## THE GLOBAL FINANCIAL SAFETY NET—A STOCKTAKING (ppea2025035)

## Source details

**Canonical URL:** [THE GLOBAL FINANCIAL SAFETY NET—A STOCKTAKING (ppea2025035)](https://www.imf.org/-/media/files/publications/pp/2025/english/ppea2025035.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/pp/2025/english/ppea2025035.pdf.md)
- [Structured JSON version](/-/media/files/publications/pp/2025/english/ppea2025035.pdf.json)

---

### Executive summary: context and core findings
- Definition and layers:
  - The Global Financial Safety Net (GFSN) consists of four layers: (i) international reserves, (ii) bilateral swap arrangements (BSAs), (iii) Regional Financing Arrangements (RFAs), and (iv) the International Monetary Fund.
- Changing global context:
  - More frequent shocks, low growth, high debt, trade tensions, digitalization, geoeconomic fragmentation, and the climate transition.
  - Eroded policy space in many countries, including major economies.
- Overall performance:
  - GFSN performed reasonably well in the Global Financial Crisis (GFC) and the Covid-19 pandemic, aided by unprecedented policy responses in major economies.
  - Aggregate GFSN size is large but access is uneven and layers are not fully substitutable across countries.
  - Coordination across layers remains largely informal and ad hoc.

### GFSN layers — roles, strengths, and limitations
- International reserves:
  - First line of defense; largest global shock absorber (reserves roughly three times the combined lending capacity of all other layers).
  - Trends and distribution:
    - EMDEs account for 60 percent of global output (PPP terms) and large EMs hold 42 percent of global reserves.
    - Reserves peaked at 7.7 percent of global external liabilities (2005–2012); now around 6 percent of global external liabilities.
    - LICs hold 1 percent of global reserves while accounting for 2 percent of world economy.
    - About 40 percent of LICs (24 countries) have reserves covering less than 3 months of imports.
  - Cost of holding reserves:
    - For EMs, average cost estimated at about 140 basis points, or roughly 0.4 percent of GDP annually (2012–2024).
    - For LICs with market access, cost estimated 400–600 basis points higher than for EMs.
  - Reserve composition:
    - Gold’s share rose to nearly 18 percent at end-2024; central banks accounted for just under a quarter of total gold purchases since 2022.
- Bilateral swap arrangements (BSAs):
  - Two main categories: (i) reserves-creating swaps among systemic AEs to ease funding markets; (ii) swaps (e.g., PBoC) often used to finance BoP needs and facilitate trade/internationalization of currencies.
  - Fed maintains unlimited standing swap lines with five central banks; temporary expansions used during GFC and Covid-19.
  - Predictability and transparency vary widely; some swaps are predictable with clear terms, others are time-limited, opaque, or conditional.
  - Example cost for Fed swaps: OIS rate plus 25 basis points.
- Central bank standing repo facilities:
  - FIMA and Eurosystem Repo Facilities allow temporary exchange of high-quality sovereign assets for dollars/euros.
  - FIMA usage peaked at US$60 billion in March 2023; facilities do not create new reserves but support reserve liquidity.
- Regional Financing Arrangements (RFAs):
  - RFAs pool member resources regionally; network expanded but capacity concentrated in Europe and Asia.
  - Collective lending capacity as of June 2025: about US$1.3 trillion.
  - Selected RFA indicators (US$ billions):
    - AMF — Lending Capacity: 4.1; Credit Outstanding: 2.2; Peak Yearly Disbursement: 0.9
    - FLAR — Lending Capacity: 10.1; Credit Outstanding: 0.3; Peak Yearly Disbursement: 1.8
    - EFSD — Lending Capacity: 10.0; Credit Outstanding: 1.9; Peak Yearly Disbursement: 1.2
    - CMIM — Lending Capacity: 240.0; Credit Outstanding: -; Peak Yearly Disbursement: -
    - EU BoP Assistance Facility — Lending Capacity: 58.6; Credit Outstanding: 0.0; Peak Yearly Disbursement: 10.4
    - EFSM — Lending Capacity: 70.3; Credit Outstanding: 43.6; Peak Yearly Disbursement: 36.2
    - ESM — Lending Capacity: 785.1; Credit Outstanding: 288.5; Peak Yearly Disbursement: 215.5
    - BRICS CRA — Lending Capacity: 84.5; Credit Outstanding: -; Peak Yearly Disbursement: -
  - RFAs are rarely used (except ESM/EFSF) and may face limits if systemic/regional shocks affect many members.
- International Monetary Fund:
  - Central role due to near-universal membership and global mandate.
  - Resource framework:
    - 16th General Review of Quotas completed in 2023; once implemented, will result in a 50 percent equiproportional increase in quotas and an equivalent rollback on the NAB and phasing out of BBAs.
    - Will maintain IMF’s lending capacity at about US$1 trillion.
  - Recent activity and toolkit:
    - Approved over US$435 billion to 97 countries since the pandemic; commitments standing at over US$250 billion at end-2024.
    - Commitments and disbursements (2020–24): financing commitments about US$386 billion; total disbursements US$160 billion.
    - Tripled PRGT lending to poor countries; created Resilience and Sustainability Trust (RST) in 2022.
  - Constraints:
    - Stigma and reluctance to seek IMF support can delay engagement; 115 out of 142 surveyed countries reported using SDR allocation to increase reserves (IMF, 2023).

### Transformative forces and forward risks
- Transformative forces:
  - Digitalization increases speed and intensity of shock transmission; introduces new contagion channels (digital currencies, pseudo-dollarization); makes rapid liquidity layers more critical.
  - Geoeconomic fragmentation may make access to BSAs and other layers depend on strategic alignment; potential erosion of international risk sharing.
  - Climate shocks increase frequency/severity of natural disasters, raising external financing needs and trade-offs between reserves and resilience investment.
- Illustrative staff simulations:
  - A sharp reversal in capital flows could generate demand for GFSN resources roughly 50 percent higher than peak annual usage during the GFC (staff illustrative simulation).
  - Digitalization in finance scenario produced the narrowest margin of aggregate resources over demand.
  - Over half of LICs may lack sufficient coverage to meet external financing needs in certain scenarios.

### Stylized scenario and network simulation methodology (high-level parameters)
- Scenarios:
  - Four scenarios: historical (central) and three focused on digitalization in finance, geoeconomic fragmentation, and climate shocks.
- Country selection:
  - Stage 1: 190 countries considered; combined score = geometric average of network-centrality and scenario-specific vulnerability; countries above the 65th percentile selected.
- Shock design:
  - Stage 2 shocks EFNs using tail assumptions: reduced debt rollover (10 percent), FDI and deposit outflows drawn from the 90th percentile of country-specific historical distribution; portfolio flows drop to zero.
  - Fiscal consolidation assumed equal to 0.7 percent of GDP in Stage 2 (subtracting from shocked EFNs).
- Network model:
  - Stage 3 assumes defaults by selected countries triggering spillovers; iterative default cascade until no further defaults.
  - Reserve usage baseline: available reserves defined as 80 percent of affected countries’ reserves.
  - Conservative sensitivity: reserves-use assumption set to 50 percent (more conservative), leading to sharply higher unmet demand.
  - IMF resource cap in baseline: 200 percent of quota.
  - Default contagion threshold: countries default if meeting increased EFNs would take reserves below 80 percent of their ARA metric.
- Key model policy response parameters:
  - Fiscal consolidation in network model: 1 percent of GDP as lower imports.
  - Exchange rate depreciation: top 25th percentile direct exposure depreciates by 10%; next 25th percentile by 5%; possible additional 5% if needed.
  - Elasticity of current account to REER: -0.2 percent of GDP per unit change.

### Simulation results: aggregate and distributional adequacy
- Aggregate adequacy:
  - Aggregate GFSN resources exceed total demand under the historical and transformative-force scenarios in the baseline (80 percent reserve usage) but non-fungibility matters.
  - Narrowest aggregate margin under the digitalization in finance scenario.
- Country-level distribution:
  - EMs: highest estimated demand across all scenarios; account for most unmet demand across scenarios. Demand peaks under the geoeconomic fragmentation scenario; historical scenario generates the largest unmet demand in some decompositions.
  - AEs: moderate demand; reserve-issuing AEs (USA, euro area, Japan, UK, China) excluded from EFN amounts as they are assumed to absorb shocks due to quasi-unlimited access; AEs’ demand peaks under the digitalization scenario.
  - LICs: low nominal demand but large unmet demand relative to available coverage; IMF covers a median of 7 percent of LICs’ external liabilities compared to a median of 3 percent across remaining income groups.
- Sensitivity to reserve use:
  - Baseline uses 80 percent of reserves; under a 50 percent reserve-use assumption, unmet demand rises sharply, increasing dependence on external layers and policy tightening costs.

### Identified core weaknesses of the GFSN
- Uneven access:
  - A small group of reserve currency issuers and those with unlimited standing swap lines enjoy quasi-unlimited protection; reserves concentrated in a few countries; RFAs provide limited coverage.
- Limited predictability:
  - Few layers beyond reserves provide automatic or pre-qualified access; uncertainty encourages excess reserve accumulation.
- Insufficient crisis prevention:
  - Outside the IMF, few layers provide strong incentives for sound macroeconomic policies or early-warning mechanisms.
- Multilayered structure and weak coordination:
  - Coordination across layers is informal and ad hoc, undermining coherence in systemic crises.

### Policy recommendations and trade-offs
- Enhance access:
  - Encourage countries lacking sufficient reserves to rebuild them via sound macroeconomic policies.
  - Explore options to improve utilization and transparency of BSAs and expand RFA coverage, especially in underserved regions.
  - Expected impact: more balanced access, strengthened resilience, reduced incentives for excessive self-insurance.
  - Risks: central bank mandates and risk aversion may limit BSA reforms; expanding RFAs requires funding and political support.
- Improve predictability and speed:
  - Explore crisis management protocols across layers, pre-qualified and contingent instruments, and strengthen IMF precautionary toolkit.
  - Expected impact: reduced excess reserve accumulation, lower crisis costs, bolstered market confidence.
  - Risks: more automatic access may weaken policy incentives unless conditioned on sound ex-ante policies and safeguards.
- Strengthen crisis prevention:
  - Enhance IMF monitoring of risks and vulnerabilities, deepen analysis of spillovers, close data gaps, and better account for transformative forces.
  - Explore stronger links between IMF surveillance and BSA/RFA policy incentives.
  - Expected impact: improved ex-ante vulnerability reduction and crisis prevention.
  - Risks: stronger ex-ante conditions may discourage timely external support or prompt less efficient self-insurance.
- Enhance coordination:
  - IMF to play a greater convening and coordinating role; explore deepened IMF–RFA collaboration and co-financing without compromising institutional autonomy.
  - Expected impact: more efficient and effective crisis responses.
  - Risks: legal and political constraints may limit formal mechanisms; informal, case-by-case cooperation may persist.

### Issues for discussion (as posed by the chapter)
- Agreement on diagnosis:
  - GFSN performed reasonably well during recent crises but faces greater strains ahead given depleted policy buffers and a challenging global landscape.
  - Uneven access and multilayered GFSN structure complicate effective crisis response.
- Critical areas for strengthening the GFSN in general.
- Fund’s role:
  - Further strengthen Fund monitoring of risks and vulnerabilities.
  - Explore options to enhance the effectiveness of precautionary facilities.
  - Strengthen coordination with RFAs and BSAs.

*Source: THE GLOBAL FINANCIAL SAFETY NET—A STOCKTAKING (ppea2025035), International Monetary Fund.*

### EXECUTIVE SUMMARY

### EXECUTIVE SUMMARY

### A stocktaking in a changing global context
- The Global Financial Safety Net (GFSN) is a set of institutions and mechanisms that provide countries with insurance against shocks and crises. It offers financing when crises occur and complements countries’ policy responses in cushioning the impact of shocks.
- The GFSN comprises four layers: (i) international reserves, (ii) bilateral swap arrangements (BSAs), (iii) Regional Financing Arrangements (RFAs), and (iv) the International Monetary Fund.
- The context for this stocktaking is a global environment characterized by:
  - more frequent shocks;
  - an economic outlook marked by low growth, high debt, and trade tensions;
  - transformative forces—digitalization, geoeconomic fragmentation, and climate transition—reshaping the global landscape;
  - eroded policy space, including in major economies, which could constrain policy responses to future crises.
- Although the GFSN is sizeable in aggregate, access is not evenly distributed, and its layers are not substitutable across countries. The paper’s purpose is to take stock of the GFSN’s effectiveness and identify areas for improvement through deeper analysis.

### GFSN performance in recent years
- The GFSN performed reasonably well during systemic crises, notably the Global Financial Crisis (GFC) and the Covid-19 pandemic, helping to cushion their impact and preventing broader economic disruptions.
- Unprecedented policy responses by major economies—including large-scale fiscal stimulus and extraordinary monetary easing—created significant positive spillovers that counteracted strains in the international financial system and softened the demand for the GFSN during the GFC and Covid-19 pandemic.
- Experience with GFSN layers:
  - International reserves:
    - remain a first line of defense and the largest global shock absorber.
    - reserve adequacy is highly uneven: some countries, especially a number of large emerging markets (EMs), have accumulated excess reserves (costly domestically and globally and contributing to global imbalances), while many other countries lack adequate reserves to manage shocks.
  - Bilateral swap arrangements (BSAs):
    - have alleviated stress in global markets during systemic crises.
    - the unlimited swap lines between the U.S. Federal Reserve (Fed) and five major central banks played a critical role during the GFC and the Covid-19 pandemic in preventing the disruption of market funding in key currencies.
    - BSAs have been less proven as a tool for providing balance of payments (BoP) support, and many lack transparency.
  - Regional Financing Arrangements (RFAs):
    - pool member resources and offer an additional layer of liquidity support.
    - remain rarely used (except during the euro-area crisis), are unevenly distributed across regions and countries, have limited policy content, and leave gaps in providing timely assistance.
  - The IMF:
    - has continued to play a central role in the GFSN, given its global mandate to safeguard the stability of the International Monetary System (IMS).
    - offers the widest form of international risk sharing thanks to its near-universal membership and has provided countries with policy advice, capacity development, and financing when needed.
    - for many countries, the IMF remains the lender of last resort, although concerns about stigma and policy conditionality may have discouraged demand for its lending in some instances.
- Overall observations:
  - The GFSN has remained heavily reliant on non-pooled resources, with uneven country access to different layers.
  - Some emerging market and developing economies (EMDEs)—particularly low-income countries (LICs)—have relied primarily on the IMF.
  - Beyond the IMF, the GFSN has had a varying degree of effectiveness in incentivizing sound macroeconomic policies and addressing vulnerabilities to prevent crises.
  - Coordination across the GFSN layers remains largely informal and ad hoc, which constrains its overall effectiveness.

### Challenges posed by transformative forces
- In a context of frequent shocks and depleted policy buffers, the GFSN could come under strain, especially during systemic crises.
  - Staff’s illustrative simulations suggest that a sharp reversal in capital flows could generate demand for GFSN resources that would be roughly 50 percent higher than the peak annual usage during the GFC.
  - This demand could increase further if the erosion of policy buffers limits the capacity of major economies to deploy stimulus on the scale seen in recent crises.
- Transformative forces and their implications:
  - Digitalization in finance accelerates the speed at which shocks can propagate and would thus require quicker deployment of resources during crises.
  - Geoeconomic fragmentation could hinder access to certain GFSN layers, prompting countries into further reserve accumulation.
  - Climate-related shocks could accentuate the trade-offs between financial buffers and investments to build resilience.
- Illustrative simulations:
  - A systemic BoP shock affecting the countries most vulnerable to transformative forces could generate substantial demand for GFSN resources.
  - Among country groups, EMs experience the largest aggregate shortfall in available GFSN resources under most scenarios given their elevated external vulnerabilities.
  - Over half of LICs may lack sufficient coverage to meet their external financing needs, underscoring uneven access to GFSN layers.

### Possible areas for strengthening the GFSN
- The paper highlights core GFSN weaknesses: uneven access, lack of predictability, limited crisis prevention, and poor coordination, and explores options to address them.
  - Enhance access:
    - Countries lacking sufficient reserves could be encouraged to rebuild them through sound macroeconomic policies.
    - Explore options to improve the utilization of BSAs and expand RFA country coverage.
  - Improve predictability:
    - Different GFSN layers could consider reforms to make access to their resources more reliable during crises.
    - Precautionary instruments and other pre-qualified financing arrangements can play an important role, particularly in improving the speed at which financing can be accessed.
  - Enhance crisis prevention:
    - Countries, especially systemic economies, could be further incentivized to pursue strong policies and maintain adequate buffers, supporting crisis prevention and mitigating spillovers.
    - The IMF could further strengthen its monitoring of risks and vulnerabilities.
    - There is scope for BSAs and RFAs to strengthen incentives for sound policies, including through closer links to IMF surveillance and programs.
- Strengthen strategic coordination:
  - To mitigate the multilayered structure, strategic coordination across the GFSN layers could be strengthened.
  - The IMF is well-positioned to lead this effort and could convene partners and help coordinate crisis prevention as well as crisis responses across the layers.
- Expected outcome:
  - Progress in these areas would improve the effectiveness of the GFSN and strengthen its adequacy in the face of a shock-prone and uncertain global economy shaped by transformative forces, delivering important benefits for the stability of the IMS.

*EXECUTIVE SUMMARY — THE GLOBAL FINANCIAL SAFETY NET—A STOCKTAKING, International Monetary Fund.*

### 8.      EMDEs have become increasingly important to the world economy, gaining a larger

### 8.      EMDEs have become increasingly important to the world economy, gaining a larger share of global GDP, trade, and investment

### EMDEs’ evolving role and key statistics
- EMDEs now account for 60 percent of global output in PPP terms.
- EMDEs account for 40 percent of global trade.
- EMDEs have a growing share of inward and outward FDI, with China singled out as a transformative force—emerging as the world's second-largest economy and a central node in international trade and manufacturing.
- Other EMs noted for rising prominence include India (an engine of post-pandemic global growth) and Brazil (a major destination for FDI).
- LICs experienced steady but modest growth; some LICs graduated to EM status while others saw little or no increase in per capita income.
- Improved macroeconomic management in many EMDEs—more flexible exchange rates and higher reserve buffers—has strengthened resilience during recent shocks.

### Global financial structure and currency dominance
- Advanced economies (AEs) continue to dominate global finance; the U.S. dollar remains the preeminent international currency.
- The dollar accounts for the largest share of global reserves, trade invoicing, and financial transactions, giving U.S. economic and financial policies outsized influence over the IMS.
- Monetary policy tightening by the Federal Reserve has historically triggered capital outflows, currency depreciations, and higher borrowing costs in EMDEs; Fed actions to support U.S. dollar liquidity have also been instrumental during systemic crises.
- ECB and Bank of Japan actions have materially shaped global liquidity and asset prices.
- Despite structural changes, the IMS has remained centered on the U.S. dollar, though emerging economic, technological, and geopolitical trends could reshape the system.

### Outlook, risks, and implications for the Global Financial Safety Net (GFSN)
- The global landscape is characterized by elevated uncertainty, major shifts in trade and other policies, ongoing conflicts, and rising domestic instability in some countries.
- Policymakers face depleted policy buffers and historically high debt levels after extensive policy support deployed in response to successive shocks.
- Potential growth has slowed across most countries due to flagging productivity, sluggish investment, and demographic headwinds.
- The combination of low growth and low buffers increases trade-offs between macroeconomic adjustment and financing in future crises:
  - Many LICs already contend with high debt service burdens and declining aid flows.
  - AEs and some EMs face growing fiscal demands—from population aging to rising national security needs.
  - For countries with elevated fiscal vulnerabilities, future shocks may leave little room for borrowing, increasing the need for timely, credible, and sustained policy adjustment calibrated to protect investment and preserve social cohesion.
  - Structural reforms that unlock private sector–led growth are essential to mitigate adverse effects of adjustment.
- Reduced policy space in major economies could limit their ability to respond to future crises, placing additional strain on the GFSN:
  - If systemic countries cannot deploy fast and large-scale fiscal and monetary stimulus as in past crises, shocks could propagate more rapidly and intensify.
  - Delays or withholding of swap line activation by major central banks would increase pressure on other GFSN layers, necessitating greater resources and swifter action.
- Transformative forces—digitalization, geoeconomic fragmentation, and the climate transition—are expected to increase demands on the GFSN, including its size, coverage, and the predictability, speed, and affordability of access to resources.

### GFSN layers — International reserves (features, evolution, and coverage)
- International reserves are the most reliable form of self-insurance, offering immediate and unconditional liquidity under national authorities’ direct control.
- Reserves constitute the largest layer of the GFSN—amounting to roughly three times the combined lending capacity of all other layers.
- Functions of reserves:
  - Rapidly deployed to meet balance-of-payments needs and provide foreign currency liquidity for market funding and financial sector support.
  - Critical signaling role: demonstrate ability to meet short-term external obligations, improve market perception, lower sovereign risk, and improve access to financing.
- Reserve accumulation trends:
  - Between 2005 and 2012, global reserves tripled, peaking at 7.7 percent of global external liabilities.
  - Since then, reserve growth has slowed; reserves now stand at around 6 percent of global external liabilities.
  - In U.S. dollar terms, the total stock of reserves has continued to increase despite moderation in accumulation rates.
- Distribution and adequacy:
  - Large EMs account for 42 percent of global reserves—considerably higher than their share in the world economy.
  - LICs hold just 1 percent of global reserves despite accounting for 2 percent of the world economy.
  - About 40 percent of LICs (24 countries) have reserves covering less than 3 months of imports, leaving them vulnerable to external shocks and reliant on other GFSN layers.
  - Reserve adequacy (IMF ARA metric) tends to be higher for large EMs.
- Reserve composition:
  - Reserves continue to be held primarily in traditional reserve currencies.
  - The share of gold in global reserves has gradually increased in recent years.

### Gold and reserve composition
- Gold has become the second-largest component of central bank reserves after the U.S. dollar.
- Gold’s share in global reserves rose to nearly 18 percent at end-2024, largely driven by the sharp rise in gold prices.
- Central banks stepped up net gold purchases since 2022; on a gross basis, central banks accounted for just under a quarter of total gold purchases during this period—more than double their 10 percent market share in 2010–21.
- Despite large purchases, net additions to official reserves have been modest, as some central banks appear to be holding newly acquired gold outside their reserves.
- Constraints on gold’s role as a core reserve asset include liquidity considerations, restricted transactional utility, volatility, storage and security challenges, and potential market impact during sales.

### Cost of holding reserves
- Holding reserves entails an opportunity cost because reserve assets must be highly liquid and are typically held in short-term government securities that yield low returns.
- Aggregate and group-level cost estimates:
  - For EMs, the average cost of holding reserves is estimated at about 140 basis points, or roughly 0.4 percent of GDP annually, between 2012 and 2024.
  - For LICs with market access, the cost of holding reserves is estimated to be 400–600 basis points higher than for EMs—primarily due to LICs’ elevated sovereign risk and higher funding costs.
- Broader effects:
  - Excessive reserve accumulation can be inefficient globally, exacerbating the shortage of safe assets and contributing to macroeconomic imbalances.
  - Demand from large reserve holders can compress yields and increase the cost borne by smaller holders, such as LICs.
- Alternative measures and mitigation:
  - A shadow-liquidity premium approach suggests diversified mixes (ultra‑short supranational bonds, agency paper, modest equity) could deliver 50–150 basis points higher returns than Treasury bills at comparable daily volatility.
  - Stronger policy frameworks and improved macroeconomic fundamentals can help lower the cost of holding reserves and unlock alternatives to self-insurance.

*Source: THE GLOBAL FINANCIAL SAFETY NET—A STOCKTAKING, International Monetary Fund*

### 19.      Beyond their cost, the utility of reserves as a liquidity buffer may be limited by market

### 19.      Beyond their cost, the utility of reserves as a liquidity buffer may be limited by market perceptions of their large-scale use

### Limits to reserves as a liquidity buffer
- Some EMs hesitate to draw down reserves significantly during times of stress, fearing it could signal distress and weaken investor confidence.
- Empirical examples:
  - During the GFC, most EMs refrained from deploying more than a quarter of their reserves (Aizenman and Sun, 2012).
  - Only 6 out of 87 EMs reduced their reserves by more than 20 percent during the Covid-19 pandemic in 2020.

### Bilateral Swap Arrangements (BSAs): purpose and evolution
- BSAs are contingent agreements between central banks to exchange currencies, primarily during times of market stress.
- The network of BSAs expanded significantly after the GFC, though access remains limited to a relatively narrow group of countries.
- Two broad categories of BSAs:
  - Arrangements among systemic AEs designed to ease pressures in funding markets and prevent contagion to BSA-providing countries.
  - Arrangements established for other purposes, such as to directly finance BoP needs or to facilitate bilateral trade and investment.

### BSAs focused on funding markets (role and features)
- BSAs aimed at easing funding market pressures in AEs have become a critical component of the GFSN in containing systemic crises.
- Primary goal: ensure major currency funding markets continue to operate during crises; not designed to provide BoP support and generally unavailable to smaller AEs and most EMDEs.
- The largest network of BSAs (by size) is maintained by the Fed:
  - During the GFC, the Fed established swap lines with five major AE central banks: the ECB, the Bank of England, the Bank of Japan, the Swiss National Bank, and the Bank of Canada.
  - As the crisis intensified, the Fed extended temporary BSAs to central banks of other AEs and large EMs.
  - When the Covid-19 pandemic struck, the Fed reactivated its extended swap line network.
  - Temporary BSAs were allowed to expire as market conditions normalized, but the Fed maintains unlimited standing swap lines with the five original central banks.
- Footnote example: temporarily expanded access during the GFC to central banks of Australia, Brazil, Denmark, Mexico, Korea, New Zealand, Norway, Singapore, and Sweden.

### BSAs outside reserve-currency AEs (China, India) — different objectives
- The People’s Bank of China (PBoC) has established swap lines with over 40 countries (McDowell, 2019).
  - These arrangements differ from Fed/ECB BSAs: they are not primarily aimed at stabilizing currency funding markets but have become a crucial source of financing for countries facing BoP gaps and liquidity shortages.
  - Increasingly used to facilitate trade and investment with China, including settlements in renminbi, as part of China’s broader strategy to internationalize its currency.
- The Reserve Bank of India has developed swap lines with several countries in the South Asian Association for Regional Cooperation to further enhance regional financial stability.

### Variations in predictability, terms, and transparency of BSAs
- BSAs rely on central banks’ consent and remain subject to domestic considerations.
- BSAs between reserve currency-issuing AEs generally offer predictable access and clear terms, with limited restrictions on usage and duration, and relatively low costs.
  - Example cost (Fed’s swap lines): central bank borrowing dollars from the Fed pays interest at OIS rate plus 25 basis points.
- Evidence that established BSAs provide rapid liquidity, support market functioning, and bolster confidence (Perks and others, 2021; Bahaj and Reis, 2022).
- Other swap arrangements are often less transparent, time-limited, subject to predetermined access limits, and may have more restrictive terms.
- Lack of standardized terms and limited disclosure regarding activation conditions and usage:
  - Hampers accurate assessment of countries’ reserve and external positions.
  - May hinder effectiveness, including in deterring market pressures (IMF, 2025b).

### Central bank standing repurchase (repo) facilities
- The Fed and the ECB maintain standing repurchase facilities for other central banks (the Fed’s Foreign and International Monetary Authorities (FIMA) Repo Facility and the ECB’s Eurosystem Repo Facility).
- Function: allow other central banks to temporarily exchange U.S. Treasuries (Fed) or euro-denominated debt issued by euro-area governments (ECB) for U.S. dollars or euros.
- FIMA facility:
  - A wide range of central banks and monetary authorities hold a FIMA account and are eligible to apply for the Fed’s FIMA repo facility.
  - Usage has remained limited and infrequent, peaking at US$60 billion in March 2023.
- Role relative to BSAs and reserves:
  - Aim to alleviate pressures in global dollar and euro funding markets and contain potential spillovers into U.S. and euro-area financial markets.
  - Unlike BSAs, these facilities do not expand the supply of reserves and cannot be considered an additional GFSN layer; they support functioning of countries’ own reserves by ensuring existing assets can be liquidated in stress.
  - Collateral-based liquidity facilities enable central banks to keep high-quality assets and access rapid foreign currency funding instead of selling them outright, easing the liquidity–return trade-off.
  - LICs may face operational, legal, and eligibility hurdles in using these facilities.

### Regional Financing Arrangements (RFAs): scope and coverage
- RFAs pool financial resources among countries, typically within the same region, to provide crisis financing.
- The network of RFAs has expanded significantly over past decades and is an essential element of regional financial stability, but most resources remain concentrated in Europe and Asia.
- RFAs vary widely by mandate and scope; some include surveillance and technical assistance.
  - Examples: CMIM and ESM primarily provide crisis financial support; AMF and EFSD also extend development financing to promote long-term regional cooperation and integration.
- As of June 2025, RFAs collectively had a lending capacity of about US$1.3 trillion (Table 1).
  - Despite sizeable capacity, use has generally remained limited.
  - ESM and its predecessor, the EFSF, are notable exceptions with significant disbursements during the euro-area crisis.
  - Other large RFAs, such as CMIM and BRICS CRA, remain largely untested; their ability to sustain support during prolonged crises is uncertain.
  - RFA resources may be less reliable when systemic or regionally concentrated shocks affect a large portion of membership.
- RFAs can be a cost-effective complement to reserve accumulation by pooling resources to reduce costs of protection against uncorrelated shocks.
  - Lending rates are generally lower than market rates, though often not explicitly specified.
  - Example: In 2014, countries like Spain and Cyprus paid lending rates between 60 and 110 basis points on ESM loans, far below market rates.
  - Deterrents to usage include variable spreads across RFAs, shorter maturities, lengthy processes, and policy conditionality.

### Selected RFA lending indicators (Table 1, in US$ billions)
- AMF — Lending Capacity: 4.1; Credit Outstanding: 2.2; Peak Yearly Disbursement: 0.9
- FLAR — Lending Capacity: 10.1; Credit Outstanding: 0.3; Peak Yearly Disbursement: 1.8
- EFSD — Lending Capacity: 10.0; Credit Outstanding: 1.9; Peak Yearly Disbursement: 1.2
- CMIM 2/ — Lending Capacity: 240.0; Credit Outstanding: -; Peak Yearly Disbursement: -
- EU BoP Assistance Facility — Lending Capacity: 58.6; Credit Outstanding: 0.0; Peak Yearly Disbursement: 10.4
- EFSM — Lending Capacity: 70.3; Credit Outstanding: 43.6; Peak Yearly Disbursement: 36.2
- ESM 3/ — Lending Capacity: 785.1; Credit Outstanding: 288.5; Peak Yearly Disbursement: 215.5
- BRICS CRA 2/ — Lending Capacity: 84.5; Credit Outstanding: -; Peak Yearly Disbursement: -

Notes on Table 1 (preserved wording and dates):
- Lending capacity data refers to July 3, 2025 for ESM and EFSF; end-June 2025 for FLAR, EU BoP Assistance Facility, and EFSM; end-May 2025 for EFSD; and end-March 2025 for AMF.
- Credit outstanding data refers to July 3, 2025 for ESM and EFSF; June 2025 for AMF and EU BoP; end-May 2025 for EFSD; and end-2024 for FLAR.
- For CMIM and BRICS CRA, which are based on swap arrangements, the stated figures for lending capacity represent committed resources, with actual capacity constituting only a portion of that amount.
- ESM includes its inactive predecessor, the European Financial Stability Facility (EFSF). The EU BoP Assistance Facility and EFSM are administered by the European Commission (EC), while the ESM has a separate governing body.

### The International Monetary Fund (IMF): role, resources, and adaptation
- The IMF’s unique capacity for global risk-sharing and crisis prevention places it at the center of the GFSN.
  - Near-universal membership, inclusive governance, and mandate for macroeconomic surveillance differentiate it from other GFSN layers.
  - Through assessments of its 191 member economies, the IMF works to prevent crises by identifying vulnerabilities and providing policy advice.
  - Provides financial support to countries undertaking macroeconomic adjustment to shocks, backed by policy conditionality and designed to catalyze additional financing.
  - Helps build resilience through capacity development that strengthens institutions and policy frameworks.
- Resource adequacy and reform:
  - The IMF relies on members’ quotas as its permanent and reliable source of financing and undertakes periodic reviews to ensure adequacy.
  - Quotas are supplemented when needed by temporary borrowing through the New Arrangements to Borrow (NAB) and Bilateral Borrowing Arrangements (BBAs).
  - The 16th General Review of Quotas was completed in 2023 and, once implemented, will result in a 50 percent equiproportional increase in quotas and an equivalent rollback on the NAB and phasing out of BBAs.
  - This will restore quotas as the primary source of IMF funding while maintaining the IMF’s lending capacity at about US$1 trillion.
- Evolution of IMF lending toolkit and recent activity:
  - Post-GFC innovations include the Flexible Credit Line and Precautionary and Liquidity Line; in 2020 launched the Short-term Liquidity Line.
  - IMF precautionary facilities function as liquidity insurance, enhancing preventive capacity and reducing likelihood of costly bailouts.
  - Since the pandemic, the IMF scaled up lending significantly:
    - Approved over US$435 billion to 97 countries since the pandemic (with commitments standing at over US$250 billion at end-2024).
    - Tripled lending to poor countries under the Poverty Reduction and Growth Trust (PRGT).
  - In 2024, the IMF increased overall annual and cumulative access limits to its general resources by 38 percent, reduced borrowing costs, and approved reforms to preserve interest-free loans for the poorest countries.
  - Created the Resilience and Sustainability Trust (RST) in 2022 to help countries reduce prospective BoP risks by enhancing resilience to climate change and pandemics.

### Stigma and reluctance to seek IMF support
- Some countries hesitate to seek IMF support in a timely manner due to stigma and policy conditionality:
  - Concerns that IMF engagement could signal economic distress, weakening investor confidence and triggering capital outflows, risking a self-fulfilling crisis.
  - Domestic political costs of reforms, especially politically sensitive and socially costly reforms, deter timely engagement.
  - Some large EMs may view IMF financing as inconsistent with efforts to project economic self-reliance and regional influence.
- Consequences:
  - Countries may delay requesting IMF financial support or turn to other GFSN layers, even when alternatives are more costly or offer weaker incentives for sound policy adjustment (di Mauro and Zettelmeyer, 2017).
  - Timely engagement and realistic and sustained adjustment are key drivers of program success and could help mitigate stigma over time.

### SDR allocations and reserve supply
- The IMF has occasionally enacted SDR allocations to boost the supply of global reserves.
  - SDR allocations make the SDR unconditionally available to countries, boosting their reserves.
  - A total of SDR 660.7 billion (equivalent to about US$943 billion) has been allocated to date, with the last two allocations implemented during the GFC in 2009 and the Covid-19 pandemic in 2021.
  - SDR allocations are made in proportion to countries’ quota shares at the IMF, meaning EMDEs received about one-third of the 2021 allocation.
  - Some countries channeled their SDRs to the PRGT and RST to help the Fund support vulnerable low and middle-income countries.
  - Survey evidence: Most countries (115 out of 142 surveyed), including a large majority of EMs (61 out of 72), reported using the SDR allocation to increase reserve levels (IMF, 2023).
  - LICs reported a wider variety in usage, with a majority (33 out of 53) at least partly using their allocation for fiscal support, including for pandemic-related spending.

### Features of GFSN layers and complementarities
- Distinct strengths and weaknesses across layers determine use and timing:
  - Reserves: highly predictable and immediately accessible but costly to hold.
  - BSAs (other than between reserve currency-issuing AEs): less predictable and can be relatively costly.
  - RFAs: rarely used; questions about speed of resource deployment.
  - IMF lending: universally available and provides strong policy content but may give rise to stigma.
- Given the mix of strengths, the layers can complement each other in responding to crises when coordinated effectively.

*Source: IMF staff calculations and analysis in "THE GLOBAL FINANCIAL SAFETY NET—A STOCKTAKING" (excerpt provided).*

### 34.      Quick and unconditional access to some GFSN layers— such as reserves or certain

### 34.      Quick and unconditional access to some GFSN layers— such as reserves or certain

### Tension between rapid access and policy incentives
- Quick and predictable access to GFSN layers (e.g., reserves, certain swap lines) can reassure markets at shock onset and limit the risk of a generalized loss of confidence.
- Unconditional access can weaken countries’ incentives to address macroeconomic imbalances.
- Tools with strong conditionality and qualification criteria (such as IMF arrangements) promote policy discipline but may be less accessible or carry stigma, which can cause countries to delay seeking support and potentially worsen crises.
- Ensuring predictable and rapid access to liquidity while preserving strong incentives for sound policies creates an inherent tension in GFSN design.

### Reserve-creating versus reserve-channeling GFSN layers
- When activated, BSAs provided by reserve currency issuers (such as the Fed) temporarily create new reserves for recipient central banks, boosting the global supply of reserve liquidity.
- The IMF expands global reserves when it lends to members or implements an SDR allocation.
- BSAs between non-reserve-issuing countries and RFAs channel existing reserves between countries and among members without creating new reserves.

### Country-level coverage and disparity
- A small group of countries enjoy quasi-unlimited access by virtue of reserve currency issuance or standing, unlimited swap lines: the United States, China, the United Kingdom, Japan, the euro area; Canada and Switzerland are added due to standing, unlimited Fed swap lines.
- The ability to issue reserve currencies provides significantly greater policy space (including enhanced capacity for countercyclical fiscal stimulus and central bank balance sheet expansion) but can contribute to policy complacency.
- For non-reserve-currency countries, GFSN coverage is highly uneven:
  - Estimates measure reserves, BSA access, RFA access, and IMF borrowing capacity relative to external liabilities.
  - Assumptions: countries prepared to draw down only part of reserves (shown as above/below horizontal axis); RFA and BSA access based on access limits; IMF access illustrated in the 200–600 percent-of-quota range (normal annual and cumulative access to GRA and PRGT resources).
  - Median GFSN access across all countries is 17 percent of external liabilities; half the sample has access either below 10 percent or above 30 percent.
  - Median access within country groups is highest for LICs and lowest for AEs, largely reflecting LICs’ significantly lower external liabilities.
- Assumed reserve-use policy for the exercise: countries willing to use up to 50 percent of their reserves provided reserves do not fall below 50 percent of the ARA metric (where available) or 1.5 months of imports.

### Composition of protection for EMDEs and LICs
- Self-insurance (reserves) and potential recourse to IMF lending remain the main forms of GFSN protection for the vast majority of EMDEs, especially LICs.
- AEs and large EMs typically have more diversified liquidity sources (RFAs, BSAs, sizeable reserves).
- LICs are typically more reliant on IMF financing due to insufficient reserves and limited BSA/RFA coverage.
- For LICs, the IMF covers a median of 7 percent of their external liabilities compared to a median of 3 percent across the remaining income groups.

### Performance through recent crises (empirical patterns)
- Covid-19 pandemic and subsequent shocks prompted unprecedented policy responses that softened demands on the GFSN:
  - Fed asset purchases reached 21 percent of U.S. GDP (US$4.5 trillion).
  - ECB asset purchases reached 17 percent of euro-area GDP (US$2.3 trillion).
  - Global public debt approached 100 percent of GDP by 2021.
- GFSN usage during the pandemic and recent crises:
  - BSAs: Fed swap usage peaked at over US$400 billion in spring 2020; outstanding Fed BSA balances declined by half by summer 2020 and were almost entirely unwound by October 2020.
  - Other swap lines (ECB, China, India) also saw increased usage; ECB central bank liquidity lines peaked in September 2021 at around EUR3.5 billion.
  - International reserves: 2022 saw the largest year-on-year decline in global reserves in over two decades, driven by capital outflows and interventions amid dollar appreciation.
  - RFAs: Limited usage during the pandemic; AMF, EFSD, and FLAR approved small loans in 2020–21 and limited lending continued 2021–23.
  - IMF Lending: From 2020 to 2023, 83 countries used emergency lending and 96 countries entered program arrangements (35 with concessional lending). Financing commitments (including precautionary arrangements) totaled about US$386 billion in 2020–24, with total disbursements reaching US$160 billion.
- Historical pattern: Fed swap lines were widely used in the GFC, peaking at over US$500 billion; reserves drawdowns were smaller in the GFC than in 2015–16; coordination across GFSN layers has been largely ad hoc with no structured mechanism to align actions.

### Forward-looking risks and implications of transformative forces
- Transformative forces (digitalization in finance, geoeconomic fragmentation, climate shocks) are likely to increase the risk, speed, scale, and complexity of systemic shocks, posing challenges for GFSN responsiveness.
- Digitalization and AI in finance:
  - Can increase intensity and accelerate transmission of financial shocks.
  - Rise of digital currencies introduces new contagion channels (e.g., pseudo-dollarization).
  - Nonbanks, holding nearly half of global financial assets, can amplify capital flow volatility and reduce effectiveness of capital controls.
  - Implication: Layers that provide rapid liquidity (reserves and BSAs) become increasingly critical; adoption of alternative reserve assets could affect reserve reliability; financial technologies can also enable faster, more targeted crisis responses.
- Geoeconomic fragmentation:
  - Tariffs, industrial policies, sanctions, and national security measures could fragment trade and financial linkages and hamper coordination during crises.
  - Swap line availability could become more dependent on strategic alignment than economic factors.
  - Potential outcome: reduced international risk sharing and weaker coordinated policy spillovers; erosion of trust in global institutions over time.
- Climate shocks:
  - Increasing frequency/severity of natural disasters can trigger capital flow reversals, currency pressures, and strain public finances via emergency spending and reconstruction.
  - Repeated shocks raise likelihood of countries turning to the GFSN for support and create an accentuated trade-off between holding reserves for self-insurance and investing in long-term resilience.
- Tabled qualitative implications (summarized by layer):
  - Reserves: well suited for fast-moving shocks; incentive to accumulate reserves if other layers viewed as less reliable; trade-off with resilience investment under climate shocks.
  - BSAs: well suited for fast-moving shocks; may face pressure to expand; reliability could become uncertain under fragmentation and political considerations.
  - RFAs: may face pressure to boost capacity and respond faster; potential for greater demand to organize support at regional/political bloc level; limited coverage for climate-vulnerable countries.
  - IMF: growing role for macro-financial surveillance and policy advice; potentially greater demand for precautionary/rapid financing; uncertainty regarding mandate and resources if shocks are regional in nature; Fund has demonstrated ability to operate effectively in complex environments.

*Source: IMF staff calculations.*

### 46.      This section employs a stylized scenario analysis to illustrate the potential demand on

### ppea2025035 - 46.      This section employs a stylized scenario analysis to illustrate the potential demand on

### Scenario design and methodology
- Four scenarios produced: a central, historical scenario, and three scenarios considering transformative forces (digitalization in finance, geoeconomic fragmentation, and climate shocks).
- Each scenario applies a systemic BoP shock to a distinct group of countries deemed both most vulnerable under the scenario and systemically important in global trade and financial networks.
- Stage 1:
  - Covers 190 countries representing nearly 100 percent of global GDP.
  - Constructs a country-specific composite index combining a country’s vulnerability under each scenario with its systemic importance measured by centrality in trade and financial networks.
  - Countries above the 65th percentile of the composite index are selected.
- Stage 2:
  - For each selected country, external financing needs (EFNs) are shocked by assuming reduced debt rollover rates and FDI and deposit outflows drawn from the 90th percentile of their country-specific historical distribution.
  - Rationale: Choosing the tail of the distribution reflects a conservative approach to ensure preparedness for a severe crisis.
- Stage 3:
  - A network model estimates spillover effects through trade and financial links by assuming selected countries default on external obligations, generating second-round EFNs among partner countries.
  - Affected countries are assumed to tighten fiscal and monetary policies and draw down reserves to reduce EFNs.
  - If reserves fall below pre-defined thresholds (set based on the relationship between country risk premia and level of reserves), these countries are also assumed to default, triggering further rounds.
  - Iteration continues until no further defaults occur. Cumulative EFNs across rounds yield estimated total demand for GFSN resources.
- Country selection indicators (Table 3):
  - Historical Scenario: Estimated risk of sharp reversals of capital flows based on historical data.
  - Digitalization in Finance: Financial Development Index as proxy for exposure to digitalization-in-finance risks.
  - Geoeconomic Fragmentation: Capital flows at risk from politically distant creditors; vulnerability to tensions between major blocs; conflict metric from ICRG.
  - Climate Shocks: Vulnerability to climate change (IMF-adapted Notre-Dame Gain index score); preparedness to strengthen resilience.

### Aggregate adequacy findings
- Aggregate GFSN resources exceed total demand under the historical scenario and the scenarios illustrating transformative forces.
- Supply definition:
  - Available reserves are defined as 80 percent of affected countries’ reserves.
  - Countries’ access to external GFSN resources is included, with IMF resources capped at 200 percent of quota.
- Narrowest margin of aggregate resources over demand observed under the digitalization in finance scenario.
- Caveat: Non-fungibility of resources implies adequacy must be assessed country-by-country by comparing each affected country’s demand with resources specifically available to it.
- Reserve issuers:
  - EFNs of reserve issuers (USA, euro area, Japan, UK, China) are excluded from amounts; their spillover effects through trade and financial linkages are included.

### Country-level distribution of demand and shortfalls
- EMs (emerging markets):
  - Have the highest estimated demand across all scenarios due to significant external liabilities, elevated vulnerability, and high exposure to spillovers.
  - Demand peaks under the geoeconomic fragmentation scenario.
  - Account for most of the unmet demand across scenarios, with the historical scenario generating the largest unmet demand.
- AEs (advanced economies):
  - Exhibit more moderate demand for GFSN resources.
  - Reserve-issuing countries (all AEs except China) assumed to absorb full impact of their shocks due to quasi-unlimited access to reserves and are not presented in demand figures.
  - Demand from AEs peaks under the digitalization in finance scenario.
- LICs (low-income countries):
  - Low nominal demand across scenarios reflects small size of economies and limited trade and financial integration.
  - Limited GFSN coverage leaves LICs exposed across all four scenarios; many have sizeable unmet demand such that, even in aggregate, GFSN resources available to them would not be adequate to meet total demand.
- Country-level results indicate sizeable unmet demand in all four scenarios; unmet demand concentrated among EMs and LICs.
- Visual decomposition used in Figures 12 and 13 (described):
  - Demand covered only by reserves.
  - Demand requiring both reserves and external layers.
  - Demand on countries with insufficient coverage (portion met by GFSN).
  - Demand on countries with insufficient coverage (portion not met by GFSN).

### Sensitivity: reserves usage assumption
- Baseline assumption: countries can use 80 percent of their reserves.
- More conservative assumption: countries are assumed to be able or willing to use only 50 percent of their reserves.
  - Outcome: unmet demand for GFSN resources rises sharply under the 50 percent assumption.
  - Consequences: accentuated dependence of EMs and LICs on external GFSN layers; more countries needing external layers to meet demand; additional burden on policy tightening with attendant costs for growth and investment.

### Interpretation and implications of scenario results
- Simulations suggest most AEs and many EMs hold reserves or have access to external GFSN layers broadly sufficient to meet EFNs under transformative-force scenarios.
- For a group of EMs (including some large economies) and most LICs, available GFSN resources would not be sufficient under the historical scenario or scenarios with transformative forces.
- Although gaps may appear modest relative to overall GFSN size, they are substantial relative to resources available to individual economies.
- To close gaps, vulnerable countries need to build reserves and restore policy space via sound macroeconomic policies.
- Need to strengthen policy content of the GFSN to incentivize resilience building and crisis prevention, especially for large economies whose strong buffers and policies could prevent crisis spread and relieve pressure on other GFSN layers.

### Conclusions and policy implications (four core weaknesses and reform directions)
- Four core weaknesses identified:
  - Uneven access:
    - Small number of reserve currency-issuing countries and those with permanent unlimited swap lines enjoy quasi-unlimited protection.
    - Reserves concentrated among few countries; only limited coverage by RFAs; most EMDEs, especially LICs, rely on self-insurance and IMF support.
  - Limited predictability:
    - Few GFSN layers beyond reserves provide truly automatic or pre-qualified access, encouraging excess reserve accumulation and quasi-fiscal costs.
  - Insufficient crisis prevention:
    - Outside the IMF, few layers contain strong incentives for sound macroeconomic policies or offer mechanisms for early warning and crisis prevention.
  - Multilayered structure:
    - Coordination across GFSN layers remains largely informal and ad hoc, undermining coherence and effectiveness in systemic crises.
- Global outlook risks:
  - Depleted buffers and challenging landscape (elevated uncertainty, policy shifts, low growth, reduced policy space) could increase reliance on the GFSN.
  - Transformative forces (digitized finance, geoeconomic fragmentation, more frequent natural disasters) amplify GFSN weaknesses and require larger, more agile, better coordinated, and more predictable layers.

### Recommended reforms and expected impacts / risks
- Enhancing Access to GFSN Layers:
  - Rebuild reserves, where inadequate, through sound macroeconomic policies by countries, especially LICs and vulnerable EMs.
  - Explore options to strengthen the effectiveness of BSAs, including greater public disclosure of terms, activation criteria, and, as appropriate, usage.
  - Explore options for expanding coverage and capacity of RFAs, especially in underserved regions.
  - Expected Impact: More balanced access, strengthened resilience, reduced incentives for excessive self-insurance, greater transparency supporting confidence.
  - Risks/obstacles: Limited support for BSA reforms due to central bank mandates and risk aversion; expanding RFA coverage requires funding and political support.
- Improving Predictability and Speed:
  - Explore crisis management protocols across layers to accelerate and harmonize responses to systemic shocks.
  - Explore options to enhance predictability of access, including pre-qualified and contingent instruments.
  - Continue strengthening IMF precautionary toolkit to support countries with sound policies.
  - Expected Impact: Reduced incentive for excessive reserve accumulation, lower crisis costs, bolstered market confidence.
  - Risks/obstacles: More automatic access may weaken incentives for stronger policies unless conditioned on sound ex-ante policies and safeguards.
- Strengthening Crisis Prevention:
  - Further strengthen IMF monitoring of risks and vulnerabilities via enhanced surveillance, deeper analysis of spillovers and macrofinancial linkages, efforts to close data gaps, and improved understanding of transformative forces’ effects.
  - Explore options for BSAs and RFAs to strengthen incentives for sound policies, leveraging IMF assessments to better align policy advice and lending.
  - Expected Impact: Better monitoring and stronger policy content would incentivize vulnerability reduction and crisis prevention.
  - Risks/obstacles: Stronger ex-ante conditions may discourage countries from seeking external support or push them toward less efficient self-insurance.
- Enhancing Coordination Across GFSN Layers:
  - Strengthen coordination, with IMF playing a greater convening and coordinating role for crisis prevention and response.
  - Explore deepening IMF-RFA collaboration, including co-financing mechanisms without compromising institutional autonomy.
  - Expected Impact: Improved efficiency and effectiveness of GFSN interventions during systemic events.
  - Risks/obstacles: Legal and political constraints may limit formal structured coordination; informal, case-by-case mechanisms may remain more feasible.

### Issues for discussion
- Do Directors agree with the diagnosis:
  - The GFSN has performed reasonably well during recent crises, helping cushion impact and prevent broader disruptions.
  - Depleted policy buffers and challenging global landscape could place greater strains on the GFSN in future crises.
  - Uneven country access and the multilayered nature of the GFSN make efficient crisis response more difficult.
- Which areas are critical for strengthening the GFSN in general?
- Regarding the Fund’s role, do Directors support:
  - Further strengthening Fund monitoring of risks and vulnerabilities;
  - Exploring options to enhance effectiveness of precautionary facilities;
  - Efforts to strengthen coordination with RFAs and BSAs?

*Source: IMF staff calculations and analysis presented in the chapter excerpt.*

### Annex I. Regional Financing Arrangements: Main Features

### Annex I. Regional Financing Arrangements: Main Features

### Basic information on RFAs (Table 1 — selected entries)
- Arab Monetary Fund (AMF), 1976  
  - Mandate: Address BoP issues and support monetary cooperation.  
  - Members: Algeria, Bahrain, Comoros, Djibouti, Egypt, Iraq, Jordan, Kuwait, Lebanon, Libya, Mauritania, Morocco, Oman, Palestine, Qatar, Saudi Arabia, Somalia, Sudan, Syria, Tunisia, United Arab Emirates, and Yemen.  
  - Resources: US$3.7 bn of subscribed capital, with paid-in capital of US$3.6 bn. (end-June 2025).

- Latin American Reserve Fund (FLAR), 1978 (Andean Reserve Fund); opened to all LATAM countries in 1989  
  - Mandate: Address temporary and structural BoP imbalances.  
  - Members: Bolivia, Colombia, Costa Rica, Ecuador, Paraguay, Peru, Uruguay, and Venezuela; Chile (associate member).  
  - Resources: US$4.5 bn of subscribed capital, with paid-in capital of US$3.4 bn. (end-June 2024).

- Eurasian Fund for Stabilization and Development (EFSD), 2009  
  - Mandate: Address budget/BoP issues and support long-run development.  
  - Members: Armenia, Belarus, Kazakhstan, Kyrgyz Republic, Russian Federation, and Tajikistan.  
  - Resources: US$8.5 bn of subscribed capital, with paid-in capital of $3.1 bn. (end-May 2025).

- Chiang Mai Initiative Multilateralization (CMIM), 2010  
  - Mandate: Address liquidity issues.  
  - Members: China, Hong Kong SAR, Japan, and Korea; Brunei Darussalam, Cambodia, Indonesia, Lao PDR, Malaysia, Myanmar, Philippines, Singapore, Thailand, and Vietnam.  
  - Resources: US$240 bn of commitments, based on a multilateral currency swap arrangement. No paid-in capital.

- European Stability Mechanism (ESM), 2012 (replacing the European Financial Stability Facility (EFSF) created in 2010)  
  - Mandate: Stability support to euro area member states experiencing, or at risk of severe financial distress.  
  - Members: Euro area member countries (Austria, Belgium, Croatia, Cyprus, Estonia, Finland, France, Germany Greece, Ireland, Italy, Latvia, Lithuania, Luxembourg, Malta, Netherlands, Portugal, Slovakia, Slovenia, and Spain.)  
  - Resources: US$835.3 billion (€708.5 billion) of subscribed capital, with paid-in capital of US$94.5 billion (€81.0 billion) as of July 2025.

- BRICS Contingent Reserve Arrangement (CRA), 2014  
  - Mandate: Address liquidity issues.  
  - Members: Brazil, China, India, Russia, and South Africa.  
  - Resources: US$100 bn of commitments, based on a currency swap arrangement. No paid-in capital contribution.

- Arrangements administered by the European Commission (selected)  
  - Balance of Payments Assistance Facility, 2002 — BoP support to EU-members outside the euro area. No specific committed resources. Financed by EU borrowing, which is backed by the EU budget’s headroom.  
  - European Financial Stability Mechanism (EFSM), 2010 — BoP support to all EU-member states. No specific committed resources. Financed by EU borrowing, which is backed by the EU budget’s headroom.

### IMF–RFA collaboration (Table 2 — examples)
- Collective engagement with all RFAs:
  - Joint IMF-RFA High-Level Dialogue: Annual event since 2016, to foster exchanges among principals on the outlook and policy challenges, and promote information sharing.
  - Joint RFA Research Seminar with IMF participation: Annual event since 2016, to foster analysis and discussion of common topics of interest and promote knowledge sharing.

- IMF engagement with individual RFAs:
  - Annual high-level events: Regular IMF participation in RFA annual conferences.
  - Joint conferences and workshops: Regular events organized jointly by IMF and an RFA on common topics of interest.
  - Regular and ad-hoc exchanges: Exchanges between IMF country teams and individual RFAs on regional and individual country outlooks and macroeconomic policies, discussion of IMF policy changes.
  - Co-financing: Joint lending to expand the financing envelope and benefit from complementarities between the IMF and the RFA.
  - Enhancement of GFSN operational readiness: Joint test runs, staff exchanges and in-depth discussions of refinements to RFA frameworks.
  - Capacity development: Regular participation of RFA staff in training courses held online, at IMF headquarters and at regional TA centers.

### Simulation methodology (Annex II — overview)
- Objective: Estimate potential medium-term (MT) demand on the GFSN through indicative tail-risk scenarios that shock external financing needs (EFNs) of selected countries using stylized simulations.
- Shock design: Capital flow shocks are drawn from the 90th percentile of past distributions to estimate potential impact if trends generate extreme shocks.
- Scenarios: Four distinct scenarios calibrated by selection of countries vulnerable to three transformative forces (digitalization in finance; geoeconomic; and climate shock) plus a historical scenario where countries are selected based on historical vulnerability to capital outflows.

### Simulation stages (3-stage approach)
- Stage 1 — Selection of Countries:
  - Combined score = geometric average of: (i) Network-Centrality score (same for all scenarios); and (ii) vulnerability indicator specific to the scenario.
  - Selection rule: Countries falling in the top 65 percentile rank of the combined score are selected for a given scenario.
  - Notes on indicators:
    - Network-Centrality score: geometric average of a network score (between 0-1) and weighted count of trade and financial linkages by size and direction. Formula components: L ∈ {X, M, P, FDI, B}; α centrality parameter set equal to 0.5.
    - Financial Development Index (IMF) used to proxy exposure to new risks from digitalization in finance.
    - Geoeconomic Fragmentation Vulnerability Indicator: simple average of 3 percentile ranks based on reliance on FDI and portfolio from politically distant creditors (UN voting “ideal points” method), vulnerability to trade tensions (general equilibrium trade model), and other components.
    - Conflict metric: simple average of “civil war and coup d’etat threat” (Internal Conflict Dimension) and “Cross-Border Conflict” (External Conflict Dimension) from ICRG.
    - Climate Change Vulnerability Indicator: IMF-adapted Notre-Dame Gain Index score.
    - Sudden Stops Vulnerability Indicator: used for the historical scenario; based on Sudden Stops with Growth Impact (SSGIs).

- Stage 2 — Shock EFNs of selected countries:
  - Calculate MT gross EFNs based on World Economic Outlook (WEO) 5-year ahead forecasts.
  - Calculate MT EFNs (subtracting FDI, deposits and short-term and medium-term borrowing).
  - Shock assumptions:
    - 90th percentile of lower FDI and deposits, and 10 percent debt rollover.
    - Portfolio flows drop to zero.
  - Subtract a fiscal consolidation impact equal to 0.7 percent of GDP from the shocked EFNs.
  - Aggregate the shocked net EFNs of the selected countries.

- Stage 3 — Network model:
  - Assume selected countries default on external obligations. Spillovers emerge from direct trade and financial exposures and are amplified through changes in risk premia and asset price co-movements, triggering cascading defaults among trade and financial partners.
  - Affected countries implement fiscal consolidation, exchange rate adjustments and tap into their reserves to mitigate contagion.
  - Reserve dynamics: each country i has reserves R_i,0 at t=0. Change in reserves ∆R_i,t follows the expression given in the source (Trade revenues + Return on net foreign assets - Return on net foreign liabilities + Reserve usage).
  - Default cascade: process continues until no more defaults. Affected countries default if meeting increased EFNs would take reserves below 80 percent of their ARA metric.
  - Policy response assumptions:
    - Fiscal consolidation equal to 1 percent of GDP as lower imports.
    - Exchange rate depreciation rules: a non-failed country in the top 25th percentile of direct exposure depreciates by 10%; a country in the next 25th percentile depreciates by 5%. If needed, a country may depreciate another 5% in a subsequent period.
    - Elasticity assumed for the drop in the CA deficit in response to unit change in REER: -0.2 percent of GDP.
  - Higher interest payments: added to counterparties of defaulting countries based on higher risk premium.
  - Threshold for contagion: determined by a threshold of ARA consistent with empirical estimates; affected countries default if meeting increased EFNs would take reserves below 80 percent of their ARA metric.

### Available GFSN supply and gap assessment
- Available supply definitions in baseline comparison:
  - Available reserves defined as 80 percent of the affected country’s reserves.
  - IMF resources capped at 200 percent of quota.
  - For limited-amount BSAs, contractual limit is considered.
  - For unlimited swaps, amount is estimated based on known past usage.
  - For RFAs, their lending capacity is used.
- Conservative reserve-use simulation:
  - Availability assumed limited to preserve 50 percent of current reserves, 50 percent of ARA metric (where available), and 1.5 months of imports.
- Gap calculation:
  - Add results from Stage 2 and Stage 3 to estimate total potential demand on the GFSN from selected and affected countries; compare to available GFSN supply per country to assess gaps.

### Annex III — Brief summary of GFSN stocktaking papers (assessment and recommendations)
- 2011 — Assessment highlights:
  - Global liquidity support by reserve-issuing central banks was effective but uneven due to domestic mandates. Fed swaps are selective, with uncertain availability, and may be limited in systemic crises.
  - Extraordinary central bank support cannot be taken for granted; suited to address interbank funding pressures.
  - Fund precautionary financing helped restore market access more rapidly, but Fund resources were still tapped with a lag and not fully utilized to ring-fence countries with strong fundamentals.
  - RFAs, bilateral support, and private sector involvement important, but focused on crisis resolution, not prevention.
  - Recommendations included evenhanded and predictable liquidity provision to strong performers, minimizing systemic risks through strengthened regulation and surveillance, and greater coordination on policy spillovers and crisis response.

- 2016 — Assessment highlights:
  - GFSN serves well reserve-issuing AEs but has serious shortcomings for other borrowers and may not prevent global spillovers and contagion.
  - Borrowers’ domestic policies and shock nature may restrict available GFSN elements.
  - GFSN fails to provide adequate policy incentives, leading to reserve accumulation and fragmentation.
  - Recommendations centered on Fund reforms: provide predictable and reliable financial support, use global reserve pooling, revisit the lending toolkit (immediate liquidity support, coverage for full shock duration, prequalification, state-dependency), provide incentives for sound policies, and strengthen cooperation and monitoring.

- 2025 — Assessment highlights:
  - The GFSN held up reasonably well during systemic shocks, helped by unprecedented policy measures.
  - GFSN layers are not substitutable among countries, access remains uneven, and global coordination is largely ad hoc.
  - Few layers offer reliable access, prompting countries to accumulate excess reserves—incurring quasi-fiscal costs and contributing to global imbalances. Most EMDEs rely heavily on self-insurance and IMF support.
  - Outside the IMF, few GFSN layers provide strong incentives for sound macroeconomic policies and crisis prevention.
  - A more challenging global environment and transformative forces could increase pressure on the GFSN.
  - Recommendations: enhance access to GFSN layers, especially for EMDEs, through rebuilding reserves and exploring options to strengthen BSAs and RFAs; improve predictability and speed via triggers, crisis management, pre-qualified and contingent instruments, and strengthen the IMF precautionary toolkit; strengthen crisis prevention through monitoring of risks, stronger BSA and RSA incentives, and early warning systems; enhance coordination across GFSN layers, with the IMF playing a convening role, and further deepen IMF-RFA collaboration.

*Source: Annex I–III, "The Global Financial Safety Net—A Stocktaking" (selected annexes and tables).*

### References

### References

### Multilateralism, geopolitics, and fragmentation
- Aiyar, S., Chen, J., Ebeke, C., Garcia-Saltos, R., Gudmundsson, T., Ilyina, A., Kangur, A., Kunaratskul, T., Rodriguez, S., Ruta, M., Schulze, T., Soderberg, G., & Trevino, J., (2023), “Geoeconomic fragmentation and the future of multilateralism,” International Monetary Fund Staff Discussion Note 2023/001.
- Catalán, M. and T. Tsuruga., (2023), Geopolitics and financial fragmentation: Implications for macro-financial stability. Chapter 10 in CEPR book on "Geoeconomic Fragmentation: The Economic Risks from a Fractured World Economy," edited by Shekhar Aiyar, Andrea Presbitero, and Michele Ruta, October 2023.
- International Monetary Fund (2023), “Geopolitics and Financial Fragmentation: Implications for Macro-financial Stability,” IMF Global Financial Stability Report, Chapter 3, April 2023.
- Gopinath, G. & Stein, J., (2021), "Banking, Trade, and the Making of a Dominant Currency," The Quarterly Journal of Economics, vol. 136(2), pages 783-830.
- Gopinath, G., Boz, E., Casas, C., Díez, F., Gourinchas, P.O., & Plagborg-Møller, M., (2020), “The dominant currency paradigm,” American Economic Review, 110(3), 677–719.

### Global financial safety net, IMF policy, and governance
- International Monetary Fund (2011), “Analytics of Systemic Crises and the Role of Global Financial Safety Nets,” International Monetary Fund Policy Paper.
- International Monetary Fund (2012), “Enhancing surveillance: Interconnectedness and clusters—Background paper,” International Monetary Fund Background Paper.
- International Monetary Fund (2016), “Adequacy of the Global Financial Safety Net,” International Monetary Fund Policy Paper.
- International Monetary Fund (2017a), “Collaboration between regional financing arrangements and the IMF,” International Monetary Fund Background Paper.
- International Monetary Fund (2017b), “Adequacy of the Global Financial Safety Net—Considerations for Fund Toolkit Reform,” International Monetary Fund Policy Paper.
- International Monetary Fund, (2017c), "Fifteenth General Review of Quotas—Adequacy of Fund Resources—Further Considerations" International Monetary Fund Policy Paper.
- International Monetary Fund (2024), “Digital money, cross-border payments, international reserves, and the global financial safety net: Preliminary considerations,” IMF Note 2024/001.
- International Monetary Fund (2025), Managing Director’s Global Policy Agenda: “Medium-Term Direction for the IMF: Anchoring Stability and Promoting Balanced Growth,” Spring Meetings, 2025.
- International Monetary Fund (2025b), “Guidance for staff on the Recording of Central Bank FX Liquidity Lines in IMF Staff Reports,” International Monetary Fund Staff Guidance Note, 2025.
- Iancu, A., Seunghwan, K., & Miksjuk, A., (2021), "The Global Financial Safety Net during the COVID-19 Crisis: An Interim Stock-Take," International Monetary Fund Special Series on COVID-19.
- Independent Evaluation Office of the International Monetary Fund (2023), “IMF’s Emergency Response to the COVID-19 Pandemic,” Evaluation Report.

### Central bank swap lines, bilateral swaps, and lender-of-last-resort issues
- Bahaj, S. and Reis, R., (2022), “Central Bank Swap Lines: Evidence on the Effects of a Lender of Last Resort,” Review of Economic Studies, 89 (4): 1654–1693.
- Perks, M., Rao, Y., Shin, J., and Tokuoka, L., (2021), “Evolution of Bilateral Swap Lines,” International Monetary Fund Working Paper 21/210.
- Jakree Koosakul, and Alexei Miksjuk., "Central Banks Casting a Global Financial Safety Net: What Drives the Supply of Bilateral Swaps?," IMF Working Papers 2024, 088 (2024), accessed August 22, 2025, https://doi.org/10.5089/9798400273759.001
- McDowell, D., (2019), "The (Ineffective) Financial statecraft of China's bilateral swap agreements," Development and Change 50, no. 1 (2019): 122-143.
- Cheng, G., Miernik, D., Barnieh, Y., Han, B., Sari, I., Pang, F., Kostanyan, T., Efimov, A., Houdart, M., de Carvalho, A., Giraldo, C., Monroy, V., (2018), “IMF-RFA Collaboration: Motives, State of Play, and Way Forward – A Joint RFA Staff Proposal,” ESM Discussion Paper Series no.4.
- Bahaj, S. and Reis, R., (2022), “Central Bank Swap Lines: Evidence on the Effects of a Lender of Last Resort,” Review of Economic Studies, 89 (4): 1654–1693.

### Financial stability, non-bank intermediation, and regulation
- Financial Stability Board (2023), “2023 bank failures: Preliminary lessons learnt for resolution,” https://www.fsb.org/2023/10/2023-bank-failures-preliminary-lessons-learnt-for-resolution/.
- Financial Stability Board (2024), “Global Monitoring Report on Non-Bank Financial Intermediation,” https://www.fsb.org/2024/12/global-monitoring-report-on-non-bank-financial-intermediation-2024/
- Denbee, E., Jung, C., & Paternò, F., (2016), "Stitching together the global financial safety net," Bank of England Financial Stability Paper, number 36.
- Grabel, I., (2018), “When things don’t fall apart: Global financial governance and developmental finance in an age of productive incoherence,” Cambridge: MIT Press.
- Copeland, R., (2019), “A Global Stablecoin: Revolutionary Reserve Asset or Reinventing the Wheel?,” Social Science Research Network (SSRN).

### Reserves, safe assets, and valuation
- Aizenman, J., & Sun, Y., (2012), "The financial crisis and sizable international reserves depletion: From ‘fear of floating’ to the ‘fear of losing international reserves’?," International Review of Economics & Finance, Elsevier, vol. 24(C), pages 250-269.
- Aizenman, J., Cheung, Y. W., and Ito, H. (2014), “International Reserves Before and After the Global Crisis: Is There No End to Hoarding?” NBER Working Paper No. 20386.
- Caballero, Ricardo J., Emmanuel Farhi, and Pierre-Olivier Gourinchas, (2017), "The Safe Assets Shortage Conundrum," Journal of Economic Perspectives 31 (3): 29–46.
- International Monetary Fund, "Review of the Method of Valuation of the SDR," International Monetary Fund Policy Paper 020 (2022).
- International Monetary Fund (2014), “Revised Guidelines for Foreign Exchange Reserve Management,” International Monetary Fund Manual and Guides.

### Climate, nature-related risks, and financial development
- International Monetary Fund, Monetary and Capital Markets Department, (2020), "Chapter 5. Climate Change: Physical Risk and Equity Prices," In Global Financial Stability Report, April 2020. USA: International Monetary Fund: https://doi.org/10.5089/9781513529196.082.ch005
- Gardes-Landolfini, C., Oman, W., Fraser, J., Montes de Oca Leon, M., & Yao, B., (2024),“Embedded in nature: Nature-related economic and financial risks and policy considerations,” International Monetary Fund, Staff Climate Notes, 2024/002.
- University of Notre Dame, Notre Dame Global Adaptation Initiative: https://gain.nd.edu/our-work/country-index/

### Methods, indices, and data sources
- Bailey, M., Al, A. Strezhnev., and Voeten, E., “Estimating dynamic state preferences from United Nations voting data,” Journal of Conflict Resolution 61.2 (2017): 430-456.
- International Monetary Fund Financial Development Index Database: https://legacydata.imf.org/?sk=f8032e80-b36c-43b1-ac26-493c5b1cd33b.
- The International Country Risk Guide (ICRG) by the PRS Group. https://www.prsgroup.com/explore-our-products/icrg/.
- Obstfeld, M., (2011), “The International Monetary System: living with asymmetry,” NBER Working Paper 17641.
- Rhee, C., (2024), “An Integrated Policy Framework - application to Korea's monetary policy,” keynote speech at the Korea International Economic Association Winter Conference, Seoul, 23 December 2024, https://www.bis.org/review/r250109b.pdf
- Bahaj, S. and Reis, R., (2022), “Central Bank Swap Lines: Evidence on the Effects of a Lender of Last Resort,” Review of Economic Studies, 89 (4): 1654–1693.

*Source: ppea2025035 - References.*

---


_Source: https://www.imf.org/-/media/files/publications/pp/2025/english/ppea2025035.pdf_