a firm making an environmental investment. The probit models include the same
control variables and fixed effects as in panel 1. In panel 3, t = 0 is the year of the shock. The Chicago Board Options Exchange Volatility Index (VIX) shock is the
average value of the VIX over the calendar year. The solid line denotes the response to a 16.3 point increase in the VIX (corresponding to the difference in the average
value of the VIX in 2020, using data up to July 31, 2020, relative to the average value in 2019). The dashed lines denote 90 percent confidence intervals. Responses
are obtained with the local projection approach from firm-level panel regressions that include firm-level controls, country-specific output gaps, the price of oil, and
country and industry fixed effects. Panel 4 shows interaction terms at a one-step horizon between the VIX shocks and the lagged firm-level financial constraint
variables. The same control variables as in panel 3 are used. In panels 1, 2, and 4, solid bars indicate significance at the 10 percent level. ICR = interest coverage ratio.
green projects. To quantify the extent of the impact, The analysis shows that a sudden jump in the VIX,
two types of shocks are analyzed here: (1) a global comparable to the average level that prevailed in the
financial stress shock (proxied by the Chicago Board first half of 2020 during the COVID-19 pandemic,
Options Exchange Volatility Index [VIX]) and (2) a would lead to a persistent drop in firms’ environmental
real economic activity shock capturing a sudden drop performance by up to 5 points, with the pre-shock
in domestic output.8 performance level not attained for at least three years
after the shock (Figure 5.3, panel 3). Absent policy
8See Online Annex 5.3. actions and behavioral changes, this would imply that

Figure 5.4. Economic Shocks and Environmental Performance
Contractionary economic shocks lead to lower corporate environmental ... and carbon intensity deteriorates following contractionary economic
performance ... shocks.
1. Response of Environmental Score (y-axis) over Time (x-axis) to a 2. Response of the Logarithm of Total CO2 Emissions Relative to Revenues
Fall in the Output Gap (y-axis) over Time (x-axis) to a Fall in the Output Gap
(Index) (Percent)
–0.5 20
–1.0
15
–1.5
10
–2.0
5
–2.5
0
–3.0
–5
–3.5
–4.0 –10
0 1 2 3 0 1 2 3
Years Years
Sources: Refinitiv Datastream; and IMF staff calculations.

Note: In panels 1 and 2, the real economic activity shock is scaled as a 10 percentage point drop in the output gap. The regression includes firm-level controls (log of
total assets, earnings, and a dividend dummy variable), the price of oil (log West Texas Intermediate), the Chicago Board Options Exchange Volatility Index, and
country and sector fixed effects. Dashed lines represent 90 percent confidence interval.
average corporate environmental performance would decline in firms’ environmental performance in the
return to the levels that were last observed in 2006. medium term (Figure 5.4, panel 1).10 Similarly,
Moreover, the adverse effect of global financial shocks firms’ carbon intensity—captured by their total car-
on environmental performance is magnified when bon emissions relative to revenue—could increase by
firms are financially constrained (Figure 5.3, panel 4). up to 8.5 percent in the medium term after such a
For example, for firms with an interest coverage ratio decline in the output gap (Figure 5.4, panel 2), even
below 1 or for unrated firms in 2019, the global finan- though the initial response of carbon intensity to
cial stress shock observed thus far in 2020 is estimated economic shocks may be small because of the cycli-
to lower environmental performance by 2 additional cal dynamics of carbon dioxide emissions observed
points, compared to firms with an interest coverage during recessions (Figure 5.1, panel 1; Hale and
ratio above 1 or rated firms.9 Leduc 2020).
A large decline in the output gap (10 percentage In addition to direct global financial and eco-
points, about 50 percent larger than that observed nomic shocks, changes in oil prices could also impact
in the Group of Seven [G7] economies during the corporate environmental performance by affecting
global financial crisis), would lead to a 3 point
10Other more global measures of economic activity shocks
such as the forecast error for the current-year global GDP growth
9These economic effects are calculated by multiplying the interac- relative to the World Economic Outlook projection, or the global
tion term by a 16.3 point increase in the VIX (corresponding to the economic activity shock from Baumeister and Hamilton (2019)
difference in the average value of the VIX in 2020, using data up to also lead to a fall in corporate environmental performance in the
July 31, 2020, relative to the average value in 2019). medium term.

firms’ incentives and their financial constraints. The Figure 5.5. Oil Market Shocks and Environmental Performance
onset of the COVID-19 crisis was accompanied by
Lower oil prices due to demand factors are associated with lower
a steep decline in the international price of oil.11 corporate environmental performance.
The effect of such a decline in oil prices on firms’ Response of Environmental Scores to Oil Market Shocks that Lower the
environmental performance is, however, ambiguous. Real Price of Oil across all Industries
(Index)
On the one hand, it may relax firms’ financial con- 4
straints and reduce the incentives for businesses to
improve their energy efficiency and shift away from 3
fossil fuels, including by hindering the development
of clean energy sources by making investments in new 2
projects less profitable.12 On the other hand, low oil

1
prices could benefit the energy transition by hurting
the profitability of the oil sector and leading to lower 0
investments in the fossil fuel sector and a decline in
production, thereby making it easier for clean energy –1
firms to compete.
–2
In principle, the effect of an oil price shock on
environmental performance is likely to depend on
–3
the underlying source of the shock—that is, whether
it is a demand- or supply-driven shock. A negative –4
global demand shock associated with a decline in Oil consumption demand shock Oil supply shock
economic activity that reduces the demand for oil
Sources: Refinitiv Datastream; and IMF staff calculations.
could be associated with lower corporate environ- Note: The oil market shocks are obtained from Baumeister and Hamilton (2019).
mental performance as investments into cleaner All shocks are unit shocks that lead to a fall in the real price of oil. Responses at a
two-year horizon are represented. Controls in the regression are the log of total
energy sources are delayed because of already tight assets, earnings, a dividend dummy variable, country-specific output gaps, the
financial conditions for firms. Conversely, a drop in Chicago Board Options Exchange Volatility Index, and the price of oil (log West
Texas Intermediate). The regressions include country and sector fixed effects.
oil prices due to an oil supply shock could trigger an Solid bars indicate significance at the 10 percent level.
increase in global economic activity (Baumeister and
Hamilton 2019), easing firms’ financial constraints
and allowing them to improve their environmental demand-driven shock, firms’ environmental perfor-
performance. mance is thus likely to suffer.13
Econometric analysis suggests that the source of the Overall, these results indicate that tighter financial
oil price fluctuation is indeed key to understanding constraints are associated with weaker corporate environ-
firms’ environmental response to a shock. Histori- mental performance. Adverse global financial and output
cally, when oil prices have fallen due to demand-side shocks that increase uncertainty and amplify firms’ finan-
factors, environmental corporate performance has been cial constraints weigh significantly on their environmental
weaker. By contrast, when oil prices have declined due performance. Furthermore, a reduction in oil prices
to an oil supply shock, environmental performance against the backdrop of a decline in global economic
of firms has improved (Figure 5.5). To the extent that activity is unlikely in itself to lift corporate environmen-
the COVID-19-induced oil price shock is largely a tal performance. Thus, absent strong supportive policy
actions, tighter financial constraints and weaker economic
activity related to the COVID-19 crisis are likely to act as
11Global energy demand declined by 3.8 percent in the first a drag on firms’ environmental performance in the future.
quarter of 2020. The demand for oil, coal, and to a lesser extent gas
and nuclear energy is projected to decline substantially by the end of
2020 (IEA 2020). 13Difficulties to reach an agreement among the OPEC+ coalition
12Acemoglu and others (2019) discuss the long-term effects of the also contributed to the collapse in oil prices in early 2020, but a
shale gas boom, which reduces carbon dioxide emissions from coal in decomposition of the oil price shock in March and April 2020
the short term, while increasing aggregate production and directing suggests that it was largely driven by demand-side factors. See
energy innovation to shift away from clean energy to fossil fuels. Online Annex 5.3.

Conclusions and Policy Recommendations investments, in combination with steadily rising car-
The COVID-19 crisis has resulted in a temporary bon prices, is critical (October 2020 World Economic
decline in global carbon emissions, but its long-term Outlook; October 2019 Fiscal Monitor). Public policies
impact is uncertain. On the one hand, the crisis may and green recovery packages are important to offset
increase awareness of catastrophic risks and bring about the potential deterioration in firms’ environmental
a major shift in consumer preferences, corporate actions, performance resulting from the crisis (see the October
and investor behavior. On the other hand, the historical 2020 Fiscal Monitor).
evidence presented in this chapter suggests that there is a In addition, to alleviate firms’ financial constraints
real possibility that, barring public interventions, invest- and to aid green investment, it will be key to put in
ment by firms to improve their environmental perfor- place policies that support the sustainable finance sec-
mance may decline in this time of macro-financial stress. tor, such as better disclosure standards, development of
To achieve the reduction in emissions needed to green taxonomies, and product standardization (see the
keep global warming below 2°C, an increase in green October 2019 Global Financial Stability Report).

Box 5.1. Climate Index Based on Firms’ Earnings Calls

To measure how firms’ exposure to and awareness or “emissions.” A sharp increase in discussions
of climate change have evolved over time, a firm-level involving climate change topics is observed in 2020,
climate index was constructed for this chapter based coinciding with the COVID-19 pandemic. This
on quarterly earnings call transcripts using a climate could, for example, be the result of the COVID-19
change dictionary built from four climate change glos- crisis increasing firms’ focus on catastrophic events
saries.1 To construct the index, earnings call transcripts and long-term risks.
from 4,109 firms located in 46 countries are used. The climate change discussion index is then con-
Panel 1 of Figure 5.1.1 shows the share of earnings structed for each firm by assigning a value of 1 to
call transcripts that mention specific phrases related each earnings call transcripts that contains a phrase
to climate change, such as “climate change,” “CO2,” included in the dictionary. Panel 2 shows the aver-
age of the index over time. It is noteworthy that in
This box was prepared by Alan Feng and Germán Villegas Bauer. the earnings calls of energy sector firms, mentions of
1Following a similar approach as Engle and others (2020), the
climate-change-related terms spiked after the Paris
glossaries are obtained from the British Broadcasting Corpo-
Agreement in 2016, highlighting the importance of
ration, the Intergovernmental Panel on Climate Change, the
United Nations, and the US Environmental Protection Agency. policy risk for this sector. The increase in discussions
See Online Annex 5.4 for a list of all terms. All annexes are involving climate change over the past few years is
available at www.imf.org/en/Publications/GFSR. consistent across countries (Online Annex 5.4).
Figure 5.1.1. Climate Index
Climate change discussions have increased during After the Paris Agreement, firms in sectors exposed
the COVID-19 crisis. to transition risk became more aware of climate
risks—or opportunities.
1. Annual Share of Earnings Call Transcripts Containing 2. Quarterly Share of Firms with Climate Discussions,
Specific Climate-Change-Risk-Related Terms All Sectors and Energy Sector
(Percent) (Percent)
6 60 25
Climate change Paris Energy sector Paris
CO2 Agreement All sectors Agreement
Greenhouse gas (right scale)
5 Renewable energy 50
20
Energy efficiency
Transcripts containing the term
Environmental impact
4 40
15
3 30
10
2 20
5
1 10
0 0 0
2004
05
06
07
08
09
10
11
12
13
14
15
16
17
18
19
20
2004
05
06
07
08
09
10
11
12
13
14
15
16
17
18
19
20
Sources: FactSet; and IMF staff calculations.

References Hale, Galina, and Sylvain Leduc. 2020. “COVID-19 and CO2.”
Acemoglu, Daron, David Hemous, Lint Barrage, and Philippe FRBSF Economic Letter, July 6, Federal Reserve Bank of
Aghion. 2019. “Climate Change, Directed Innovation, and San Francisco.
Energy Transition: The Long-Run Consequences of the Shale Hong, Harrison, Jeffrey D. Kubik, and José A. Scheinkman.
Gas Revolution.” 2019 Meeting Papers 1302, Society for 2012. “Financial Constraints on Corporate Goodness.”
Economic Dynamics. NBER Working Paper 18476, National Bureau of Economic
Baumeister, Christiane, and James D. Hamilton. 2019. “Struc- Research, Cambridge, MA.
tural Interpretation of Vector Autoregressions with Incomplete International Energy Agency. 2020. “Global Energy Review
Identification: Revisiting the Role of Oil Supply and Demand 2020.” https://www.iea.org/reports/global-energy-review-2020
Shocks.” American Economic Review 109 (5): 1873–910. Le Quéré, Corinne, Robert B. Jackson, Matthew W. Jones,
Engle, Robert F., Stefano Giglio, Bryan Kelly, Heebum Lee, Adam J. P. Smith, Sam Abernethy, Robbie M. Andrew,
and Johannes Stroebel. 2020. “Hedging Climate Change Anthony J. De-Gol, David R. Willis, Yuli Shan, Josep G.
News.” Review of Financial Studies 33 (3): 1184–216. Canadell, Pierre Friedlingstein, Felix Creutzig, and Glen P.
Geman, Ben. 2020. “Survey: Florida Voters Link Climate and Peters. 2020. “Temporary Reduction in Daily Global CO2
COVID-19.” Axios (June 30). https://www.axios.com/florida Emissions during the COVID-19 Forced Confinement.”
-voters-climate-change-coronavirus-0e7182d8-81ae-45bf-a7d1 Nature Climate Change 10: 647–53.
-a910a094725f.html UN Environment Programme. 2019. Emissions Gap Report. Nairobi.

IMF Special Series on COVID-19
The IMF has responded to the COVID-19 crisis by quickly deploying financial assistance,
developing policy advice, and creating special tools to assist member countries.
The Special Notes Series (IMF.org/COVID19notes) features the latest analysis and research
from IMF staff in response to the pandemic. Below are four recent Notes from the dozens
published to date.
Banking Sector Regulatory Unconventional Monetary

and Supervisory Response Policy in Emerging Market and
to Deal with Coronavirus Impact Developing Economies
(with Q and A) David Hofman and Gunes Kamber
Rachid Awad, Caio Ferreira, Ellen Gaston, and Luc Riedweg This note discusses the use of unconventional
This note discusses the challenges that the monetary policies in emerging market and de-
COVID-19 pandemic poses for the banking veloping economies with a focus on two ob-
sector and possible regulatory and superviso- jectives: (i) increasing monetary policy space
ry responses that can maintain the balance be- to help central banks meet their output and
tween preserving financial stability, maintain- inflation goals; and (ii) mitigating limitations to
ing banking system soundness, and sustaining monetary transmission that may hamper the
economic activity. provision of credit where it is most needed.
Considerations for Designing Monetary and Financial Policy

Temporary Liquidity Support Responses for Emerging Market
to Businesses and Developing Economies
Phakawa Jeasakul Thomas Harjes, David Hofman, Erlend Nier, and
This note discusses key considerations for Thorvardur Olafsson
designing temporary liquidity support to This note provides an overview of appropriate
otherwise viable businesses to allow them central bank policy responses to the severe
to continue operations during the COVID-19 economic and financial impact of the
pandemic. COVID-19 pandemic in emerging market and
developing economies. It covers monetary,
exchange rate, and macroprudential policies,
as well as capital flow measures.
The views expressed in these notes are those of the author(s) and do not necessarily represent the views of the IMF,
its Executive Board, or IMF management.
COVID-19 Policy Tracker

This periodically updated policy tracker summarizes the key economic responses
196 governments are taking to limit the human and economic impact of the pandemic.
IMF.org/COVID19policytracker
I N T E R N A T I O N A L M O N E T A R Y F U N D
IN THIS ISSUE:
CHAPTER 1
Global Financial Stability Overview:
Bridge to Recovery
CHAPTER 2
Emerging and Frontier Markets:
A Greater Set of Policy Options
to Restore Stability
CHAPTER 3
Corporate Funding:
Liquidity Strains Cushioned
by a Powerful Set of Policies
CHAPTER 4
Bank Capital:
COVID-19 Challenges and Policy
Responses
CHAPTER 5
Corporate Sustainability:
Firms’ Environmental Performance
and the COVID-19 Crisis
GLOBAL FINANCIAL STABILITY REPORT OCTOBER 2020

Text

Uploaded by

Copyright:

Available Formats

Text

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Text

Uploaded by

Copyright:

Available Formats

INTERNATIONAL MONETARY FUND

IMF CSF Creative Solutions Division

Names: International Monetary Fund.

ISBN 978-1-51355-422-8 (Paper)

Please send orders to:

Assumptions and Conventions vi

Chapter 1 Global Financial Stability Overview: Bridge to Recovery 1

Chapter 3 Corporate Funding: Liquidity Strains Cushioned by a Powerful Set of Policies 53

Chapter 4 Bank Capital: COVID-19 Challenges and Policy Responses 71

International Monetary Fund | October 2020 iii

iv International Monetary Fund | October 2020

Editor's Note (November 13, 2020)

International Monetary Fund | October 2020 v

vi International Monetary Fund | October 2020

Corrections and Revisions

Print and Digital Editions

Copyright and Reuse

International Monetary Fund | October 2020 vii

viii International Monetary Fund | October 2020

International Monetary Fund | October 2020 ix

x International Monetary Fund | October 2020

Bridge to Recovery: October 2020 Global Financial Stability Report at a Glance

Monetary policy—Maintain accommodation to support the recovery

Pandemic under Control

Monetary policy—Maintain accommodation until monetary policy objectives are achieved

Post-pandemic Financial Reform Agenda

International Monetary Fund | October 2020 xi

global economy is expected to grow by 5.2 percent in

in September (Figure 4), though still above its pre– 0.30

tries depending on the spread of the virus, the scope of

in China and the United States have outperformed other 0.10

markets, driven by technology stocks (dark and light

June 2020 Global Financial Stability Update, persists. 20

For example, analysis of year-to-date US stock market 15

sentiment despite higher economic uncertainty (see –20

Chapter 1). Similarly, the decline in corporate bond –25

tained for some time. However, if investors reassess the 2

scope for policy support or if the recovery is delayed, 0

the odds of a sharp adjustment may rise. –2 40

positions, as well as smaller firms—have experienced –8 0

greater financial stress than their peers in the early stages

which have taken a beating from COVID-19.

could see their capital buffers depleted to the levels

xiv International Monetary Fund | October 2020

abilities have increased across multiple sectors, with 0.9 OFIs

Financial vulnerability indices (percentile score)

ties in the corporate, banking, and sovereign sectors 0.6

(Figure 11). 0.5

Because of the pandemic, the financing needs of 0.4

policymakers should be mindful of intertemporal

vulnerabilities. Sources: Bank for International Settlements; Haver Analytics; Institute of

the COVID-19 crisis also presents an opportunity 3

to engineer a transition to a greener economy (see 2

After the pandemic is fully under control, policy 0

support can be gradually withdrawn and policy priori- –1

ties should focus on rebuilding bank buffers, strength- –2

ening regulation of nonbank financial institutions –3

International Monetary Fund | October 2020 xv

xvi International Monetary Fund | October 2020