## _wp07103

## Source details

**Canonical URL:** [_wp07103](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2007/_wp07103.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2007/_wp07103.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2007/_wp07103.pdf.json)

---

### Introduction
- Jordan’s economy has strengthened substantially under successive Fund-supported programs.
- Key macro outcomes:
  - International reserves increased from less than US$½ billion in 1989 to more than US$6 billion at end-2006.
  - The public debt burden has been lowered substantially.
  - The banking system is stronger.
  - Inflation has been cut.
- Medium-term challenges:
  - Large current account deficit of 16 percent of GDP.
  - Vulnerability to sudden shifts in investor sentiment and regional political uncertainties, despite high-quality external financing mainly in the form of long-term foreign direct investment (FDI).
- Policy context:
  - The dinar has been pegged to the U.S. dollar since the creation of the dinar in 1950, except for a short period in the late 1980s.
  - Holding international reserves is desirable to deal with possible sudden stops and to support the exchange rate peg, and to provide comfort to market participants and foreign investors.

### Vulnerabilities and sudden stops
- Crises in emerging market economies provoked by "sudden stops" of capital inflows have become more prevalent since the 1990s; typically result in large currency devaluations and severe output losses.
- Roles of reserves:
  - Reserves can be used to roll over debt in a sudden stop crisis.
  - Reserves may also impact the probability of a sudden stop by acting as a crisis-preventing signal to investors.

### Methodology and model
- Analytical approach:
  - Cross-country analysis using traditional measures of reserves adequacy (rules of thumb).
  - Modified Jeanne-Rancière (J-R) model allowing reserves to both help rollover debt in a sudden stop and affect the probability of a sudden stop (πt modeled as a function of Rt).
- Model features and assumptions:
  - Single open economy in discrete time; Yt = Y0(1+g)^t.
  - Private consumption budget constraint includes private foreign debt Lt, government transfers Zt, and interest rate r.
  - Government issues only short-term debt Nt to finance reserves Rt (Rt = Pt Nt).
  - Price of government security: Pt = (1 + r + δ - πt)^-1.
  - Sudden stop: occurs with probability πt; private foreign borrowing Lt falls to zero and output experiences penalty γ (Yt = Yt-1 (1+g)(1-γ)).
  - Government smooths consumption by distributing reserves during sudden stops.
  - Government maximization with CRRA utility: u(Ct) = (Ct^(1-σ) - 1)/(1-σ), σ = coefficient of relative risk aversion.
  - First-order condition incorporates ∂πt/∂Rt, the marginal effect of reserves on sudden stop probabilities.
  - Simplifications: short-term government debt only; representative consumer; sudden-stop consequences assumed non-permanent.

### Key empirical finding: reserves reduce sudden-stop probability
- Probit regression (sample: 34 EM countries, 1980–2003) with dependent variable: occurrence of sudden stop (capital account as % of GDP falls by more than 5 percentage points year-on-year).
- Main estimated coefficients (standard errors in parentheses):
  - Reserves to Short Term Debt: -0.123 (0.047)**  
  - Real Exchange Rate (deviation from HP trend): 0.01 (0.008)  
  - Public Debt to GDP: 0.506 (0.227)*  
  - Foreign Liability to Money: -0.001 (0.046)  
  - Real GDP Growth: -2.329 (1.420)  
  - Gross Inflows to GDP: 6.781 (1.751)**  
  - Constant: -1.677 (0.205)**  
  - Observations: 647
- Notes on estimation:
  - + indicates significance at 10% level, * at 5% level, and ** at 1% level.
  - Standard errors clustered at country level.
  - Reserves to short term debt are lagged two periods.
  - Real GDP Growth is two-year average of t-1 and t-2.
  - All other explanatory variables lagged one period.
- Interpretation:
  - Negative and statistically significant relationship between reserves (as % of short-term debt) and sudden stop probability, with caveats about reverse causality and specification dependence.

### Calibration and optimal reserves for Jordan (baseline and scenarios)
- Baseline calibration parameters:
  - ∂πt/∂Rt taken from Table 2 marginal probabilities.
  - g = 6 percent.
  - γ = 9.2 percent.
  - r = 5 percent.
  - δ = 1.5 percent.
  - σ = 2.
  - λ = 0.11 (11 percent of GDP).
- Baseline model outcomes:
  - Optimal reserves ≈ 14 percent of GDP.
  - Monetary translation: optimal ≈ US$2¼ billion, equals just over 2¼ months of imports and about 10 percent of broad money.
  - Historical context: since mid-1990s reserves rose from about 15 percent of GDP to over 50 percent of GDP; actual reserves consistently higher than calculated optimum except for a few late-1980s periods.
- Sensitivity analysis (selected results):
  - Using the lower 95 percent confidence interval for reserve effect (beta = -0.19): optimal reserves increase by about 0.5 percentage point of GDP relative to baseline (and by about 2.5 percentage points vs the case where reserves have no impact).
  - Doubling risk aversion (sigma = 4): increases optimal reserves by about 0.6 percentage point of GDP; increasing sigma to 6 yields marginal additional increases.
  - Increasing output cost during sudden stop (gamma = 0.17): optimal reserves increase substantially (example: increases by 8 percent of GDP in 2004).
  - Larger sudden stop size (lambda = 0.38, as in 1992 Jordan episode): optimal reserves increase substantially, though still below actual holdings in baseline comparisons.
  - Combined severe shocks (gamma = 19.7 percent and lambda = 0.38): optimal reserves increase by 29 percentage points of GDP relative to baseline and approach actual holdings.
  - Adding both severe shocks and a stronger crisis-preventing effect of reserves can reduce optimal reserves in recent years because very high reserves dramatically lower sudden-stop probability, reducing consumption-smoothing need.
- Model extensions that raise optimal reserves:
  - Heterogeneous consumers (two-type approximation): raises optimal reserves by about 1 percentage point of GDP.
  - Persistent output loss over multiple periods: two-period loss raises optimal reserves by about 3½ percentage points of GDP; three periods adds a further 3½ percentage points.
  - Uncertainty about sudden-stop consequences (with risk aversion): likely increases optimal reserves in an order similar to heterogeneity effects.
  - Combined extensions could raise baseline optimal reserves by about 9 percentage points of GDP.
- Bottom-line numeric mappings and policy context:
  - Modified model extensions alone suggest optimal reserves around 23 percent of GDP (≈ a little over US$3½ billion, or 3¾ months of imports).
  - Given the exchange rate peg, sensitivity analysis outcomes, and traditional reserve indicators, policymakers may reasonably choose reserves above these modified levels for added caution.
  - Overall conclusion: Jordan’s reserve holdings are sufficient to support the dinar peg and withstand extreme capital account disruptions.

### Traditional measures of reserves adequacy (cross-country comparison)
- Reserve cover metrics for Jordan:
  - Reserves cover short-term external debt (by remaining maturity) more than seven times.
  - Reserves/broad money: Jordan falls in the lower half of the EM sample.
  - Reserves/FX deposits: Jordan falls in the lower half of the EM sample (reflecting financial depth and extensive dollarization).
  - Reserves/imports: Jordan is in the upper half of the sample.
  - Comprehensive “reserve cover” (short-term debt + current account deficit adjusted for FDI inflows + foreign currency deposits): Jordan has far lower coverage.
- Note: traditional indicators do not individually capture all motivations for holding reserves; they provide background for model-based analysis.

### Policy-relevant implications
- Empirical and model evidence support precautionary reserve holdings to:
  - Roll over debt during sudden stops.
  - Reduce the probability of sudden stops by signaling sound external liquidity.
- Quantitative guidance for policymakers:
  - Baseline optimal reserve target ≈ 14 percent of GDP (≈ US$2¼ billion).
  - Extended-model and adverse-shock considerations can justify raising targets toward ≈ 23 percent of GDP (≈ a little over US$3½ billion).
  - Given the dinar peg and potential for extreme shocks, maintaining reserves above model optima can be prudent.

*Source: Excerpt from _wp07103 - Section V concludes (IMF PDF).*

### References..............................................................................................................

### _wp07103 - References

### Introduction
- Jordan’s economy has strengthened substantially in recent years under successive Fund-supported programs.
- Key macro outcomes:
  - International reserves increased from less than US$½ billion in 1989 to more than US$6 billion at end-2006.
  - The public debt burden has been lowered substantially.
  - The banking system is stronger.
  - Inflation has been cut.
- Despite accomplishments, Jordan faces medium-term challenges:
  - Large current account deficit of 16 percent of GDP, one of the highest in emerging market (EM) economies.
  - Vulnerability to sudden shifts in investor sentiment and regional political uncertainties, despite high-quality external financing mainly in the form of long-term foreign direct investment (FDI).
- Policy context:
  - The dinar has been pegged to the U.S. dollar since the creation of the dinar in 1950, except for a short period in the late 1980s.
  - Holding international reserves is desirable to deal with possible sudden stops and to support the exchange rate peg, and to provide comfort to market participants and foreign investors.

### Vulnerabilities and sudden stops
- Definition and prevalence:
  - Footnote: "Crises in emerging market economies provoked by 'sudden stops' of capital inflows have become more prevalent since the 1990s. Typically, these crises result in large currency devaluations and severe output losses (Calvo, 2006)."
- Role of reserves:
  - Reserves can be used to roll over debt in a sudden stop crisis.
  - Reserves may also impact the probability of a sudden stop by acting as a crisis-preventing signal to investors (cited literature: Hashimoto, 2007; Garcia and Soto, 2004; Chamon, Manasse, and Prati, 2006).

### Methodology and model
- Analytical approach:
  - Cross-country analysis using widely-used traditional measures of reserves adequacy (rules of thumb).
  - Formulation of a modified J-R model that allows reserves to both help rollover debt in a sudden stop and affect the probability of a sudden stop.
- Rationale for model choice:
  - The J-R approach incorporates the crisis-moderating effect of reserves; if reserves reduce the probability of crises, this effect should influence the optimal level of reserves.

### Key findings
- Traditional metrics:
  - Traditional measures of reserves adequacy (rules of thumb) do not suggest that Jordan’s reserve holdings are out of line with its peers.
  - Jordan’s reserve holdings are the highest in the EM sample compared to its short-term debt exposure.
- Model-based estimates:
  - Model-based estimates of “optimal” reserves are much lower than actual reserve holdings.
- Robustness to adverse scenarios and extensions:
  - Incorporating possible severe adverse shocks (including large sudden stops and associated large output reductions), and sensible model extensions—factoring in heterogeneous consumers, longer-lasting output responses to sudden stops, and uncertainty—suggests that the gap between optimal and actual reserve holdings may not be that large.
- Overall assessment:
  - The detailed analysis provides support for the view that Jordan’s reserve holdings are sufficient to cope with the most extreme of economic circumstances and therefore provide important backing for the dinar peg to the U.S. dollar.

### Structure of the paper (excerpted)
- Section II: Examines Jordan’s reserve holdings in a cross-country context using several widely-used rules of thumb and simple reserve level comparisons.
- Section III: Introduces a modified J-R model that incorporates the crisis-preventing attributes of reserves.
- Section IV: Calibrates the model and provides optimal reserve estimates for Jordan.

*Source: Excerpt from _wp07103 - References (IMF PDF).*

### Section V concludes.

### _wp07103 - Section V concludes.

### Traditional Measures of Reserves Adequacy
- Jordan’s reserves cover short-term external debt (by remaining maturity) more than seven times.
- Reserve cover relative to:
  - Broad money: Jordan falls in the lower half of the EM sample.
  - Foreign exchange deposits: Jordan falls in the lower half of the EM sample (reflecting financial depth and extensive dollarization).
  - Imports: Jordan is in the upper half of the sample.
  - A comprehensive “reserve cover” (short-term debt + current account deficit adjusted for FDI inflows + foreign currency deposits): Jordan has far lower coverage.
- The section presents a taxonomy of traditional indicators (Table 1) including:
  - Reserves/short-term external debt
  - Reserves/broad money
  - Reserves/FX deposits
  - Reserves/imports (guideline: at least three to four months of imports for less developed economies)
  - Reserve cover (short-term debt + current account deficit adjusted for FDI + FX deposits)
- Note: traditional indicators do not individually capture all motivations for holding reserves; they provide background for model-based analysis.

### Modified J‑R Model (reservation and crisis-prevention aspects)
- The model builds on Jeanne and Rancière (J-R) but adds an extension: reserves can reduce the probability of a sudden stop (πt is modeled as a function of Rt).
- Key modeling assumptions and structure:
  - Single open economy in discrete time; domestic production grows at rate g with Yt = Y0(1+g)^t.
  - Private consumption budget constraint (equation (1)) includes private foreign debt Lt, government transfers Zt, and interest rate r.
  - Government issues only short-term debt Nt to finance reserves Rt (Rt = Pt Nt).
  - Price of government security: Pt = (1 + r + δ - πt)^-1 (expressed in the source as equation (2) with δ term premium and πt).
  - Sudden stop: occurs with probability πt; private foreign borrowing Lt falls to zero and output experiences penalty γ (Yt = Yt-1 (1+g)(1-γ)).
  - Government smooths consumption by distributing reserves during sudden stops.
- Government maximization problem framed with CRRA utility: u(Ct) = (Ct^(1-σ) - 1)/(1-σ), σ = coefficient of relative risk aversion.
- First-order condition (equation (8)) incorporates marginal effect of reserves on sudden stop probabilities ∂πt/∂Rt.
- Important simplifications: government debt is short-term for tractability; representative consumer assumed; sudden stop consequences assumed non-permanent for analytical simplicity.

### Econometric Results: Effect of Reserves on Preventing Crises
- Empirical approach:
  - Probit regression using sample of 34 emerging market countries (including Jordan) from 1980 to 2003.
  - Dependent variable: occurrence of a sudden stop (capital account as % of GDP falls by more than 5 percentage points year-on-year).
  - Key regressors (lagged): reserves (excluding gold) as fraction of short-term external debt (lagged two periods), real effective exchange rate deviation from HP trend, public debt to GDP, foreign liabilities in banking sector to money, absolute value of gross inflows to GDP, and averaged GDP growth.
- Main regression results (Table 2 coefficients; standard errors in parentheses):
  - Reserves to Short Term Debt: -0.123 (0.047)**  
  - Real Exchange Rate (deviation from HP trend): 0.01 (0.008)  
  - Public Debt to GDP: 0.506 (0.227)*  
  - Foreign Liability to Money: -0.001 (0.046)  
  - Real GDP Growth: -2.329 (1.420)  
  - Gross Inflows to GDP: 6.781 (1.751)**  
  - Constant: -1.677 (0.205)**  
  - Observations: 647
  - Notes: + indicates significance at 10% level, * at 5% level, and ** at 1% level. Standard errors clustered at country level. Reserves to short term debt are lagged two periods. Real GDP Growth is two-year average of t-1 and t-2. All other explanatory variables lagged one period.
- Interpretation:
  - A negative and statistically significant relationship between reserves (as % of short-term debt) and sudden stop probability is found, interpreted cautiously given potential reverse causality and specification dependence.

### Calibration and Optimal Reserves for Jordan
- Baseline calibration choices:
  - ∂πt/∂Rt taken from Table 2 marginal probabilities.
  - g (average real growth rate) = 6 percent (Jordan-specific, vs J-R EM average 3.3 percent).
  - γ (average output loss during a sudden stop) = 9.2 percent.
  - r (risk-free rate) = 5 percent.
  - δ (term premium) = 1.5 percent.
  - σ (risk aversion) = 2.
  - λ (net capital flows decrease during sudden stop for representative EM) = 0.11 (11 percent of GDP).
- Baseline model outcome:
  - Baseline model suggests optimal reserves ≈ 14 percent of GDP (Table 3, column 2).
  - Translating to monetary values: optimal ≈ US$2¼ billion, which equals just over 2¼ months of imports and about 10 percent of broad money.
  - Jordan’s actual reserves have been consistently higher than the calculated optimum except for a few late-1980s periods; since mid-1990s reserves rose from about 15 percent of GDP to over 50 percent of GDP.
- Sensitivity analysis results (summary of key scenarios from Table 3):
  - Using the lower 95 percent confidence interval for the reserve effect (beta = -0.19): optimal reserves increase by about 0.5 percentage point of GDP relative to baseline (and by about 2.5 percentage points vs the case where reserves have no impact).
  - Doubling risk aversion (sigma = 4): increases optimal reserves by about 0.6 percentage point of GDP; increasing sigma to 6 yields marginal additional increases.
  - Increasing output cost during sudden stop (gamma = 0.17, i.e., 17 percent output loss): optimal reserves increase substantially (example: increases by 8 percent of GDP in 2004).
  - Larger sudden stop size (lambda = 0.38, i.e., 38 percent of GDP not rolled over, as in 1992 Jordan episode): optimal reserves increase substantially, though still below actual holdings in baseline comparisons.
  - Combined severe shocks (gamma = 19.7 percent output loss and lambda = 0.38): optimal reserves increase by 29 percentage points of GDP relative to baseline and approach actual holdings.
  - Adding both severe shocks and a stronger crisis-preventing effect of reserves can reduce optimal reserves in recent years because very high reserves dramatically lower sudden-stop probability, reducing consumption-smoothing need.
- Model extensions increasing optimal reserves:
  - Heterogeneous consumers: a two-type approximation raises optimal reserves by about 1 percentage point of GDP.
  - Persistent output loss (sudden-stop costs lasting multiple periods): output loss over two periods raises optimal reserves by about 3½ percentage points of GDP; over three periods adds a further 3½ percentage points.
  - Uncertainty about sudden-stop consequences (with risk aversion): likely increases optimal reserves in an order similar to heterogeneity effects.
  - Combined extensions could raise baseline optimal reserves by about 9 percentage points of GDP.
- Bottom-line numeric mappings and policy context:
  - Modified model extensions alone suggest optimal reserves around 23 percent of GDP (≈ a little over US$3½ billion, or 3¾ months of imports).
  - Given the exchange rate peg, sensitivity analysis outcomes, and traditional reserve indicators, policymakers may reasonably choose reserves above these modified levels for added caution.
  - Overall conclusion: Jordan’s reserve holdings are sufficient to support the dinar peg and to withstand extreme capital account disruptions.

### Conclusions (Section V)
- The paper develops a formal framework weighing costs and benefits of reserves for Jordan, combining traditional indicators and a reserves-optimizing model (based on Jeanne and Rancière (2006) with an added effect of reserves on sudden-stop probabilities).
- Key takeaways:
  - Traditional metrics do not indicate Jordan’s reserves are out of line with peers, though simple model-based optimal reserves appear lower than actual holdings.
  - Translating model outputs into common metrics, comprehensive sensitivity analysis, and plausible model extensions narrow the gap between optimal and actual reserves.
  - Formal analysis supports the view that Jordan’s reserve holdings provide solid support for the dinar peg and are adequate to handle extreme capital account disruptions.

*Source: _wp07103 - Section V concludes.*

### REFERENCES

### _wp07103 - REFERENCES

### International reserves, precautionary motives, and mercantilism
- Aizenman, Joshua and Jaewoo Lee, 2006, “Financial Versus Monetary Mercantilism: Long-Run View of Large International Reserves Hoarding,” IMF Working Paper 06/280 (Washington: International Monetary Fund).
- Aizenman, Joshua and Jaewoo Lee, “International Reserves: Precautionary versus Mercantilist Views, Theory and Evidence,” forthcoming Open Economies Review.
- Flood, Robert and Nancy Marion, 2002, “Holding International Reserves in an Era of High Capital Mobility,” IMF Working Paper 02/62 (Washington: International Monetary Fund).
- Garcia, Pablo S. and Claudio Soto, 2004, “Large Holding of International Reserves: Are They Worth it?” (mimeo; Chile: Bank of Chile).
- Hashimoto, Yuko, 2007,“Too Much Self-Insurance? Asian Foreign Reserves,” (mimeo; Tokyo: Tokyo University).
- Jeanne, Olivier and Romain Rancière, 2006, “The Optimal Level of International Reserves for Emerging Market Countries: Formulas and Applications,” IMF Working Paper 06/229 (Washington: International Monetary Fund).
- Summers, Lawrence H, 2006, “Reflections on Global Account Imbalances and Emerging Markets Reserve Accumulation,” paper presented at the L.K. Jha Memorial Lecture, Reserve Bank of India, Mumbai, India, March.

### External vulnerability, liquidity, and contagion in emerging markets
- Bussière, Matthieu and Christian Mulder, 1999, “External Vulnerability in Emerging Market Economies: How High Liquidity can Offset Weak Fundamentals and the Effects of Contagion,” IMF Working Paper 99/88 (Washington: International Monetary Fund).
- Mody, Ashoka and Mark P. Taylor, 2002, “International Capital Crunches: The Time Varying Role of Information Asymmetries,” IMF Working Paper 02/43 (Washington: International Monetary Fund).
- Calvo, Guillermo A., 2006, “Monetary Policy Challenges in Emerging Markets: Sudden Stop, Liability Dollarization, and Lender of Last Resort,” NBER Working Paper No. 12788 (Massachusetts: National Bureau of Economic Research).
- Sachs, Jeffrey, Aaron Tornell, and Andres Velasco, 1996, “Financial Crises in Emerging Markets: The Lessons from 1995,” Brookings Papers on Economic Activity, Vol. 27, No. 1, pp. 147–99.

### Crisis models, capital account crises, and recovery
- Krugman, Paul, 1979 “A Model of Balance-of-Payments Crises,” Journal of Money, Credit and Banking, Ohio State University Press, Vol. 11(3), pp. 311–25.
- Chamon, Marcos, Paolo Manasse, and Alessandra Prati, 2006, “Can We Predict the Next Capital Account Crisis?” paper presented at the seventh Jacques Polak Annual Research Conference, November.
- Cerra, Valerie and Sweta Chaman Saxena, 2005, “Did Output Recover from the Asian Crisis?” Staff Papers, International Monetary Fund, Vol. 52, Issue 1, pp. 1–23.

### Standards, market indicators, and exchange spreads
- Christofides, C., Christian B. Mulder, and Andrew J. Tiffin, 2003, “The Link Between Adherence to International Standards of Good Practices, Foreign Exchange Spreads, and Ratings,” IMF Working Paper 03/74 (Washington: International Monetary Fund).

*Source: _wp07103 - REFERENCES*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2007/_wp07103.pdf_