## CHAPTER 1 Beneath the ResIlIence: RIsIng RIsks, UnResOlved vUlneRaBIlItIes

## Source details

**Canonical URL:** [CHAPTER 1 Beneath the ResIlIence: RIsIng RIsks, UnResOlved vUlneRaBIlItIes](https://www.imf.org/-/media/files/publications/gfsr/2026/october/english/ch1.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/gfsr/2026/october/english/ch1.pdf.md)
- [Structured JSON version](/-/media/files/publications/gfsr/2026/october/english/ch1.pdf.json)

---

### Financial market backdrop and overall assessment
- Financial markets have remained resilient since the April 2026 Global Financial Stability Report.
- Economic growth has held up and corporate earnings have been strong, particularly among firms related to artificial intelligence (AI).
- Since the second quarter of 2026, 14 of 33 major central banks have hiked policy rates, up from just 2 in the first quarter.
- Inflation has been more contained than initially feared, leading to only a modest rise in policy rates.
- Financial conditions tightened somewhat in the first quarter of 2026, but subsequent tailwinds from corporate valuations more than offset the tightening from higher interest rates, leaving overall financial conditions accommodative on net.
- Non-China emerging market economies experienced tightening in financial conditions this year to date, mostly in the first quarter.

### Salient near-term risks (selection and assessment)
- Three salient risks identified:
  - Artificial Intelligence (AI)-driven market correction.
  - Sharp rise in sovereign bond yields.
  - Cyber threats.
- Selection criteria: risks were chosen based on the likelihood of triggering shocks and the severity of potential market impact.
- Assessment: individually these risks do not necessarily threaten financial stability, but they could become systemic if amplified by underlying vulnerabilities that impair financial intermediation and credit provision.

### Summary of risks, vulnerabilities, and policy recommendations (high-level points)
- Artificial Intelligence (AI)-driven market correction
  - Trend: About the same
  - Triggering shocks: weaker or less durable AI investments and profitability
  - Impact: Could be significant given the growing size of AI-related assets; leverage is rising in the AI ecosystem
  - Policy recommendations: Contain leverage used by investors to gain exposures to AI where excessive (including limits on derivatives, exchange-traded products, and margin loans); strengthen system resilience using tools such as sector capital requirements, concentration limits, and capital buffers, as needed.
- Sharp rise in sovereign bond yields
  - Trend: Worse
  - Triggering shocks: resurgent inflation, term premium rises due to fiscal shocks
  - Impact: Could be severe given the foundational importance of sovereign bond markets; high debt and rollover risks remain unresolved
  - Policy recommendations: Put public debt on a sustainable path and improve bond market structure, like central clearing.
- Cyber threats
  - Trend: Worse
  - Triggering shocks: successful cyberattacks on critical financial institutions or market infrastructure
  - Impact: Could be severe as rapid digitization has increased the attack surface; frontier AI models could be used in both attacks and defense
  - Policy recommendations: Build specialized cyber capacity and enhance incident reporting, third-party risk oversight, and operational resilience.
- Weak banks in a stress scenario
  - Trend: About the same
  - Summary: Global banking system remains resilient in an adverse scenario, but vulnerable institutions remain and linkages of some banks to NBFIs have deepened
  - Policy recommendations: Ensure prudent provisioning and robust buffers, monitor interconnections with NBFIs, and implement international standards.
- Emerging market buffer erosion and alternative financing
  - Trend: Worse
  - Summary: Foreign exchange reserve buffers of weaker emerging markets have eroded; frontier economies are increasingly using less transparent and alternative forms of financing
  - Policy recommendations: Rebuild buffers, ensure policy mix is consistent with the IMF Integrated Policy Framework, and ensure robust disclosure of alternative borrowing sources.
- Liquidity mismatches of institutional investors
  - Trend: Worse
  - Summary: Insurer and pension fund holdings of illiquid assets have increased and become more concentrated, with liquidity mismatches rising
  - Policy recommendations: Monitor and stress test liquidity gaps, align fund liquidity with underlying asset liquidity.
- Leverage and liquidity risks in private credit
  - Trend: About the same
  - Summary: Default rates have fallen, but liquidity pressure is elevated and borrowers are more leveraged
  - Policy recommendations: Strengthen surveillance of interconnected risks and address excessive exposures of regulated entities to private credit.
- Household debt burden
  - Trend: Better
  - Summary: Debt burden has eased and households have sizable assets, but tighter financial conditions could erode wealth
  - Policy recommendations: Use macroprudential tools to ensure ongoing household resilience.

### Artificial Intelligence–driven market correction: evidence, exposures, and transmission
- AI investment has supported broader economic activity and helped offset possible negative effects from the war in the Middle East and overall policy uncertainty.
- The World Economic Outlook considers a scenario in which an unwinding of AI investment reduces global GDP, investment, and consumption by as much as 0.4, 1.3, and 0.6 percent, respectively, by 2028.
- Leverage and concentration indicators in the AI ecosystem:
  - Since 2024, hyperscalers’ share of global investment-grade bond issuance has increased to 10 percent from less than 1 percent.
  - Hedge funds’ gross equity positions plus equity margin loans have grown from 10 percent to about 15 percent of advanced economy GDP.
  - The share of single stocks in leveraged exchange-traded funds has increased to 28 percent from 16 percent.
- Circular financing: Fourteen publicly disclosed transactions since 2023 account for approximately $300 billion of aggregate equity investments led by hyperscalers in financially weaker AI firms, supplemented by $650 billion in debt, in exchange for $1.4 trillion in future revenue commitments.
  - About $320 billion of the $650 billion debt is taken on by investees in circular financing deals, including debt backed by investor guarantees, contingent financing, and other off-balance-sheet debt.
- Transmission and cross-border sensitivities:
  - Advanced Asian equity markets with high betas to a proxy of global AI equity returns will likely suffer most in a sharp slowdown in AI investment.
  - Many jurisdictions increased holdings of US equities from 2022 to 2026, reflecting greater allocations to hyperscalers and other AI-related firms, which could transmit stress even to jurisdictions with low direct equity betas to AI.

### Sovereign bond yield risk and market functioning
- Global sovereign bond yields have risen, driven by higher energy prices and increases in term premiums.
- Increase in 10-year bond yields between February 27 and September 30, 2026 has been driven by both higher-than-expected short-term rates and higher term premiums demanded by investors.
- Higher expected short-term rates have driven yields in most advanced economies; term premiums have played a larger role in emerging markets.
- Current term premiums are at the upper end of historical ranges, especially for advanced economy bonds.
- Downside propagation channels highlighted:
  - rollover risks for governments that recently shortened debt maturities, potentially raising debt burdens and term premiums;
  - selling of longer-term bonds by price-sensitive investors—foreign investors, investment funds, households and other funds—amplifying upward pressure on term premiums;
  - deterioration in market liquidity when volatility rises, given dependence on dealers for intermediation and balance-sheet constraints.
- Some market-functioning developments:
  - Repo borrowing by hedge funds collateralized by sovereign bonds in the euro area and the United States has partly unwound without notable pressure in repo markets.
  - Market liquidity metrics have improved for the euro area and the United Kingdom; metrics have continued to deteriorate in the Japanese government bond market.
  - In the US Treasury market, yield curve fitting errors, implied volatility, and bid-ask spreads point to favorable liquidity conditions in recent months.
- Policy recommendations:
  - Place public debt on a sustainable path.
  - Reduce rollover risks by establishing a more durable maturity profile for sovereign debt outstanding.
  - Strengthen market structure by greater central clearing of both bonds and funding instruments, namely repos.
  - Ensure central bank liquidity facilities are operationally ready to stem market dysfunction.

### Cyber threats and digital resilience
- Cyber disruptions are a growing source of systemic risk; a major attack affecting payments, trading, or access to deposits could cause immediate losses and undermine confidence.
- Three trends driving cyber risks:
  - rapid digitalization expanding the attack surface (cloud services, third-party providers, real-time payment and settlement systems);
  - growing reliance on a small number of critical technology providers and complex third-party dependency chains;
  - cyber threats becoming more sophisticated, including state-sponsored actors, ransomware groups, insiders, and frontier AI models with strong capabilities in network-attack simulations.
- Defensive and capacity challenges:
  - In more than 70 percent of cases thus far in 2026, software providers have had almost no time to patch vulnerabilities before exploitation began, a jump of 20 percentage points from just a year ago.
  - Regions such as Asia face significant shortages of cybersecurity talent.
- Policy recommendations:
  - Strengthen foundational cyber hygiene, incident reporting, crisis preparedness, and oversight of critical third-party providers—particularly in emerging markets.
  - Build specialized cyber expertise and adopt risk-based, proportionate approaches to AI governance in finance.

### Vulnerabilities in the global banking system and 2026 stress-test findings
- IMF’s 2026 Global Bank Stress Test scenario assumptions and impacts:
  - reassessment of AI productivity gains triggers a sharp correction in technology valuations;
  - short- and long-term bond yields rise by 150 and 300 basis points, respectively;
  - global real GDP growth falls about 3 percentage points below baseline in the first two years of the scenario and only partly recovers thereafter;
  - global unemployment rises 3.5 percentage points;
  - inflation increases by over 4 percentage points by 2027;
  - the global banking system’s aggregate Common Equity Tier 1 (CET1) ratio declines by about 150 basis points to approximately 11.5 percent, driven primarily by credit impairments partly offset by stronger net interest income.
- Aggregate resilience masks pockets of weakness:
  - Banks representing 22 percent of global banking assets are classified as weak—weak banks being those with post-stress CET1 ratios falling below 7 percent.
  - The share of weak banks increased from 18 percent of global banking assets in the 2025 stress test to 22 percent in 2026.
- Regional outcomes and exposures:
  - North America: increase in weak banks driven by market risk as sharp increases in long-term government bond yields trigger asset valuation losses, particularly in a small number of large banks.
  - Asia: share of weak institutions reflects significant capital depletions in oil- and gas-importing economies.
  - Most other regions, particularly emerging markets, experience more limited capital depletion.
- Banks’ exposures to NBFIs and private assets:
  - In the United States, loans to NBFIs increased by 18 percent in 2026; exposures to private equity and private credit firms rose by 30 percent.
  - In a scenario where bank losses equal 20 percent of their exposures to private assets, aggregate CET1 capital would decline by about 2 percentage points, and banks representing less than 1.6 percent of assets in the sample would see their Tier 1 capital ratios fall below 7 percent.
  - In the European Union and the United Kingdom, under the same shock, losses represent 1 percentage point of CET1 capital and no banks fall below the 7 percent Tier 1 threshold.
- Policy recommendations:
  - Ensure prudent provisioning and robust buffers.
  - Monitor interconnections with NBFIs and implement international standards.
  - Complete implementation of Basel III and other internationally agreed-upon standards.

### Emerging market and frontier economy buffers, reserve adequacy, and funding
- Authorities deployed a broad range of measures to cushion the energy shock: targeted fiscal support, monetary policy action, targeted currency interventions, and capital flow measures.
- Reserve adequacy:
  - The median reserve adequacy ratio as of June 2026 for investment-grade countries is 135 percent and for below-investment-grade countries is 117 percent.
  - For weaker emerging markets that are both below-investment-grade and had a reserve adequacy ratio below 100 percent, the median ratio fell from 72 percent in the first quarter of 2026 to 63 percent in the second quarter.
- Drawdowns in official reserves have been highly uneven, with greater erosion among below-investment-grade sovereigns.
- Foreign exchange interventions were concentrated in Europe and the Middle East in March 2026, rotating toward Asia and Africa in the second and third quarters.
- Frontier economies:
  - Frontier spreads widened by roughly 60 basis points at the onset of the war in the Middle East in late February 2026 but have more than retraced since.
  - Market access broadened since February, with frontier sovereign issuers accounting for close to one-fifth of emerging market sovereign supply in the first half of 2026.
  - The current combination of historically compressed sovereign credit spreads alongside a 10-year US Treasury yield above 4.5 percent is historically unusual and was last observed in 2007.
- Policy recommendations:
  - Rebuild policy buffers and ensure an overall policy mix consistent with the IMF’s Integrated Policy Framework.
  - Rebuild external buffers where they have eroded.
  - Strengthen medium-term fiscal and debt management frameworks and enhance disclosure and comprehensive debt transparency.

### Institutional investors, private credit, and household finances
- Institutional investors and illiquidity:
  - Pension funds and insurers have nearly doubled their share of illiquid investments over the past decade.
  - Pension funds and insurers hold more than one-quarter of publicly traded corporate bonds in the United States and the euro area.
  - The three largest managers could now account for more than half of institutional investments in private equity and private credit.
  - Policy recommendations: monitor and stress test liquidity gaps, align fund liquidity with underlying asset liquidity, and close data gaps.
- Private credit and direct lending:
  - Direct lending now accounts for nearly $2 trillion globally in loans.
  - The ratio of the median borrower’s debt-to-EBITDA has risen from 2024 through 2026, especially in the software and services industries.
  - Semiliquid vehicles accounted for more than 50 percent of new fundraising in the US in 2025 prior to the recent redemption wave.
  - Among 10 BDCs with $300 billion in total assets, 4—representing about $90 billion in assets—have commitments exceeding their available liquidity buffers.
  - In a stress scenario, borrowers could draw heavily on contingent facilities just as funds’ liquidity resources are eroded by lower collateral values.
  - Policy recommendations: strengthen systemwide monitoring, close data gaps on leverage and contingent liquidity commitments, and enhance disclosures to support stress testing.
- Household finances:
  - Household debt-service ratios have broadly improved in recent quarters and remain appreciably lower than peak levels reached after central banks began to hike interest rates in 2022.
  - Households’ equity and real estate holdings account for roughly 20 and 40 percent of their assets, respectively, or 1.2 and 2.5 times global GDP.
  - Lower-income households could be disproportionately affected by higher interest rates and weaker real income growth; delinquency rates on consumer debt tied to households, like credit card debt in the United States, have risen.
  - Sensitivity analysis: investment exposure, including equities, has risen by 1 to 3 percentage points for the bottom 20 percent of households in the United States, the euro area, and Japan over the past five years.

### Corporate zombies and corporate distress in Asia
- At the end of 2025, credit in Asia averaged 107 percent of GDP, compared with an average of 99 percent for emerging market economies and 93 percent globally, up from 81 percent in 2007.
- Across 10 Asian economies, zombie firms—defined as firms for which earnings have not covered interest expenses for at least three consecutive years—now account for 21.4 percent of corporate debt outstanding in the region, having increased from under 10 percent.
- Counterfactual estimation for the 10 Asian economies: if the stock of zombie debt had been subjected to hazard and recovery rates typical of firms with sustained distress, almost 23 percent of their debt would have been written off or restructured as of 2025.
  - That 23 percent of zombie debt amounts to more than 5 percent of total corporate debt.
  - Annual default hazard used in the central calibration: 9 percent.
  - Creditor recovery upon default: 35 percent.
- Bank forbearance: banks with Tier 1 capital below the median of their country counterparts have 16 percent of their loans with zombie firms, compared with 12 percent for better capitalized banks.

### Policy priorities and cross-cutting recommendations
- Contain excessive leverage used to finance the AI ecosystem where excessive, including by restricting equity margin financing, limiting leveraged exchange-traded funds, and containing short-term options.
- Improve transparency and reporting of structured financing instruments such as sovereign total return swaps.
- Manage liquidity and leverage risks in direct lending and address contingent liquidity commitments that exceed available buffers.
- Monitor and address concentration risks in the institutional investment chain to reduce the likelihood of correlated portfolio adjustments during stress.
- Use a mix of policy tools—monetary tightening alongside foreign exchange intervention where appropriate—to strengthen exchange rate recoveries, recognizing outcomes depend on underlying vulnerabilities and credible macroeconomic policies.
- Strengthen cyber resilience by prioritizing foundational cyber hygiene, incident reporting, crisis coordination, and supervisory capacity, tailored to country maturity.

*International Monetary Fund, Global Financial Stability Report, October 2026.*

### Introduction

### Introduction

### Financial market backdrop and overall assessment
- Financial markets have remained resilient since the April 2026 Global Financial Stability Report.
- Economic growth has held up and corporate earnings have been strong, particularly among firms related to artificial intelligence (AI).
- Since the second quarter of 2026, 14 of 33 major central banks have hiked policy rates, up from just 2 in the first quarter.
- Inflation has been more contained than initially feared, leading to only a modest rise in policy rates.
- Financial conditions tightened somewhat in the first quarter of 2026, but subsequent tailwinds from corporate valuations more than offset the tightening from higher interest rates, leaving overall financial conditions accommodative on net.
- Non-China emerging market economies experienced tightening in financial conditions this year to date, mostly in the first quarter.

### Salient near-term risks identified
- The chapter identifies three salient risks ahead: an AI-driven market correction, a sharp rise in sovereign bond yields, and cyber threats.
- These risks were selected based on the likelihood of their triggering shocks and the severity of their potential market impact.
- Individually, these risks do not necessarily threaten financial stability, but they could become systemic if amplified by underlying vulnerabilities that impair financial intermediation and credit provision.

### Vulnerabilities that could amplify risks
- Global stress tests suggest the banking system would remain adequately capitalized in a bond yield spike and AI-led correction scenario, although a significant share of vulnerable institutions would emerge and some banks’ growing interconnectedness with nonbank financial intermediaries (NBFIs) could prove a source of fragility.
- Emerging market vulnerabilities include eroding external buffers in some economies and growing reliance on alternative, often opaque, financing among frontier economies.
- Vulnerabilities persist among institutional investors through rising liquidity mismatches and concentrated investments, particularly in private credit direct lending through higher leverage and liquidity mismatches.
- On the positive side, household debt remains relatively low in aggregate.

### Summary table of risks, vulnerabilities, and policy recommendations (high-level points)
- Artificial Intelligence (AI)-driven market correction
  - Trend: About the same
  - Triggering shocks: weaker or less durable AI investments and profitability
  - Impact: Could be significant given the growing size of AI-related assets; leverage is rising in the AI ecosystem
  - Policy recommendations: Contain leverage used by investors to gain exposures to AI where excessive (including limits on derivatives, exchange-traded products, and margin loans); strengthen system resilience using tools such as sector capital requirements, concentration limits, and capital buffers, as needed.
- Sharp rise in sovereign bond yields
  - Trend: Worse
  - Triggering shocks: resurgent inflation, term premium rises due to fiscal shocks
  - Impact: Could be severe given the foundational importance of sovereign bond markets; high debt and rollover risks remain unresolved
  - Policy recommendations: Put public debt on a sustainable path and improve bond market structure, like central clearing.
- Cyber threats
  - Trend: Worse
  - Triggering shocks: successful cyberattacks on critical financial institutions or market infrastructure
  - Impact: Could be severe as rapid digitization has increased the attack surface; frontier AI models could be used in both attacks and defense
  - Policy recommendations: Build specialized cyber capacity and enhance incident reporting, third-party risk oversight, and operational resilience.
- Weak banks in a stress scenario
  - Trend: About the same
  - Summary: Global banking system remains resilient in an adverse scenario, but vulnerable institutions remain and linkages of some banks to NBFIs have deepened
  - Policy recommendations: Ensure prudent provisioning and robust buffers, monitor interconnections with NBFIs, and implement international standards.
- Emerging market buffer erosion and alternative financing
  - Trend: Worse
  - Summary: Foreign exchange reserve buffers of weaker emerging markets have eroded; frontier economies are increasingly using less transparent and alternative forms of financing
  - Policy recommendations: Rebuild buffers, ensure policy mix is consistent with the IMF Integrated Policy Framework, and ensure robust disclosure of alternative borrowing sources.
- Liquidity mismatches of institutional investors
  - Trend: Worse
  - Summary: Insurer and pension fund holdings of illiquid assets have increased and have become more concentrated, with liquidity mismatches rising
  - Policy recommendations: Monitor and stress test liquidity gaps, align fund liquidity with underlying asset liquidity.
- Leverage and liquidity risks in private credit
  - Trend: About the same
  - Summary: Default rates have fallen, but liquidity pressure is elevated and borrowers are more leveraged
  - Policy recommendations: Strengthen surveillance of interconnected risks and address excessive exposures of regulated entities to private credit.
- Household debt burden
  - Trend: Better
  - Summary: Debt burden has eased and households have sizable assets, but tighter financial conditions could erode wealth
  - Policy recommendations: Use macroprudential tools to ensure ongoing household resilience.

### Artificial Intelligence–driven market correction: evidence and transmission channels
- AI investment has supported broader economic activity and helped offset possible negative effects from the war in the Middle East and overall policy uncertainty.
- Firms at the center of the AI value chain have experienced strong equity price appreciation and exceptional earnings growth; forward price-to-earnings ratios have eased in several equity markets.
- The World Economic Outlook considers a scenario in which an unwinding of AI investment reduces global GDP, investment, and consumption by as much as 0.4, 1.3, and 0.6 percent, respectively, by 2028.
- Leverage sources in the AI ecosystem:
  - Since 2024, hyperscalers’ share of global investment-grade bond issuance has increased to 10 percent from less than 1 percent.
  - Hedge funds’ gross equity positions plus equity margin loans have grown from 10 percent to about 15 percent of advanced economy GDP.
  - The share of single stocks in leveraged exchange-traded funds has increased to 28 percent from 16 percent.
- Circular financing: Fourteen publicly disclosed transactions since 2023 account for approximately $300 billion of aggregate equity investments led by hyperscalers in financially weaker AI firms, supplemented by $650 billion in debt, in exchange for $1.4 trillion in future revenue commitments.
  - About $320 billion of the $650 billion debt is taken on by investees in circular financing deals, including debt backed by investor guarantees, contingent financing, and other off-balance-sheet debt.
- Transmission and cross-border sensitivities:
  - Advanced Asian equity markets that exhibit high betas to a proxy of global AI equity returns will likely suffer the most in a sharp slowdown in AI investment.
  - Many jurisdictions increased holdings of US equities from 2022 to 2026, reflecting greater allocations to hyperscalers and other AI-related firms, which could transmit stress even to jurisdictions with low direct equity betas to AI.

### Sovereign bond yield risk and market functioning
- Global sovereign bond yields have risen, driven by higher energy prices and increases in term premiums.
- High debt and rollover risks remain unresolved, making the bond market susceptible to more frequent fiscal and inflation shocks even though market functioning has been orderly to date.

*Source: International Monetary Fund, October 2026.*

### CHAPTER 1 Beneath the ResIlIence: RIsIng RIsks, UnResOlved vUlneRaBIlItIes

### CHAPTER 1 Beneath the ResIlIence: RIsIng RIsks, UnResOlved vUlneRaBIlItIes

### Sovereign bond market fragilities
- Increase in 10-year bond yields between February 27 and September 30, 2026 has been driven by both higher-than-expected short-term rates and higher term premiums demanded by investors (Figure 1.4, panel 1).
- Higher expected short-term rates have driven yields in most advanced economies; term premiums have played a larger role in emerging markets.
- Across economies, the rise in term premiums reflects compensation for several risks, most notably renewed concerns about persistent fiscal deficits and high government debt levels (see April 2026 Fiscal Monitor) and the associated forthcoming increase in bond supply.
- For emerging market economies, higher term premiums may have resulted from spillovers from higher yields in advanced economies.
- Surging AI-related debt issuances that may compete with sovereign bonds for investor demand has been cited as an additional source of pressure on bond markets.
- Current term premiums are at the upper end of historical ranges, especially for advanced economy bonds (Figure 1.4, panel 2).
- Downside scenario: further rises in energy prices could reignite inflation and force tighter monetary policy, propagating through:
  - rollover risks for governments that recently shortened debt maturities, potentially raising debt burdens and term premiums;
  - selling of longer-term bonds by price-sensitive investors—foreign investors, investment funds, households and other funds—amplifying upward pressure on term premiums;
  - deterioration in market liquidity when volatility rises, given dependence on dealers for intermediation and balance-sheet constraints.
- The World Economic Outlook analyzes macroeconomic outcomes under a scenario featuring such a shock.
- Some fragilities have improved:
  - Repo borrowing by hedge funds collateralized by sovereign bonds in the euro area and the United States has partly unwound without notable pressure in repo markets (Figure 1.4, panel 3).
  - Market liquidity metrics have improved for the euro area and the United Kingdom, especially yield curve fitting errors and market breadth metrics (Figure 1.5).
  - Market liquidity metrics have continued to deteriorate in the Japanese government bond market, driven by yield volatility since the end of yield curve control and ongoing monetary policy normalization, and wider bid-ask spreads, especially for long-term bonds.
  - In the US Treasury market, yield curve fitting errors, implied volatility, and bid-ask spreads point to favorable liquidity conditions in recent months.
  - Fiscal initiatives have reduced sovereign funding spreads in some economies (see Box 1.2 for a case study on such a program in the euro area).
- Other fragilities have worsened:
  - Rollover risk remains high and could be exacerbated by debt-management strategies that replace long-term debt with short-term debt or available cash at scale.
  - Historically high government debt levels remain unresolved, and persistently high fiscal deficits in major economies continue to add pressure.
- A crystallization of these fragilities could lead to market turmoil: a sharp rise in bond yields would depreciate asset valuations through higher discount rates (including on AI-related assets) and erode the value of government bonds used as collateral.

*Sources: International Monetary Fund; April 2026 Fiscal Monitor; World Economic Outlook; Financial Times (September 27, 2026); PIMCO Perspectives (Seidner and Dhawan 2026).*

### Cyber threats and digital resilience
- As finance becomes more digital and interconnected, cyber disruptions have emerged as a growing source of systemic risk; a major attack affecting payments, trading, or access to deposits could cause immediate losses and undermine confidence.
- Three trends driving cyber risks:
  - rapid digitalization expanding the attack surface (cloud services, third-party providers, real-time payment and settlement systems);
  - growing reliance on a small number of critical technology providers and complex third-party dependency chains, increasing concentration risk and common points of failure;
  - cyber threats becoming more sophisticated, including state-sponsored actors, ransomware groups, insiders, and frontier AI models with strong capabilities in network-attack simulations (Adrian and others 2026).
- Cyber defense is becoming more challenging:
  - In more than 70 percent of cases thus far in 2026, software providers have had almost no time to patch vulnerabilities before exploitation began, a jump of 20 percentage points from just a year ago (Figure 1.6, panel 1).
  - Regions such as Asia face significant shortages of cybersecurity talent (Figure 1.6, panel 2).
- AI can help address shortages through stronger monitoring, threat detection, and automated response, but AI also increases urgency to strengthen foundational cyber resilience, including cyber hygiene, incident reporting, crisis preparedness, and oversight of critical third-party providers—particularly in emerging markets where capacity constraints remain significant.

*Sources: Adrian and others (2026); Global Cybersecurity Forum (2024); zerodayclock.com; International Monetary Fund.*

### Vulnerabilities in the global banking system and stress test findings
- The IMF’s 2026 Global Bank Stress Test evaluates banks’ ability to withstand an adverse scenario combining weaker-than-expected productivity gains from AI with higher interest rates, bond market fragilities, and energy supply disruptions.
- Scenario assumptions and impacts:
  - reassessment of AI productivity gains triggers a sharp correction in technology valuations, generating broad spillovers across global financial markets;
  - short- and long-term bond yields rise by 150 and 300 basis points, respectively;
  - global real GDP growth falls about 3 percentage points below baseline in the first two years of the scenario and only partly recovers thereafter;
  - global unemployment rises 3.5 percentage points;
  - inflation increases by over 4 percentage points by 2027;
  - the global banking system’s aggregate Common Equity Tier 1 (CET1) ratio declines by about 150 basis points to approximately 11.5 percent, driven primarily by credit impairments partly offset by stronger net interest income.  
- Aggregate resilience masks pockets of weakness:
  - Banks representing 22 percent of global banking assets are classified as weak—weak banks being those with post-stress CET1 ratios falling below 7 percent.
  - The share of weak banks increased from 18 percent of global banking assets in the 2025 stress test to 22 percent in 2026.
- Regional stress-test outcomes:
  - North America: increase in weak banks driven by market risk as sharp increases in long-term government bond yields trigger asset valuation losses, particularly in a small number of large banks.
  - Asia: share of weak institutions reflects significant capital depletions in oil- and gas-importing economies.
  - Most other regions, particularly emerging markets, experience more limited capital depletion, supported by stronger net interest and fee income generation.
- Banks’ exposures to NBFIs and private assets:
  - In the United States, loans to NBFIs increased by 18 percent in 2026; exposures to private equity and private credit firms rose by 30 percent.
  - In a scenario where bank losses equal 20 percent of their exposures to private assets, aggregate CET1 capital would decline by about 2 percentage points, and banks representing less than 1.6 percent of assets in the sample would see their Tier 1 capital ratios fall below 7 percent (Figure 1.7, panel 2).
  - In the European Union and the United Kingdom, under the same shock, losses represent 1 percentage point of CET1 capital and no banks fall below the 7 percent Tier 1 threshold.
  - Stand-alone NBFI and private asset stress events appear absorbable by banks, but vulnerabilities could be amplified if such losses coincide with a severe economic downturn.

*Sources: IMF 2026 Global Bank Stress Test; Background Analysis 1.2, “2026 IMF Global Banking Stress Test”; Fitch Connect; Fitch Solutions; IMF staff calculations; Consolidated Reports of Condition and Income; S&P Capital IQ Pro.*

### Emerging market and frontier economy buffers
- Emerging market and frontier economies have proved relatively resilient during recent shocks by deploying policy measures; risk premiums retraced to levels last seen before the war in the Middle East.
- Widening of emerging market risk premiums at the outbreak of the war was concentrated mainly in currencies, local rates, and equities, but premiums have retraced to historically compressed levels (Figure 1.8, panel 1).
- Authorities deployed a broad range of measures to cushion the energy shock: targeted fiscal support, monetary policy action, targeted currency interventions, and capital flow measures.
- Policy buffers have weakened in some cases:
  - Measures to cushion the energy shock have weighed on fiscal space and reserve buffers, leaving less room to maneuver if energy price pressures intensify or global financial conditions tighten.
- Reserve adequacy:
  - The median reserve adequacy ratio as of June 2026 for investment-grade countries is 135 percent and for below-investment-grade countries is 117 percent (Figure 1.8, panel notes).
- Historical episodes used for comparison include the April 2025 US tariff announcement; February 2022 Russia-Ukraine war; February 2020 COVID-19 outbreak; February 2018 “Volmageddon”; August 2015 renminbi devaluation; and May 2013 “taper tantrum.”

*Sources: IMF staff calculations; Bloomberg Finance L.P.; IMF, International Reserve and Foreign Currency Liquidity dataset.*

*International Monetary Fund | October 2026*

### 1. Emerging Market Composite Risk Index

### 1. Emerging Market Composite Risk Index

### Reserve adequacy and foreign exchange interventions
- Drawdowns in official reserves have been highly uneven, with greater erosion among below-investment-grade sovereigns.
- Investment-grade economies drew down reserves broadly in line with past stress episodes; below-investment-grade sovereigns depleted reserves far more sharply, starting from already weaker adequacy levels (Figure 1.8, panel 2).
- Lower gold prices compounded the decline where gold had a larger share in official reserves.
- Reserve adequacy remains comfortable in aggregate but across most emerging markets remains below pre-war levels.
- For weaker emerging markets that are both below-investment-grade and had a reserve adequacy ratio below 100 percent, the median ratio fell from 72 percent in the first quarter of 2026 to 63 percent in the second quarter.
- Foreign exchange interventions were concentrated in Europe and the Middle East in March 2026, rotating toward Asia and Africa in the second and third quarters.
- Several Asian economies used foreign exchange interventions and in some cases targeted capital flow management measures; economies that relied solely on foreign exchange interventions saw little meaningful recovery in exchange rates from their weakest level since end February, whereas economies that complemented intervention with monetary tightening generally experienced stronger exchange rate recoveries.

### Frontier economies: market access, funding costs, and external debt
- Frontier spreads widened by roughly 60 basis points at the onset of the war in the Middle East in late February 2026 but have more than retraced since, with spreads narrowing toward levels last seen in late 2024.
- Rising US Treasury yields have left all-in hard-currency funding costs little changed since February, but these costs are still down by roughly 200 basis points since early 2024.
- Market access broadened since February, with frontier sovereign issuers accounting for close to one-fifth of emerging market sovereign supply in the first half of 2026, including several first-time borrowers with lower sovereign credit ratings.
- The current combination of historically compressed sovereign credit spreads alongside a 10-year US Treasury yield above 4.5 percent is historically unusual and was last observed in 2007.
- Among frontier economies with the largest external debt increases, bond markets funded about one-fifth of the debt increase between 2018 and 2024 for those rated B− or higher.
- For sovereigns rated CCC+ or below with large debt increases, bond exposure declined even as external debt increased; these sovereigns relied mainly on official financing.
- Some sovereigns have begun to raise funding using sovereign total return swaps; these transactions can create vulnerabilities when borrowing costs rise and are publicly reported in only a limited number of economies, weakening transparency.

### Institutional investors: illiquid asset exposure and concentration
- Pension funds and insurers have nearly doubled their share of illiquid investments over the past decade (Figure 1.11, panel 1).
- Pension funds and insurers hold more than one-quarter of publicly traded corporate bonds in the United States and the euro area.
- A small group of investment consultants plays an increasingly influential role in advising pension funds; the three largest managers could now account for more than half of institutional investments in private equity and private credit (Figure 1.11, panel 2).
- Growing concentration across the investment chain could increase the likelihood that institutions respond to liquidity pressures in similar ways, amplifying market stress.

### Private credit direct lending: leverage and contingent liquidity risks
- Direct lending now accounts for nearly $2 trillion globally in loans.
- The ratio of the median borrower’s debt-to-EBITDA has risen from 2024 through 2026, especially in the software and services industries (Figure 1.12, panel 1).
- A persistently large share of borrowers are unable to generate sufficient cash flow to meet debt-servicing obligations and remain reliant on sponsor support, lender forbearance, or continued access to external financing.
- Semiliquid vehicles accounted for more than 50 percent of new fundraising in the US in 2025 prior to the recent redemption wave.
- Among 10 BDCs with $300 billion in total assets, 4—representing about $90 billion in assets—have commitments exceeding their available liquidity buffers (Figure 1.12, panel 2).
- In a stress scenario, borrowers could draw heavily on contingent facilities just as funds’ liquidity resources are eroded by lower collateral values, potentially transforming liquidity pressures into a contraction in credit supply.

### Household finances
- Household debt-service ratios have broadly improved in recent quarters across countries and remain appreciably lower than peak levels reached after central banks began to hike interest rates in 2022 (Figure 1.13).
- Countries with more floating-rate mortgages tend to have higher debt-service ratios and are more sensitive to interest rate resets.
- Households’ equity and real estate holdings account for roughly 20 and 40 percent of their assets, respectively, or 1.2 and 2.5 times global GDP.
- Lower-income households could be disproportionately affected by higher interest rates and weaker real income growth; delinquency rates on consumer debt tied to households, like credit card debt in the United States, have risen.

### Policy recommendations and priorities
- Contain excessive leverage used to finance the AI ecosystem where excessive.
- Improve transparency and reporting of structured financing instruments such as sovereign total return swaps to mitigate vulnerabilities and underpin investor confidence.
- Manage liquidity and leverage risks in direct lending, including addressing contingent liquidity commitments that exceed available buffers.
- Monitor and address concentration risks in the institutional investment chain to reduce the likelihood of correlated portfolio adjustments during stress.
- Use a mix of policy tools—monetary tightening alongside foreign exchange intervention where appropriate—to strengthen exchange rate recoveries, recognizing outcomes depend on underlying vulnerabilities and credible macroeconomic policies.

*International Monetary Fund, Global Financial Stability Report, October 2026.*

### Chapter 1 of the April 2025 Global Financial Stability Report

### Chapter 1 — Beneath the Resilience: Rising Risks, Unresolved Vulnerabilities

### Sensitivity analysis of household debt and household finances
- The chapter contains a sensitivity analysis of global household debt to interest rates and income changes.
- Investment exposure, including equities, has risen by 1 to 3 percentage points for the bottom 20 percent of households in the United States, the euro area, and Japan over the past five years (latest available data from the Federal Reserve’s Flow of Funds; the European Central Bank’s Household Finance and Consumption Survey; and Japan’s Family Income and Expenditure Survey).
- Household debt servicing burden has improved in jurisdictions with less floating rate mortgages.
- Debt service ratio definition and data sources:
  - The debt service ratio is defined as required debt payments (interest plus principal) divided by disposable income.
  - Calculations are made by the Bank for International Settlements for countries aside from Brazil, Mexico, and South Africa.
  - For BIS-derived or Brazil and South Africa numbers, “latest observation” is the fourth quarter of 2025, and “prior observation” is fourth quarter of 2024.
  - For Mexico, the “latest observation” is for the fourth quarter of 2024 and the “prior observation” is for the fourth quarter of 2023.
  - The share of fixed-rate mortgages is based on the definition of De Stefani and Mano (2025).

### AI, leverage, and market amplification risks
- Findings and concerns:
  - AI-related equity price declines could trigger forced deleveraging and volatility amplification.
  - AI accelerates vulnerability discovery, exploitation, and attack in cybersecurity, compressing the attack–defense cycle to machine speed.
- Policy recommendations:
  - Increase oversight of means to obtain leverage, including by retail investors, for example by placing restrictions on equity margin financing, limiting the proliferation of leveraged exchange-traded funds, and containing the use of short-term options (see the April 2026 Global Financial Stability Report).
  - Financial sector supervisors and macroprudential authorities should assess potential systemwide losses arising from AI-related exposures, including amplification from highly interconnected and circular financing arrangements.
  - Deploy appropriate tools where vulnerabilities could be material, including sectoral capital requirements; concentration limits, including in prime brokerage relationships; and capital buffers.

### Bond market fragilities and sovereign debt management
- Core recommendations:
  - Place public debt on a sustainable path as the first line of defense against bond market fragilities.
  - Reduce rollover risks by establishing a more durable maturity profile for sovereign debt outstanding.
  - Strengthen market structure to prevent a seizing of market liquidity during stress episodes by greater central clearing of both bonds and funding instruments, namely repos.
  - Ensure central bank liquidity facilities are operationally ready to stem market dysfunction.
- Central bank role:
  - Deliver on mandates, preserve independence, and communicate monetary policy decisions clearly.
  - Communications should explain the policy reaction function, how incoming data are likely to inform decisions, and the risks surrounding the outlook to help anchor medium- and long-term inflation expectations and mitigate market sensitivity to individual data releases.

### Cybersecurity and operational resilience
- Policy priorities by country maturity:
  - Advanced economies: strengthen operational resilience, third-party risk oversight, and resilience testing; identify critical business services; set impact tolerances; map operational dependencies; conduct regular resilience testing.
  - Emerging markets and developing economies: prioritize foundational cyber hygiene, incident reporting, crisis coordination, and supervisory capacity before adopting advanced AI-related oversight frameworks.
  - Low-income countries: focus scarce resources on essential controls and critical infrastructure protection.
- Cross-jurisdiction recommendations:
  - Supervisors should build specialized cyber expertise, adopt risk-based and proportionate approaches, and ensure robust governance, operational resilience, and human oversight of AI systems.

### Banks, nonbank financial institutions, and systemwide safeguards
- Banks:
  - Although banks remain resilient, an adverse scenario with lower growth, higher inflation and interest rates, energy shocks, and weaker AI monetization increases the share of banks that appear weak.
  - Supervisors in energy-sensitive jurisdictions should ensure timely and accurate asset classification and adequate provisioning to cover expected credit losses.
  - Complete implementation of Basel III and other internationally agreed-upon standards to ensure banks maintain adequate capital and liquidity buffers.
  - Reviews to reduce unnecessary complexity should not weaken resilience, create regulatory arbitrage, or undermine international consistency.
  - Intensify monitoring of banks’ exposures to NBFIs and assess how stress at lightly regulated institutions could affect bank solvency and liquidity.
- Institutional investors and NBFIs:
  - The growing size, concentration, and illiquidity of institutional investor portfolios calls for broader risk monitoring across pension funds and insurers.
  - Authorities should identify common exposures, leverage, and liquidity mismatches, and increase the use of solvency and liquidity stress tests.
  - Close data gaps and strengthen information sharing domestically and across borders.
  - Ensure liquidity of investment products remains aligned with the liquidity of underlying assets through appropriate redemption terms and notice periods.
- Private credit:
  - Strengthen systemwide monitoring across the entire financing chain, from portfolio companies and special-purpose vehicles to funds and investors.
  - Close data gaps on leverage, contingent liquidity commitments, and cross-sectoral and cross-border exposures.
  - Enhanced disclosures to support stress testing and surveillance tools that assess shock propagation.

### Emerging markets: policy mix, buffers, and debt transparency
- Policy recommendations:
  - Rebuild policy buffers and ensure an overall policy mix consistent with the IMF’s Integrated Policy Framework.
  - Rebuild external buffers where they have eroded to preserve market access and the capacity to respond to shifts in global risk sentiment.
  - Calibrate monetary, exchange rate, macroprudential, and capital flow measures to country-specific circumstances and deploy interventions only to address frictions or disorderly market conditions.
  - Strengthen medium-term fiscal and debt management frameworks to safeguard debt sustainability and limit refinancing risks.
  - Ensure spending aimed at protecting vulnerable groups from inflation or an energy shock is temporary and targeted.
  - Enhance disclosure and comprehensive debt transparency given increased recourse to alternative borrowing sources.

### Risks versus vulnerabilities: selection and assessment framework
- Conceptual distinctions:
  - Risks: plausible near-term catalysts for market turmoil combining likelihood of triggering shocks and potential severity of impact.
  - Vulnerabilities: structural imbalances that can amplify risks (elevated leverage, liquidity mismatches, concentrated exposures, interconnectedness, weak external or fiscal positions).
  - Financial instability/systemic risk arises when risks are amplified by vulnerabilities, impairing key financial institutions and markets and disrupting credit or financial services.
- Selection and assessment process:
  - Candidate risks and vulnerabilities are drawn from IMF staff daily monitoring of financial markets and institutions, market reports, analytical tools, and outreach.
  - Considerations for selection:
    - Risks: likelihood of shocks; severity of potential market impacts.
    - Vulnerabilities: potential to amplify risks; extent of recent changes; whether novel analyses can shed new light.
  - Staff assess trends as “better,” “about the same,” or “worse” using empirical indicators, balance sheet data, stress tests, and other analytical tools.
  - Overall financial stability risks are assessed based on the combined trajectory of risks and vulnerabilities and may be benchmarked against growth-at-risk estimates (Adrian and others 2019).

### Euro area common borrowing and sovereign bond market fragilities
- Empirical findings:
  - Yield differentials between euro area core and southern member states (Greece, Italy, Portugal, and Spain) have narrowed markedly in recent years.
  - Convergence reflects stronger fiscal positions in southern states and ECB measures (for example, the Pandemic Emergency Purchase Programme and the Transmission Protection Instrument).
  - Since the launch of the Next Generation EU (NGEU), the European Commission has become a relatively large, regular borrower across maturities; common borrowing and transfers have strengthened perceptions of fiscal risk sharing and helped reduce concerns about debt sustainability and redenomination risk.
  - Empirical evidence suggests these initiatives have dampened the effect of systemic shocks on sovereign bond markets: the association of sovereign bond asset-swap spreads to systemic stress—gauged by the ECB’s Composite Indicator of Systemic Stress—has become less negative since the introduction of the NGEU.
- Implication drawn in the chapter:
  - Risk-sharing programs, where applicable in currency unions, can supplement fiscal consolidation in mitigating bond market fragilities.
- Analysis coverage:
  - The analysis covers the period from January 2016 until the end of July 2026.

### Private sector debt and corporate distress in Asia
- Private sector debt levels:
  - At the end of 2025, credit in Asia averaged 107 percent of GDP, compared with an average of 99 percent for emerging market economies and 93 percent globally, up from 81 percent in 2007.
- Corporate distress:
  - Across 10 Asian economies, zombie firms—defined as firms for which earnings have not covered interest expenses for at least three consecutive years—now account for 21.4 percent of corporate debt outstanding in the region, having increased from under 10 percent.

*Source: Chapter 1 of the April 2025 Global Financial Stability Report, International Monetary Fund.*

### 2008. The share of zombie firms is relatively small

### 2008. The share of zombie firms is relatively small

### Prevalence and distribution
- In Latin America, Europe, and Africa, on average, about one in eight firms were zombies in 2024, and were concentrated in a handful of large borrowers.
- Estimates of corporate zombies in Asia include China, Hong Kong SAR, India, Indonesia, Korea, Malaysia, the Philippines, Singapore, Taiwan Province of China, and Thailand.
- Firms exit from zombie status after they cover their interest expenses for three consecutive years.

### Bank forbearance and mechanics
- Corporate zombies are typically sustained through bank forbearance.
- Recognizing losses on distressed borrowers consumes capital, whereas rolling over loans defers losses and preserves reported capital ratios; undercapitalized banks tend to extend their bad loan exposure.
- In the Asian economies, banks with Tier 1 capital below the median of their country counterparts have 16 percent of their loans with zombie firms, compared with 12 percent for better capitalized banks.
- Bank forbearance therefore locks in both weak borrowers and weak banks.

### Quantitative estimates and impacts
- Counterfactual estimation for the 10 Asian economies: if the stock of zombie debt had been subjected to hazard and recovery rates typical of firms with sustained distress, almost 23 percent of their debt would have been written off or restructured as of 2025.
- That 23 percent of zombie debt amounts to more than 5 percent of total corporate debt and is likely a source of future earnings and balance sheet weakness for the banks involved.
- Estimation assumptions and parameters reported:
  - Annual default hazard used in the central calibration: 9 percent—the exit rate of comparably distressed firms absent evergreening, calibrated by Faria-e-Castro, Paul, and Sánchez (2024).
  - Creditor recovery upon default: 35 percent based on Kermani and Ma (2023).
  - Counterfactual of debt resolution for a zombie firm in its dth year: [1 – (1 – 0.09)d] × (1 – 0.35).
  - Shaded band in panel 2 indicates the range of estimates based on alternative default hazards of 5 to 15 percent.
- Figure annotations and metrics included:
  - Panel 1: Share of Credit to Non‑financial Corporates (Percent) (time series spanning years including 2008 through 2025).
  - Panel 2: Estimates of Zombie Debt Resolution in Asia (Percent).
  - Reported summary statistics: 22.7 (Percent of GDP), 5.1 (Percent of zombie firm debt or Percent of total corporate debt as annotated).

### Data sources and sample construction
- The samples for emerging markets, the global aggregate, and Asia are based on data from the Bank for International Settlements.
- Other data sources cited: Orbis; Moody’s Analytics; IMF staff calculations.
- Notes: Data from Orbis; firms defined as exiting zombie status after covering interest expenses for three consecutive years.

---


_Source: https://www.imf.org/-/media/files/publications/gfsr/2026/october/english/ch1.pdf_