## _021411b

## Source details

**Canonical URL:** [_021411b](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/np/pp/eng/2011/_021411b.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/np/pp/eng/2011/_021411b.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/np/pp/eng/2011/_021411b.pdf.json)

---

### Executive summary — scope, methods, and survey evidence
- Focus: precautionary aspect of reserves—availability and liquidity for potential balance of payments needs; non-precautionary motives excluded.
- Methodology mix:
  - Simple metric-based approaches (transparent, tractable).
  - Cross-country reserve regressions (peer comparisons).
  - Scenario analysis and detailed engagement with country authorities.
- Survey of reserve managers (late 2010): "Twenty seven submissions," response rate "53 percent"; respondents view reserves as needed to meet a range of external and domestic risks.

### Key findings on reserve accumulation and adequacy
- Global and regional trends:
  - Gross reserves in EMs increased more than six-fold in the past decade to around US$5 trillion.
  - Asia contributed more than half of the increase; China’s reserves grew from US$170 billion to over US$2.4 trillion.
  - Median coverage ratios among EMs in 2009: around six months of imports; 200 percent of short-term debt; 30 percent of broad money.
- General conclusions:
  - Little consensus on an adequate level of precautionary reserves; traditional metrics provide conflicting signals; optimization approaches are sensitive to assumptions.
  - Peer-comparisons may amplify upward trends in perceived needs.
  - For EMs: most have at least adequate reserves under proposed metrics; some hold much higher reserves than suggested; only a few fall short.
  - For LICs: mixed picture; some LICs potentially need higher reserves.
  - For AMs: crisis experience raises questions; universally applicable quantitative metrics are harder to identify.

### Proposed two-stage risk-weighted metric for EMs
- Stage 1: construct a risk-weighted liability stock using observed outflows during exchange market pressure (EMP) events (focus on tenth percentile outflows).
- Stage 2: assess desired reserve coverage as a proportion of the risk-weighted liability stock.
- Empirical 10th percentile outflows (table values preserved):
  - Fixed: Exports 8.9; Broad money 12.4; Short-term debt 24.4; Other portfolio liabilities 27.6
  - Floating: Exports 2.3; Broad money 7.1; Short-term debt 24.4; Other portfolio liabilities 9.2
- Proposed round-number risk-weight formulas:
  - Fixed: 30% of STD + 15% of OPL + 10% of M2 + 10% of X
  - Floating: 30% of STD + 10% of OPL + 5% of M2 + 5% of X
- Performance and suggested adequacy range:
  - Coverage in the region of 100−150 percent of the metric flagged as potentially adequate for a typical country.
  - Empirical support:
    - Median reserve loss as a share of the metric: 10 percent.
    - Standard deviation: 15 percentage points for the metric versus 24 percentage points for STD.
    - Reserves losses of 50 percent of the metric correspond to the 95th percentile.
  - Three assessment approaches used: (i) crisis prevention (probability of EMP), (ii) crisis mitigation (consumption fall regressions), (iii) correlation with actual reserve losses.

### Cross-country reserve-demand regressions (EMs)
- Regression setup and fit:
  - Dependent variable: EM reserves/GDP, 1990-2008
  - Observations: 800
  - R-squared: 0.39
- Coefficient estimates (preserve exactly):
  - Imports/GDP 0.155**
  - M2/GDP 0.00002
  - Portfolio liabilities/GDP 0.363***
  - Terms of trade volatility 0.106***
  - Export volatility 0.267**
  - Partner growth volatility 0.024**
  - Interest rate differential with the US -0.018
  - Peg Dummy 0.017
  - Crisis Dummy 0.039**
  - Constant 0.003
- Insights:
  - Portfolio liabilities/GDP (0.363***) and Imports/GDP (0.155**) are robust determinants.
  - Short-term debt alone is a poor explanatory variable; broad money coefficient is negligible.
  - Prior crisis history (Crisis Dummy 0.039**) significant; Peg Dummy insignificant.

### LIC-specific optimal-reserves framework and empirical results
- Shock and crisis identification:
  - Shock defined as annual percentage change below the 10th percentile of country-specific distribution; crisis definition requires post-shock two-year average of real absorption below pre-shock three-year average and negative absorption growth in shock year.
  - Probability of a crisis within identified shock episodes: around 30 percent.
- Two-step empirical estimation (1990−2007):
  1. Probability of absorption drop (Panel Probit; No. obs. 445; R2 0.11). Key estimates (preserve exactly with significance):
     - Reserve, months of imports (t-1): -0.0896*** (0.0339)
     - Flexible exchange rate regime (t-1): -0.3801*** (0.1366)
     - Government balance, % of GDP (t-1): -0.0323*** (0.0125)
     - CPIA (t-1): -0.3090*** (0.1056)
     - IMF program (t): -0.3021** (0.1409)
     - Constant: 0.8648** (0.3614)
     - Finding: increasing reserves (months of imports) significantly reduces crisis probability; effect larger under fixed regimes.
  2. Magnitude of absorption drop (Panel OLS; No. obs. 418; R2 0.34). Key estimates:
     - Reserves, months of imports (t-1): -2.2403*** (0.6677)
     - Flexible exchange rate regime (t-1): -8.6983*** (2.1689)
     - External demand growth: -0.9320** (0.4356)
     - Terms of trade growth: -0.0841* (0.0484)
     - Finding: reserves reduce size of absorption losses; exchange rate flexibility facilitates adjustment.
- Calibration of optimal reserves:
  - Uses estimated coefficients, reference values for opportunity cost, and simplified risk-aversion assumptions (risk-neutrality baseline; higher risk-aversion raises optimal reserves).
  - Reference values for marginal product of capital in LICs: range 3 to 8 percent (Caselli and Freyer (2007) cited).
  - Illustrative results:
    - Insurance value of a flexible exchange rate regime: about 0.6 percent of GDP per year (or over three months of imports on average) in annual savings in the cost of holding optimal reserves (unit cost 4 percent).
    - Availability of contingent Fund support: annual savings about 0.3 percent of GDP per year (two months of imports).
  - Findings: optimal reserves generally higher for fixed regimes, commodity exporters, fragile states; three months of imports may be inadequate for some LICs.

### Reserves, costs, and reserve management
- Cost of holding reserves:
  - Net financial cost ("cost of carry") used as proxy in EMs—differential between external funding cost and return on liquid foreign assets; sovereign bond spread measures (J.P. Morgan EMBI) commonly used.
  - Median net financial cost of reserves remained "(at about 200 basis points)" in 2001−07; for the median EM country, reserves costs estimated "around 0.5 percent of GDP over 2001−09."
  - For very low external-debt countries, proxy should shift to foregone investment opportunities (e.g., investment-grade corporate bond spreads).
- Reserve management and institutional arrangements:
  - Pre-crisis: conservative SAA focused on low-yielding safe government bonds; pre-crisis years saw shifts into higher-yielding, less liquid assets.
  - Crisis-era behavior: central banks reduced exposure to nontraditional assets; assumed-liquid assets proved illiquid for many.
  - Global procyclicality: reserve managers’ flight to quality can exacerbate foreign currency funding problems elsewhere.
  - Asset-liability management (ALM) and coordination (central bank & debt management office) recommended to align reserves and sovereign liabilities.
- Empirical episodes:
  - Brazil: comfortable reserves helped limit adverse effects; ALM practices considered.
  - Korea: reserves fell from "$260 billion to $200 billion from May to November 2008"; Fed swap lines and macro-prudential measures used.
  - Russia: dual shocks led to reserve losses and use of SWF assets; institutional ALM reforms proposed.
  - Global deposit behavior: Pihlman and van der Hoorn (2010) estimate reserve managers pulled out roughly "US$500 billion of deposits from the banking sector between December 2007 and March 2009."

### Application, country judgment, and heterogeneity
- Distribution of coverage (end-2009, EM sample):
  - About 35 percent within proposed adequacy range.
  - Roughly 40 percent above the range.
  - About 25 percent below the range.
- Caveats on country-level application:
  - Among 11 countries below 100 percent: one dollarized, two with currency boards, another with extensive capital controls.
  - Metric weights should be adjusted for country features (e.g., creditor composition of STD, trade credits, commodity reliance, capital controls, dollarization, remittances, exchange-rate tolerance).
  - For LICs, supplement current-account focused approach with EM-type metrics when capital flight risk is present.
- Role of peer comparisons:
  - Regression-based fitted values explain much of average reserve increase since 2000 (dependent on premise that average precautionary insurance not systematically misaligned).
  - Large positive deviations suggest other motives or unobserved precautionary preferences.

### Advanced-market economies (AMs): considerations and lessons
- Crisis showed capital-account risks can strike AMs; reserve adequacy historically less emphasized.
- Central bank swap lines acted as critical substitutes for reserves in systemic events but are selective and limited for idiosyncratic shocks.
- Full insurance for AMs via reserves impractical given large financial markets; for some AMs modified EM-type metrics or scenario analysis may be relevant.
- Illustrative country experiences:
  - Denmark: reserves doubled since the crisis; reserves around US$80 billion (about 25 percent of GDP).
  - Sweden: borrowed about $15 billion via swap lines; reserves about US$50 billion, around 9½ percent of GDP.

### Policy implications and recommendations
- Reserves are useful to prevent crises and mitigate impact but have costs and diminishing returns.
- As reserves rise beyond adequate levels, emphasis should shift to:
  - contingent financing mechanisms (cheaper alternatives),
  - improved macroeconomic and prudential policies,
  - sovereign balance-sheet management and ALM.
- Sound policy frameworks (low sustainable public debt, low inflation, exchange rate near equilibrium, effective supervision) are first-line defenses; reserves are complements, not substitutes.
- Suggested process for assessments:
  - Start with the proposed simple cross-country metrics and regressions.
  - Follow with detailed scenario analysis and engagement with authorities to capture country-specific risks and available risk-management resources.
  - Explicitly consider contingent financing (IMF lines, swap lines), SWF liquidity, and other reserve complements in adequacy judgments.

### Remaining issues for discussion (excerpt)
- Is a move beyond traditional metrics warranted?
- For EMs, is the risk-weighted metric and suggested adequacy range (100−150 percent) appropriate?
- Does the LIC framework capture key factors bearing on reserves needs?
- Should AMs materially change their reserve approach, and how should it be calibrated?
- Do peer-comparisons provide satisfactory assessments?
- How should reserve management and sovereign balance-sheet risks be incorporated into adequacy evaluations?

_Italic source attribution: IMF staff chapter "_021411b - 12.      The global crisis and its aftermath in Europe have shown that capital account risks can strike AMs too, putting into question the typical assumption that they need little reserves." (source PDF)._

### Executive Summary ......................................................................................................

### EXECUTIVE SUMMARY

### Key findings on reserve accumulation and adequacy
- Global reserves holdings increased dramatically over the past decade and resumed increasing since the global financial crisis, "even at an accelerated pace."
- There is little consensus on what constitutes an adequate level of precautionary reserves: traditional metrics (e.g., "three months of imports" or cover of short-term debt) are narrowly-based and often provide conflicting signals; newer optimization approaches are sensitive to modeling assumptions and calibrations.
- Assessments based on peer comparisons likely amplify upward trends as perceived needs rise with observed holdings.
- For emerging market economies (EMs):
  - The proposed simple new metrics suggest that most EM countries’ reserve levels are adequate, with some holding much higher reserves than suggested and only a few falling short.
  - A separate analysis of reserve demand based on precautionary variables points to similar conclusions in terms of relative reserve holdings among countries.
- For low-income countries (LICs):
  - The picture is more mixed; some LICs potentially need to hold higher reserves according to their specific characteristics.
- For advanced-market economies (AMs):
  - The crisis raises questions about reserve adequacy, but universally applicable quantitative metrics are even harder to identify; the paper reviews qualitative considerations.

### Scope and definition of analysis
- Focus is on the precautionary aspect of reserves—their availability and liquidity for potential balance of payments needs.
- Non-precautionary motives for reserves (e.g., exchange rate policy, intergenerational savings, sovereign wealth funds) are outside the paper’s scope.
- "Precautionary need" in this paper refers specifically to the need for readily available external liquid assets.
- "Adequacy" is considered against total resources available to meet shocks (central bank swap lines, SWFs, access to IMF financing), but the analysis concentrates on officially reported gross reserves for data comparability.

### Approach and methodology
- Recognition that an ideal cost-benefit optimization is impeded by large uncertainty about utility and cost functions.
- The paper uses a mix of approaches:
  - Simple metric-based approaches that are transparent and tractable but require judgment on tail-risk coverage.
  - Cross-country reserve regressions (peer comparisons) to infer revealed precautionary preferences and relative standing.
  - Scenario analysis and detailed engagement with country authorities to assess country-specific risks and broader risk-management frameworks.
- Survey evidence: a survey of reserve managers in late 2010 yielded "Twenty seven submissions," a response rate of "53 percent." Responses indicated countries view reserves as needed to meet a range of risks, both external and domestic.

### New proposed approach for emerging markets
- A two-stage "risk-weighted" metric is proposed:
  - Stage 1: Develop a metric that reflects relative risk levels of different potential sources of balance of payments pressure, based on observed outflows in exchange market pressure events—provides a "better ruler" to measure reserves.
  - Stage 2: Assess how much reserves cover might be needed relative to this risk-weighted measure, analogous to bank capital assessed as a percentage of risk-weighted assets.
- Complemented by peer-comparison regression analysis: precautionary variables do not fully explain recent average reserves growth, but regressions provide information on precautionary factors and on countries’ holdings relative to peers; results broadly consistent with the metric approach.

### Framework for low-income countries (LICs)
- Proposes a tractable "optimal reserves" framework for LICs:
  - Empirically estimates absorption-smoothing benefits of reserves conditional on adverse external shocks using data on past severe shock episodes.
  - Calibrates optimal reserves using estimated regression coefficients, reference values for the opportunity cost of holding reserves, and simplified assumptions about risk-aversion.
  - Explicitly accounts for interactions among country fundamentals, access to Fund financing in the event of a shock, and optimal reserve holdings.
- The approach relates optimal reserve levels to economic structure and the nature and likelihood of exogenous shocks, thereby capturing country circumstances.

### Policy implications and recommendations
- Reserves are useful in both preventing crises and mitigating their impact, but:
  - Reserves are costly at national and global levels and subject to diminishing returns.
  - As reserves rise beyond adequate levels, increased attention should be paid to other elements of sovereign risk management, including contingent financing mechanisms, country insurance, and macroeconomic and prudential policies.
- Sound macroeconomic and prudential frameworks (low and sustainable public debt, policies maintaining low inflation and exchange rate near equilibrium, effective financial supervision) remain the most direct way to limit vulnerability; reserves help even where policy frameworks are weak but are not a substitute for sound policies.
- Reserve assessments should start from the simple cross-country metrics and regressions proposed, but be followed by detailed scenario analysis and engagement with authorities to capture country-specific risks and the full range of available risk-management resources.

### Governance and prior guidance
- The paper builds on past Fund discussions and guidance, including:
  - SM/00/65 (3/23/2000) and SM/01/311 (10/16/2001) emphasizing short-term debt and the need to consider various potential sources of short-term reserve demand and balance of payments stress testing.
  - More recent Board discussion in SM/10/86 (4/13/2010) where many Directors supported "further analytical work to provide guidance on appropriate levels of precautionary reserves tailored to country circumstances."

_Executive Summary from the IMF paper titled "Executive Summary" (source PDF)._

### 12.      The global crisis and its aftermath in Europe have shown that capital account

### _021411b - 12.      The global crisis and its aftermath in Europe have shown that capital account risks can strike AMs too, putting into question the typical assumption that they need little reserves.

### Implications for individual country reserves levels
- Analysis implies most EM countries have at least adequate reserves, while LICs present a more mixed picture.
- Several EM countries hold higher reserves than appear needed for precautionary purposes; only a few EMs hold clearly less than adequate levels.
- For LICs, results point to additional reserves needs for some more vulnerable categories of economies.
- Simple ―across-the-board‖ assessments of reserve adequacy are limited; judgment and detailed country information are needed for full country-level assessments.

### Outline of the paper (sections covered)
- Section II: reserves in EMs — trends, crises experience, existing approaches, costs, and reserve management.
- Section III: suggested approach to assessing reserve adequacy in EMs — derivation of a new metric and reserve demand regressions.
- Section IV: reserves issues for LICs and proposed metrics.
- Section V: preliminary considerations relevant for AMs.
- Section VI: conclusions and issues for discussion.
- Supplement: further technical details.

### Trends in reserve accumulation and crisis experience (EMs)
- Gross reserves in EMs increased more than six-fold in the past decade to around US$5 trillion.
- Asia contributed more than half of the increase; China’s reserves grew from US$170 billion to over US$2.4 trillion.
- Median coverage ratios among EMs in 2009 were:
  - around six months of imports
  - 200 percent of short-term debt
  - 30 percent of broad money
- Reserve accumulation accelerated, slowed during the crisis, and then regained speed.
- During the crisis, many individual countries experienced significant reserve use despite aggregate increases in reserves in 2008 and 2009.
- Some countries actively used reserves to reduce FX market volatility or provide foreign currency liquidity to banks; others avoided using reserves for fear that depletion would signal external weakness.

### Observed benefits of reserves during crises
- Event-study evidence: EM countries with higher reserve holdings better maintained consumption growth relative to pre-event trend during exchange market pressure periods.
- Higher reserves correlated with greater capacity to expand fiscal policy; low reserves associated with procyclical fiscal contraction.
- Diminishing marginal benefits: effects are more apparent when moving from low to moderate reserve levels than from moderate to high levels.

### Definition of Reserves (Box 1)
- Reserve assets: external assets readily available to and controlled by monetary authorities for meeting balance of payments financing needs; intervention in exchange markets; and related purposes (confidence in the currency and the economy; basis for foreign borrowing).
- Adequacy is assessed by capacity to prevent or mitigate external shocks.
- Reserve complements and caveats:
  - Assets in SWFs (particularly the liquid portion) could be employed as reserve complements; some countries arrange for SWF assets to be used for balance of payments purposes or to be qualified as reserve assets.
  - Commodity price stabilization funds can disburse when prices are weak and support balance of payments despite no explicit BOP support function.
  - Contingent credit (e.g., Fund credit lines) and central bank swap lines are reserve substitutes but have limits to substitutability; swap lines are typically extended selectively to countries with strong AM linkages and high creditworthiness.
  - During crises, some assets counted as reserve assets (e.g., covered bonds, asset and mortgage backed securities) may prove less liquid than expected.
- Analysis in the paper is limited to the standard definition of reserve assets, with discussion of supplements; data limitations are a caveat.

### Use of reserve substitutes during the crisis
- Survey of reserve managers: most recognized central bank swap lines, IFI contingent credit lines, and SWF assets as able to play the role of reserves, though few reported actual use.
- Swap lines during the crisis were available only to EMs with strong AM financial and trade linkages.
- Users of the Fund’s Flexible Credit Line (Colombia, Mexico, and Poland) used the facility to bolster market sentiment.
- Some countries used World Bank resources (e.g., Indonesia) or SWF assets instead of reserves, possibly to avoid alarming markets about falling official reserves.
- The reversal of intended roles between reserves and SWFs raises measurement issues and complicates firm guidance on reserves.

### Existing approaches to reserve adequacy (summary and critique)
- Traditional metrics remain widely used but are difficult to interpret and inherently arbitrary.
- Common rules of thumb and related commentary:
  - Import cover:
    - Measures months of prospective imports that can be sustained if inflows cease.
    - Three months’ coverage is a traditional benchmark, but empirical support for three months versus two or six months is lacking.
  - Short-term debt (STD):
    - The ―Greenspan-Guidotti‖ rule: 100 percent cover of STD (one year) is widely used for EMs.
    - The 12-month cut-off is arbitrary; crisis durations and rollover behavior vary.
    - Empirical observation: STD was a poor predictor of reserve falls in the crisis—little association between 2007 STD levels and crisis reserve losses in 08Q4–09Q1.
    - Examples: Armenia, Ecuador, Malaysia, and Morocco lost reserves over 70 percent of STD despite low STD; Baltics and Bulgaria had high STD but saw reserve losses averaging less than 10 percent of STD.
  - Broad money (typically M2):
    - Less firmly grounded; 20 percent of M2 is often cited as an upper end.
    - Intended to capture deposit outflow risk/capital flight or potential need for bank support.
    - Some studies find perverse correlations; narrow money in CBAs has statutory roles, but coverage of narrow money alone may be of limited relevance in modern banking systems.
  - GDP:
    - Sometimes used mainly as a scaling factor for cross-country analysis; little theoretical or empirical backing as a direct adequacy metric.

### Precautionary models of reserve accumulation (Box 2)
- Open-economy precautionary saving parallels closed-economy literature: determinants include shocks to incomes/output, borrowing constraints, discount rates, interest-rate endogeneity, and risk aversion.
- Optimal reserve holdings depend on shock types:
  - LICs: focus on current account–based measures (reserves to imports). Calibrated small open-economy models can yield much higher optimal reserves than heuristics:
    - Valencia (2010) for Bolivia (commodity exporter) finds optimal reserves of 29−37 percent of GDP.
    - Consistent with Cashin, Liang and McDermott (2000) and Borensztein, Jeanne and Sandri (2009) on commodity price shock persistence.
  - EMs: face capital account shocks (sudden stops, currency crises) as well as current account shocks; buffer-stock models extended to include financial deepening and sudden-stop probabilities.
    - Aizenman and Marion (2003), Aizenman and Lee (2007) provide regression evidence for precautionary motives in East Asia post-Asian Crisis.
    - Wyplosz (2007) argues total external liabilities—not just STD—may better measure capital account risk; under that metric, emerging Asia’s reserves are not unprecedented.
    - Obstfeld, Shambaugh and Taylor (2009) find M2-to-GDP performs better than STD in accounting for reserve accumulation (data up to 2004).
  - AMs: literature often abstracts from borrowing constraints and sudden-stop risk; precautionary saving models (e.g., Ghosh and Ostry (1997)) applied to several advanced economies. Some AMs that issue debt in a currency they cannot control face capital account risks similar to EMs.

*Source: IMF chapter "_021411b - 12.      The global crisis and its aftermath in Europe have shown that capital account risks can strike AMs too, putting into question the typical assumption that they need little reserves." — https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/np/pp/eng/2011/_021411b.pdf*

### 20.      Combination metrics seek to reflect a broader range of sources of risk. The most

### 20.      Combination metrics seek to reflect a broader range of sources of risk.

### Combination metrics and practical formulations
- Expanded Greenspan-Guidotti rule commonly used: "STD plus the current account deficit (if it is in deficit)" to reflect full potential 12-month financing need.
- Critique of asymmetric treatment: treating current account only when in deficit implies assuming a larger shock for surplus countries than for deficit countries; alternative suggested: "STD minus the current account balance" (positive or negative).
- In practice, traditional metrics are often used in combination, for example:
  - check reserves against the highest of "three months of imports, 100 percent of STD and 20 percent of M2."
- Other combination metrics:
  - Wijnholds and Kapteyn (2001) use STD and M2 to model debt repayments and capital outflows, incorporating exchange rate regimes and country risk.
  - The new combination metric proposed in Section III builds on these ideas to develop a more empirically-based index of broad potential needs while retaining simplicity and ease of use.

### Traditional metrics versus actual reserve holdings
- The three "traditional" metrics span a broad range: the average range covered by the three traditional metrics for EM countries is nearly "15 percent of GDP."
- Actual reserve positions relative to the three traditional metrics:
  - "only a third of countries have reserves levels within this range;"
  - "nearly 60 percent are above all three measures;"
  - "only 5 percent below all three."

### Model-based approaches to reserve adequacy
- Cost-benefit optimization models describe reserve accumulation as an explicit optimization problem: reserves chosen to provide optimal insurance against sudden drops in consumption given costs of holding reserves.
- Key components/assumptions in common models:
  - benefits: lower probability of a crisis and smoothing consumption during a crisis;
  - costs: yield on reserves lower than other assets (opportunity cost);
  - Jeanne and Rancière (2006) framework: risk-averse policymakers maximize welfare in a small open economy vulnerable to sudden stops; optimal reserves determined by size and probability of sudden stop, potential loss in output and consumption, opportunity cost, and degree of risk aversion.
- Limitations and sensitivity:
  - optimal reserves very sensitive to assumed parameters and to the stylized economic structure;
  - in capital market crises (typical for EMs), output losses are more endogenous than in externally-driven current account crises (typical for LICs), leading to a broad range of "optimal" reserves that may limit usefulness for EMs.
- LIC-focused models:
  - Barnichon (2009): insurance against natural disasters or terms-of-trade shocks;
  - Drummond and Dhasmana (2008): extend Jean-Rancière framework for aid and terms-of-trade shocks;
  - outcomes remain critically dependent on assumed parameters.

### Scenario analysis and contingent financing
- Country-specific adverse scenario analysis: shocks applied to components of capital and current accounts over time, magnitudes determined by past experience, projections, and country characteristics; resulting financing gap translated into potential drains on reserves.
- Advantages:
  - brings in rich country- and time-specific information;
  - allows judgment that numerical metrics cannot provide.
- Uses:
  - widely used in recent staff reports and by reserve managers;
  - used for contingent financing vehicles (e.g., FCL arrangements and precautionary SBAs) where potential financing gaps may be fully or partially met by Fund resources.
- Concerns and limitations:
  - difficulty determining appropriate level of stress;
  - absence of explicit accounting of costs of holding reserves;
  - difficulty ensuring consistency across countries.

### Country practices, peer comparisons, and perceptions
- Survey and staff report findings:
  - country authorities report predominant use of traditional measures and peer comparisons;
  - all three traditional measures are taken into account by many countries;
  - "about 10 percent of reports using optimality methods" as part of more in-depth examinations.
- Risk of peer-driven dynamics:
  - peer comparisons widely used by market participants, potentially contributing to an upward spiral in perceived reserve needs absent more fundamental analysis.
- Assessments of reserve adequacy (IMF survey of reserve managers and staff calculations):
  - Reserve managers’ assessments (pie categories shown in source):
    - "Very inadequate" / "Somewhat inadequate" / "About right" / "Somewhat higher than needed" / "Significantly higher than needed" (categories reported qualitatively in text).
  - In recent staff reports (percent of total):
    - "Too much/ 'ample' 5%"
    - "Comfortable/ sufficient 18%"
    - "Scope to increase/ insufficient 20%"
    - "Uncertain 4%"
    - "No assessment 53%"
  - Most reserve managers who responded to the survey considered their level of reserves to be "about right", despite other surveys noting central banks taking on more investment risk as reserve levels rose.

### Cost of holding reserves
- Importance: decisions on reserve levels should assess costs as well as benefits.
- Survey: two-thirds of countries use quantified measures of costs, often within cost-benefit frameworks; little consensus on appropriate cost measures.
- Candidate cost indicators:
  - sterilization costs;
  - exchange rate valuation losses (actual or potential);
  - opportunity cost of foregone consumption or investment;
  - cost from maturity mismatch between reserves and sovereign liabilities.
- Net financial cost ("cost of carry") as likely best proxy for opportunity cost in EMs:
  - defined as differential between external funding cost of reserves and return on liquid foreign assets—typically using sovereign bond spread measures such as the J.P. Morgan EMBI index.
  - relies on assumptions: market access and that, at the margin, reserves are effectively financed by external debt; possible matching maturities between reserves and debt.
- Endogeneity and empirical patterns:
  - empirical studies find a negative relationship between reserves and spreads—using spreads can overstate opportunity cost of reserves.
  - effect depends on relative levels of reserves and debt: when reserves are low and debt high, increases in reserves reduce government funding costs much more than when reserves are high and debt low.
  - estimates (Levy-Yeyati methodology) suggest the net financial cost effect fell dramatically in the 2001–07 period; although EMBI spreads fell sharply between 2001 and 2007, the median net financial cost of reserves remained relatively flat "(at about 200 basis points)", before rising with spreads in crisis years.
  - for the median EM country, reserves costs are estimated to have averaged "around 0.5 percent of GDP over 2001−09."
- For countries with very low external debt:
  - net financial cost of reserves implied by foregone alternative investment opportunities rather than debt financing costs;
  - investment grade corporate bond spreads in advanced economies may be a better proxy for opportunity cost.

### Reserve management and institutional arrangements
- Historical SAA and crisis-era shifts:
  - traditionally conservative reserve management: majority held in low-yielding safe liquid government bonds.
  - pre-crisis years saw shifts into higher-yielding, less liquid assets; during the crisis, central banks reduced exposure to these nontraditional assets (often from fear of losses rather than liquidity needs).
- Risks of inadequate reserve management:
  - less liquid or riskier assets may be sold only at discounts or with delay;
  - assumed-liquid asset classes proved illiquid in the crisis—especially when many central banks sold large amounts of a particular asset class.
  - global procyclicality: reserve managers reducing exposure to non-government debt exacerbated foreign currency funding problems at commercial banks, prompting domestic authorities to provide foreign currency liquidity; one central bank’s flight to quality can reduce reserves at another.
- Policy implications for SAA and incentives:
  - SAA should attach more weight to countercyclical assets that rise in value during downturns;
  - active portfolio mandates to outperform benchmarks can create wrong incentives if managers invest in cyclical assets such as credits;
  - where reserves are more than adequate, authorities may manage excess reserves as part of sovereign wealth.
- Institutional arrangements to reduce vulnerabilities:
  - public debt management and prudential regulation that address private sector imbalances can reduce calls on FX reserves;
  - large foreign currency share and short maturities of public debt increase vulnerability to FX appreciation and foreign rate increases.
- Asset and liability management (ALM):
  - ALM techniques can coordinate risk profiles of reserves and public debt; many countries match currency composition of reserves with short-term external debt but few consider the complete sovereign balance sheet;
  - greater ALM coordination (e.g., information exchange between central bank and debt management office) could reduce financial risks and vulnerabilities.
- Country examples of ALM and crisis experience:
  - Brazil: comfortable reserve level helped limit adverse effects; central bank applied ALM by accounting for characteristics of external debt.
  - Korea: reserves fell from "$260 billion to $200 billion from May to November 2008"; response included Fed swap lines, use of reserves, and macro-prudential measures to reduce maturity mismatches and limit short-term external debt buildup.
  - Russia: dual shocks (collapse in oil prices and reversal of capital flows) led to reserve losses and use of SWF assets for fiscal expansion; prompted draft law in April 2010 to create an agency to manage National Wealth Fund and public debts and introduce ALM to manage asset-liability mismatches.
- Global reserve manager behavior during crisis:
  - Pihlman and van der Hoorn (2010) estimate reserve managers pulled out roughly "US$500 billion of deposits from the banking sector between December 2007 and March 2009."

*Content derived from IMF staff analysis in the provided chapter.*

### 35.      Prudential regulation could help address private sector imbalances and reduce

### 35.      Prudential regulation could help address private sector imbalances and reduce

### Role of prudential regulation and reserves
- Strong private sector balance sheets with low exchange rate, credit, and liquidity risks can reduce the need for central bank buildup of foreign exchange reserves by insulating the private sector from sudden reversals of capital flows or market freezes.
- Strong supervision and regulation of financial institutions can minimize risks embedded in these institutions’ balance sheets and reduce the likelihood of calls on foreign exchange reserves.
- During the crisis, some countries have adopted prudential measures to manage vulnerabilities arising from external liabilities in the financial sector.

### Analytical approaches for emerging markets (overview)
- Two complementary approaches are used to assess reserve needs:
  - Cross-country experience on outflows during times of exchange market pressure to form a single risk-weighted metric of liability structure risks.
  - Regression analysis of observed reserve holdings against a range of precautionary variables to infer consistency of actual reserve levels with peers.

### Sources of balance of payments risk (identified drains)
- Multiple sources can generate balance of payments pressure, mostly in the capital account; combinations vary across crises.
- Empirical observations:
  - Russia: capital flight (outflows into foreign assets) and short-term liabilities were primary drains.
  - Indonesia: sale of equities by nonresidents was the major source of pressure.
  - Brazil: drains from lower export earnings and multiple financial account sources occurred simultaneously.
  - Comparison of Brazil and Russia (1998 vs 2008−09) shows reserves’ role: large outflow pressures produced full-blown crises only during lower reserve periods.
  - Korea: despite much higher reserves during the recent crisis, contingent reserve substitutes (foreign currency swap arrangements) played an important supplementary role.
- Four specific sources of drains emphasized (separate and essentially non-overlapping):
  - Export earnings (X) — potential loss from drop in external demand or terms of trade shock.
  - Short-term debt (STD) — external short-term liabilities at remaining maturity.
  - Medium- and long-term debt and equity liabilities (Other portfolio liabilities, OPL).
  - Broad money (M2) — proxy for domestic liquid assets that could be sold and transferred into foreign assets during a crisis.
- Exclusions and caveats:
  - FDI liabilities not included due to lack of evidence of being sold during stress in the dataset.
  - Other risks (derivatives exposures, speculative attack) excluded due to data limitations or proxying difficulty, though approach can be modified if data allow.

### Derivation of the proposed reserve-adequacy metric (two-stage approach)
- Stage 1: Estimate relative riskiness of different drains based on observed distributions of outflows during exchange market pressure (EMP) events and construct a "risk-weighted liability stock".
  - EMP identification follows an exchange market pressure index where deviations more than 1.5 times its standard deviation are considered significant pressure.
  - Distributions of annual percentage losses are estimated separately for fixed and floating exchange rate regimes.
  - The approach focuses on drains observed at the tenth percentile of each distribution to derive relative risk weights.
  - Adjustments made for alternative estimation methodology, judgment, and to avoid false precision.
- Stage 2: Gauge desired proportion of the risk-weighted liability stock to be held as liquid reserves using past crisis experience (crisis prevention, crisis mitigation, observed reserve losses).

### Estimated risk weights (10th percentile outflows during EMP events)
- Empirical 10th percentile outflows (as presented):
  - Table values shown as:
    - Fixed: Exports 8.9; Broad money 12.4; Short-term debt 24.4; Other portfolio liabilities 27.6
    - Floating: Exports 2.3; Broad money 7.1; Short-term debt 24.4; Other portfolio liabilities 9.2
- Proposed round-number risk-weight formulas (to reflect uncertainty and ease of use):
  - Fixed: 30% of STD + 15% of OPL + 10% of M2 + 10% of X
  - Floating: 30% of STD + 10% of OPL + 5% of M2 + 5% of X

### Performance of the metric and empirical evidence
- Three approaches to assess appropriate coverage against the metric:
  (i) Crisis prevention: The ratio of reserves to the metric is strongly associated with reduced probability of an EMP event; the proposed metric outperforms alternate traditional metrics (import cover, STD, M2, maximum of traditional metrics, GDP, Wijnholds and Kapteyn metric).
  - A more prudent fiscal position (cyclically adjusted primary balance) is associated with lower crisis probability; inclusion does not remove significance of the reserves/metric ratio.
  (ii) Crisis mitigation: Regressions of consumption fall in crisis events on the reserves ratio show the reserves ratio is highly significant and outperforms other metrics; other fundamentals and policy variables are also important.
  (iii) Correlation with reserve losses: The new metric correlates better with actual reserve losses than STD alone.
  - Median reserve loss as a share of the metric: 10 percent (same as for STD).
  - Standard deviation: 15 percentage points for the metric versus 24 percentage points for STD.
  - Reserves losses of 50 percent of the metric are at the 95th percentile.

### Suggested adequacy range and implications
- Evidence suggests coverage in the region of 100−150 percent of the metric might be regarded as adequate for a typical country.
  - Under method (i): crisis probabilities rapidly tail off beyond this range.
  - Under method (ii): large consumption drops were concentrated in countries with ratios below this level.
  - Under method (iii): actual reserve losses in the crisis typically at 10 percent of the metric and rarely exceed 50 percent.
- Countries may choose to hold more or less than 100−150 percent depending on circumstances and risk aversion.
- Given costs of reserves, greater emphasis should be placed on other defenses:
  - Cheaper contingent financing mechanisms.
  - Improving overall policy and sovereign risk management frameworks.

### Judgment in country application (examples of adjustments)
- Considerable judgment required when applying the metric to individual countries; examples where weights might be adjusted:
  - Reduce weight on STD where liabilities are predominantly owed to foreign parents; increase when from wholesale market.
  - If a large share of trade credits within STD, less reserve cover may be needed.
  - Increase weight on exports where reliant on particularly volatile commodities.
  - Reduce or eliminate M2 weight where effective capital controls prevent capital flight.
  - Dollarized economies: foreign currency needs may differ; absence of currency risk could lower needs but foreign currency liquidity needs of banking sector could be very high.
  - Countries relying on remittances could supplement the metric with an estimated weight on remittances.
  - Within floating regimes, tolerance for large exchange rate swings permits lower reserve coverage; countries seeking to avoid depreciation may hold more.
  - Countries with weak fundamentals might want higher coverage while addressing underlying vulnerabilities.
- Adjusting the metric per country is a first step toward full scenario analysis for liquidity buffers; keeping the metric simple aids transparency and avoids a "black box".

### Cross-country regression approach to reserve adequacy
- Modeling actual reserve holdings can shed light on precautionary motivations and allow comparison across countries.
- The estimated model for EMs fits average reserve growth reasonably well; precautionary variables explain most reserve growth over the past decade.
- Regressions do not track individual country holdings closely—some countries hold reserves for purposes beyond those identified or differ in risk aversion—but relative positions from regressions align well with the metric calculations.

*Source: IMF staff paper excerpt on reserve adequacy and prudential regulation (text provided).*

### 44.      Use of the reserve demand model for assessing the level of reserve adequacy is

### Use of the reserve demand model for assessing the level of reserve adequacy

### Model assumptions and interpretation
- The reserve demand model is predicated on the assumption that, averaged across countries and over the regression sample period, countries are neither systematically over- or under-insured from a precautionary perspective.
- Given this premise, the model indicates how individual EMs’ holdings of reserves may be expected to vary from the sample average depending on specific vulnerabilities and reserves holding costs.
- The model’s ability to explain much of the increase in average reserve holdings by EMs since 2000 suggests that—if EMs as a group were appropriately (or somewhat under-) insured during the 1990s—much of this increase is consistent with precautionary demand against rising risks for the average country.
- Substantial positive deviations of individual countries’ reserve holdings from the model prediction suggest that other motivations (precautionary or non-precautionary) are at play.

### Determinants of reserve holdings (EM regression results and interpretation)
- Regression fit and sample:
  - Dependent variable: EM reserves/GDP, 1990-2008
  - Observations: 800
  - R-squared: 0.39
- Coefficient estimates (preserve exactly):
  - Imports/GDP 0.155**
  - M2/GDP 0.00002
  - Portfolio liabilities/GDP 0.363***
  - Terms of trade volatility 0.106***
  - Export volatility 0.267**
  - Partner growth volatility 0.024**
  - Interest rate differential with the US -0.018
  - Peg Dummy 0.017
  - Crisis Dummy 0.039**
  - Constant 0.003
- Key empirical insights:
  - External environment: volatility of real GDP growth in trading partners and volatility of the terms of trade have explanatory power with expected signs; some external-uncertainty proxies (e.g., the VIX) are not significant in regressions.
  - Current account risks: the import ratio is significant and the most robust determinant; export receipt volatility is also significant. Current account variables are particularly significant in explaining the increase in reserves over the past decade.
  - Capital account risks: short-term debt is not a good explanatory variable; overall portfolio liabilities dominate (coefficient 0.36), comparable to weights used in the metric approach. Broad money has expected sign but a very low and statistically insignificant coefficient in the EM regression.
  - Cost of holding reserves: the interest rate differential proxy is of the expected negative sign but not statistically significant.
  - Other factors: prior crisis dummy is significant (countries with past crises have greater precautionary demand). Pegged exchange rate regime dummy is insignificant.

### Main conclusions from the EM analysis
- Precautionary motives appear important in explaining both time series and cross-sectional variation in reserve holdings.
- For some countries, additional reasons to hold reserves exist beyond the model’s precautionary elements; these can be non-precautionary motivations or precautionary motives not identified econometrically (e.g., generalized but unobservable increases in risk aversion).
- Traditional variables (such as imports) remain useful, but other liabilities and potential calls on reserves are clearly important.

### Application of the EM frameworks and metric-based assessments
- Distribution of reserve coverage (based on end-2009 data):
  - About 35 percent of the sample was within the proposed adequacy range.
  - Roughly 40 percent had reserves above the range.
  - About 25 percent were below the range.
- Caveats when assessing individual countries:
  - Of the 11 countries with cover below 100 percent, one is dollarized, two have currency boards, and another has extensive capital controls — factors that can significantly affect reserve adequacy assessment.
  - Some countries above the suggested range may still have important vulnerabilities (e.g., high commodity price volatility) that raise precautionary needs.
  - Large reserve holdings are not necessarily “excessive” broadly—for example, nonrenewable resource exporters with limited institutional capacity may hold precautionary savings in reserves rather than alternative vehicles.
- Consistency with traditional metrics:
  - The maximum of three traditional metrics lies within the 100−150 percent range of the new metric in more than half the cases.
  - For a number of countries, the “maximum of traditional” approach would point to much higher reserves than either the new measure or actual holdings.
  - The proposed metric extends existing practices and can serve as a better precautionary yardstick than traditional metrics alone.

### Cost–benefit application and model dependence
- A standardized cost-benefit framework calibrates reserve holdings relative to the metric as defense against a balance-of-payments shock associated with a given (exogenous) output loss.
- The benefit of avoiding the output loss is balanced against the cost of reserves (represented by EMBI spreads).
- Two illustrative optimal-reserve outcomes shown:
  - Optimal reserves under assumption of 6.5% output loss.
  - Optimal reserves under assumption of 10% output loss.
- For most countries, resulting optimal levels fall within a margin of error of the metric’s “adequacy” level — illustrating sensitivity to model parameters and structure.

### Regression-based country comparisons and consistency
- Comparison of actual reserve holdings with fitted regression values (2009):
  - Some countries show very high reserves compared to peers; others show smaller negative deviations.
  - Information from regression deviations is reasonably consistent with positions against the proposed metric: most countries are relatively high on both measures or relatively low on both; only a few cases show divergent messages.

### Reserves in LICs: trends, cushion effect, and a tailored approach
- Recent LIC reserve accumulation:
  - Median coverage ratios among LICs in 2009: around 4.7 months of imports, and 55 percent of broad money.
  - As of 2009, over a quarter of all countries had reserve levels below three months of imports.
- Evidence on cushion against shocks:
  - Event study (1980-2007, shock defined as bottom 10th percentile of country-specific distribution) indicates countries with reserves covering more than three months of imports were better able to smooth consumption and absorption.
  - Example magnitudes: cumulative consumption losses over a five-year period were about 6−17 percentage points for countries with reserve coverage of less than three months of imports; impact was limited for those with higher coverage.
  - Shock-mitigation effect of reserves is particularly pronounced in highly indebted economies, small islands, commodity exporters, and countries with fixed exchange rate regimes.
- LIC regression (1992–2001) and fit:
  - Dependent variable: LIC reserves/GDP, 1992-2001
  - Observations: 414
  - R-squared: 0.639 (regression includes a constant term)
  - Coefficient estimates (preserve exactly):
    - Income -0.0045***
    - Log(Population) -2.2280***
    - Imports/GDP 0.2611***
    - Exchange rate volatility -0.0351**
    - Export volatility 0.0482**
    - M2/GDP 0.3374***
    - Peg Dummy 1.2851*
    - Interest rate differential -0.2178
  - The empirical model accounts for over 60 percent of variation in reserves (excluding country fixed effects), suggesting precautionary motivations are important for average LIC reserve growth.
  - Out-of-sample forecasts for 2003−08 (excluding the 2009 SDR allocation) indicate LIC reserve growth was broadly in line with evolving fundamentals.
- Heterogeneity across LIC groups:
  - Commodity-exporting LICs: export volatility is a more important determinant.
  - Noncommodity-exporting LICs: import ratio is particularly significant.
  - Implication: methodologies for assessing adequacy need tailoring to country structural characteristics and risk types.

### Rationale for a LIC-specific approach
- Key differences between EMs and LICs that motivate a tailored LIC framework:
  - Limited access to international markets: external drains in LICs are primarily on the current account; current-account-based measures (e.g., months of imports) remain useful for most LICs.
  - Exogenous nature of shocks: LICs face different shocks (sharp swings in foreign aid, remittances, FDI, natural disasters) that are often exogenous; estimating reserves needed to combat these shocks is less statistically problematic than for EMs where capital-account crises can be endogenous to reserves.  

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

### 57.      The proposed approach aims to provide a tractable optimal reserves framework

### The proposed approach aims to provide a tractable optimal reserves framework for assessing reserve adequacy in LICs

### Identification of shocks and crises
- Large external shock events (1990−2009) are identified across: external demand, terms of trade, FDI, foreign aid, remittances and large natural disasters.
- Shock definition: annual percentage change of the relevant variable falls below the 10th percentile in the left-tail of the country-specific distribution.
- Crisis definition within shock episodes (both conditions must hold):
  - the post-shock two-year average of real absorption falls below the pre-shock three-year average; and
  - absorption growth is negative in the shock year.
- The probability of a crisis within identified shock episodes is around 30 percent.
- Note: There is a marked difference in real absorption growth between the crisis and non-crisis samples (of over 8 percentage points).

### Empirical estimation of crisis probability and magnitude
- Two-step estimation of absorption-smoothing benefits of reserves:
  1. Estimate probability of a drop in absorption in the event of external shocks as a function of reserves (months of imports) and pre-shock country-specific fundamentals (Panel Probit regression, 1990-2007).
     - Key regression estimates (Probability of Absorption Drop; Panel Probit, 1990-2007; No. of observations 445; R2 0.11):
       - Reserve, months of imports (t-1): -0.0896*** (0.0339)
       - Flexible exchange rate regime (t-1): -0.3801*** (0.1366)
       - Government balance, % of GDP (t-1): -0.0323*** (0.0125)
       - CPIA (t-1): -0.3090*** (0.1056)
       - IMF program (t): -0.3021** (0.1409)
       - Constant: 0.8648** (0.3614)
     - Findings from regressions:
       - Reserves have a statistically significant crisis prevention role.
       - Stronger fiscal position and better institutional quality (CPIA) reduce crisis probability.
       - Exchange rate flexibility and access to Fund financing following a large shock are associated with significantly lower crisis probability.
       - Marginal impact of reserves depends on exchange rate regime: increasing reserve coverage from three to four months of imports reduces crisis probability by about 3.5 percentage points for fixed exchange rate regimes, while a smaller reduction is achieved under flexible regimes.
  2. Regress magnitude of absorption drop (percent of GDP) against reserves ratio, exchange rate regime, and size of shocks, including country fixed effects (Panel OLS Regression, 1990-2007; No. of observations 418; R2 0.34).
     - Key regression estimates (Magnitude of Absorption Drop; Reserves are measured in logs):
       - Reserves, months of imports (t-1): -2.2403*** (0.6677)
       - Flexible exchange rate regime (t-1): -8.6983*** (2.1689)
       - External demand growth: -0.9320** (0.4356)
       - Terms of trade growth: -0.0841* (0.0484)
       - Change in FDI to GDP: -0.0159 (0.3391)
       - Change in aid to GDP: 0.0527 (0.0839)
     - Findings from regressions:
       - International reserves reduce the size of absorption losses in the event of external shocks.
       - Exchange rate flexibility facilitates adjustment to real shocks.
     - Robustness: regressions are largely robust to alternative estimation approaches and country samples (see Supplement).

### Calibration of optimal reserves: benefits, costs, and risk aversion
- Calibration approach:
  - Uses estimated regression coefficients, reference values for opportunity cost of holding reserves, and simplified assumptions about risk aversion.
  - Optimization framework assumes risk-neutrality (calibrated reserves should be considered a lower bound of optimal reserves under greater risk-aversion).
- Opportunity cost of holding reserves:
  - Several reference values considered due to lack of a clear-cut appropriate measure.
  - Reference values are based on existing estimates of the marginal product of capital and the differential between domestic and foreign real interest rates.
  - Caselli and Freyer (2007) calculate a range of 3 to 8 percent for the marginal product of capital in LICs.
- Sensitivity and illustrative results:
  - Calibrated optimal reserves vary with country characteristics and the cost of holding reserves.
  - Optimal reserve levels are generally higher for fixed exchange rate regimes, commodity exporters, and fragile states.
  - While the computed optimal reserves range encompasses the standard measure of three months of imports, this metric could be inadequate for countries with fixed exchange rate regimes.
  - For a representative LIC and assuming unit cost of holding reserves is set at 4 percent:
    - The "insurance" value of a flexible exchange rate regime is about 0.6 percent of GDP per year (or over three months of imports on average) in annual savings in the cost of holding optimal reserves.
    - Availability of (contingent) Fund support can result in annual savings in holding optimal reserves of about 0.3 percent of GDP per year (two months of imports).

### Role of policy fundamentals and heterogeneity
- Stronger fiscal positions are associated with lower optimal reserves (illustrative calibration, unit cost 4 percent).
- Trade-off between optimal reserves and policy fundamentals differs across exchange rate regimes, with higher sensitivity for fixed-rate regimes.
- Conclusion: a uniform metric (including the traditional rules of thumb) is not appropriate across all LICs given heterogeneity.

### Application to LICs and practical considerations
- The framework provides an empirically grounded metric for reserve-to-import ratios needed against exogenous shocks, but application requires judgment given heterogeneity.
- Practical supplements and caveats:
  - Capital flight risks in some LICs suggest supplementing this approach with the EM-type metric, consistent with evidence that broad money is a significant determinant of precautionary reserve holdings.
  - Calibrating using past shock history may fail to capture structural changes and new risk sources; forward-looking assessment is important for LICs undergoing rapid structural transformation and growing global integration.
  - For currency unions, assessment must account for institutional features and requirements of the currency union arrangement.
- Comparison of actual holdings against computed optimal reserves (based on end-2008 and end-2009 data):
  - On average, LIC holdings are broadly adequate.
  - Based on end-2008 data, LICs with fixed exchange regimes—particularly commodity exporters and fragile states—were, on average, below the computed adequacy range.
  - Countries with flexible regimes were well above the range, though with significant cross-country differences.
  - End-2009 picture differs slightly because the 2009 SDR allocation likely distorted reserve holdings for many LICs.
  - Caveats:
    - Flexible-regime countries tend to be more open and integrated with international financial markets; EM-like considerations may matter.
    - Nonprecautionary motives for holding reserves (monetary policy, exchange rate decisions) could be pertinent for managed float regimes.

### Lessons for advanced economies (AMs)
- Historically, little attention to reserve adequacy in AMs; AMs assumed to have strong market access and limited precautionary need for reserves.
- The recent global crisis showed relevance of reserves for AMs in some circumstances:
  - Currency mismatch and dollar shortage affected European financial institutions; lack of dollar reserves constrained central banks' lender-of-last-resort capabilities.
  - AMs with banks reliant on funding in euros and dollars faced severe funding/hedging stress when wholesale markets froze.
- Full insurance for AMs would imply impractically high reserve levels given large, globally integrated financial markets.
- Central bank swap lines can substitute for reserves in systemic events; less reliable for nonsystemic, idiosyncratic shocks.
- For some eurozone countries, larger fiscal buffers may be needed, but much higher reserve levels would likely have had limited effect where solvency, not liquidity, is the core issue.
- Metrics developed for EMs might be relevant in modified form for some AMs, but for most AMs reserve needs depend on detailed country-specific factors; scenario analysis and country-specific assessment are often necessary.

### Illustrative country experiences from the crisis
- Denmark:
  - Danmarks Nationalbank intervention and subsequent market stress; commercial paper market froze after Lehman Brothers; central bank doubled foreign exchange reserves since the crisis.
  - Reserves now stand at around US$80 billion (about 25 percent of GDP).
- Sweden:
  - Riksbank used Federal Reserve and ECB swap lines and borrowed about $15 billion externally to boost international reserves.
  - Current reserves about US$50 billion, around 9½ percent of GDP; prompted reassessment of reserve needs and investigation of bank resilience instruments.

*Source: Staff calculations and analysis in the provided IMF chapter/section.*

### 71.      The analysis presented in the paper suggests a multi-pronged approach.

### 71.      The analysis presented in the paper suggests a multi-pronged approach.

### Reserve adequacy: methodological findings
- Analyses of reserve adequacy should rely on country characteristics, directly exploring pressures against which reserves are held.
- For most EMs, which have tended to experience drains in the capital account during periods of exchange market pressure, simple rules of thumb should be complemented by a more comprehensive analysis.
- The paper suggests a new metric that builds on existing approaches to encompass various possible drains on reserves, and which appears to perform well as a more broad-based measure against which to assess reserves levels.
- A similar tractable model that relies on country-specific shocks is developed for LICs.
- Estimated models of reserve demand can complement these approaches as a means of comparing countries’ reserve levels against their peers from a precautionary perspective.

### Country-specific factors and complementary analyses
- Country-specific factors should be taken into account by considering additional measures or analyses.
- Country authorities and teams may wish to supplement adequacy assessments with alternative metrics and scenarios reflecting the relevant risk profiles.
- Reserves considerations could also take into account availability of contingent financing from the IMF—especially lines of credit with ex ante conditionality that emphasize the role of good policy fundamentals and institutions as the first line of defense.
- A direct examination of the authorities’ motives for reserve accumulation and their estimates of costs and other alternatives should also be considered.
- Scenario analysis is particularly useful if the perceived risk environment has changed or if new vulnerabilities have emerged.
- Several complementary approaches may thus be necessary to guide decisions on reserve adequacy.

### Issues for discussion (paragraph 73)
- Is a move beyond traditional metrics warranted?
- For EMs, is the risk-weighted approach to develop a broader yardstick against which to measure reserves appropriate? Does the suggested adequacy range against this measure seem reasonable?
- Does the proposed framework for LICs adequately capture factors bearing on reserves needs?
- Should advanced countries change significantly their approach to reserves? If so, how should it be calibrated?
- Do peer comparisons (cross-country reserve demand regressions) provide another satisfactory means of assessing adequacy?
- In evaluating adequacy of reserves how should issues relating to the management of reserves, and broader sovereign balance sheet risks, be factored into the analysis?

*Source: _021411b - 71.      The analysis presented in the paper suggests a multi-pronged approach.*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/np/pp/eng/2011/_021411b.pdf_