## _wp0940

## Source details

**Canonical URL:** [_wp0940](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2009/_wp0940.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2009/_wp0940.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2009/_wp0940.pdf.json)

---

### I. Introduction
- Two waves of large capital inflows to emerging market economies over the past two decades:
  - First wave: commenced in the early 1990s and ended with the Asian crisis in 1997.
  - Second wave: started in 2003, ebbed in 2008 in the wake of the global financial crisis.
- Benefits and challenges:
  - Benefits: economic gains from increased financial integration.
  - Challenges: potential overheating, loss of competitiveness, increased vulnerability to crisis.
- Common policy responses:
  - Allow exchange rates to appreciate or intervene to resist appreciation.
  - Sterilization to neutralize monetary impact of intervention.
  - Introduce/tighten controls on capital inflows; ease controls on outflows.
  - Fiscal responses varied; rising revenues often led to higher government spending.
- Paper objectives:
  - Review experience with large capital inflows across many emerging market and advanced economies.
  - Characterize policy responses and assess macroeconomic implications.
  - Address three questions: macroeconomic implications; policy challenges; which policy measures were adopted and whether they worked (including mitigation of abrupt reversals).

### II. Contributions and High-Level Findings
- Sample and identification:
  - Identified 109 episodes of large net private capital inflows to 52 countries over 1987–2007, of which 87 episodes were completed by 2006.
- Macro outcomes associated with inflow episodes:
  - Episodes associated with acceleration of GDP growth, followed by frequently significant declines after inflows abated.
  - Post-inflow decline in GDP growth is significantly larger for episodes that end abruptly.
  - Of the 87 completed episodes, 34 ended with a sudden stop and 13 with a currency crisis.
  - Large swings in aggregate demand and current account balances: strong deterioration during inflow period and sharp reversal at the end.
  - End of episodes typically involved a sharp reversal of non-FDI flows; FDI proved more resilient.
  - Surge in inflows associated with real effective exchange rate appreciation.
- Four policy-related results:
  1. Countries with more volatile macroeconomic fluctuations tend to have higher current account deficits and stronger increases in aggregate demand and real currency value during inflows.
  2. Episodes with more moderate declines in GDP growth post-inflow are associated with greater fiscal restraint during the inflow period—containing aggregate demand and limiting real appreciation. Specifically, keeping public expenditure growth steady rather than ratcheting up spending can help contain currency appreciation and foster better outcomes.
  3. Countries resisting nominal exchange rate appreciation through intervention were generally not able to moderate real appreciation when inflows persisted, and faced more serious adverse macroeconomic consequences when surges stopped.
  4. Tightening capital controls generally has not been associated with lower real appreciation nor with reduced vulnerability to sharp reversal of inflows.

### III. Database and Construction of NPCIR
- Coverage:
  - Annual data over 1985–2007 for 52 countries: 8 advanced and 44 developing.
- Definition and construction of net private capital inflows (NPCIR):
  - Based on changes in foreign assets and liabilities of the domestic private sector recorded in the IMF’s BOP database; excludes sovereign borrowing and includes private borrowing from external official sources.
  - Five-step construction:
    1. Calculate (net) FDI: direct investments into the recipient country minus direct investments abroad.
    2. Strip out assets classified under monetary authority and general government for remaining categories.
    3. Net the series of private assets and liabilities to yield net inflows for portfolio, financial derivatives, and other investments.
    4. Add FDI to net private portfolio investment, financial derivative, and other investment categories → net private capital inflows.
    5. Scale total net private capital inflows by GDP → net private capital inflows-to-GDP ratio (NPCIR).

### IV. Methodology for Identifying Episodes
- Two complementary criteria:
  - Country-specific: NPCIR must be significantly larger than the trend for that country.
  - Regional: NPCIR must exceed the 75th percentile of the regional NPCIR distribution.
- Trend extraction:
  - Rolling, backward-looking Hodrick-Prescott (HP) filter with smoothing parameter 1000; initial five years used to determine trend; for short series (mostly Central and Eastern Europe) HP filter applied to entire series with smoothing parameter 100.
- Episode identification rules:
  - Year t is an episode year if either:
    - Deviation of NPCIR from its trend in year t > one historical standard deviation and NPCIR exceeds 1 percent of GDP, or
    - NPCIR exceeds the 75th percentile of the regional distribution.
  - Episodes begin in the first year meeting criteria and continue while subsequent years meet them.
  - Clustering rules: no episodes identified in the two years prior; adjacent or one-year-separated episodes sometimes combined depending on NPCIR sign.
- Defining abrupt endings:
  - Episode ends "abruptly" if NPCIR in the year after termination is more than 5 percentage point of GDP lower than at the end of the episode.
  - Also abrupt if end coincides with a currency crisis: depreciation of at least 25 percent cumulative over 12 months and at least 10 percentage points greater than in the preceding 12 months.

### V. Stylized Facts on Episodes and the Second Wave
- Two waves in NPCIR:
  - First wave: early 1990s, ending with 1997–98 Asian crisis.
  - Second wave: building since 2002, accelerated in 2007 with inflows far exceeding prior experience.
- Characteristics of the second wave:
  - Occurred with generally stronger current account positions for most emerging markets and substantial acceleration in accumulation of foreign reserves.
  - Accompanied by a sharp increase in outflows due to global diversification of portfolios.
  - Predominance of net FDI flows relative to net “financial” flows in all four regions studied.
  - Differentiators from the 1990s: larger set of countries involved; generally more solid current account positions (exception: emerging European countries); more financially integrated world where outflows partially offset inflows.
- Temporal and regional patterns:
  - Episodes completed between 1987 and 1998 generally involved smaller volume relative to GDP but lasted longer than episodes ended between 1999 and 2006.
  - Emerging Asia and Latin America dominated the first wave; more recent episodes concentrated in emerging Europe and other emerging market countries.
  - Of 87 completed episodes, 34 ended with a sudden stop and 13 with a currency crisis; in 7 episodes a sudden stop coincided with a currency crisis.

### VI. Policy Responses — Measures and Stylized Behavior
- Overview of policy toolkit used during large inflows:
  - Exchange rate intervention and reserve accumulation.
  - Sterilization to offset monetary impact of reserve accumulation.
  - Fiscal policy adjustments (often procyclical increases in public expenditure).
  - Capital controls on inflows and outflows.
- Exchange market pressure and resistance index:
  - EMP_{i,t} = Δ%er_{i,t}/σ_{Δ%er_{i,t}} + Δres_{i,t}/σ_{Δres_{i,t}}.
  - Resistance Index_{i,t} = 1 − ( (Δ%er_{i,t}/σ_{Δ%er_{i,t}}) / EMP_{i,t} ), standardized into discrete values {0, 0.2, 0.4, 0.6, 0.8, 1} per raw index ranges.
  - Interpretation: 0 = no resistance; 1 = maximum resistance; intermediate values = proportion of EMP relieved by intervention.
- Sterilization index:
  - Regression ΔNDA_{i,t} = α_{i,t} + β_{i,t} ΔNFA_{i,t} + u_{i,t} using 12 monthly observations; sterilization index = –β_{i,t}.
  - Sterilization index = 1 implies full sterilization; = 0 implies no sterilization; can exceed 1 (over-sterilization).
  - Alternative broader index uses ΔM2 regression; authors focus on NDA-based index.
- Fiscal policy indicators:
  - Fiscal stance represented by change in growth of real noninterest government expenditure (nominal noninterest expenditures scaled by CPI; interest expenditures subtracted).
  - Complementary HP-filtered deviation measure used for structural characterization.
- Capital controls measurement:
  - Index based on IMF AREAER covering nine control dimensions; averaged to form inflow and outflow control indices.
  - AREAER allows distinguishing inflow vs outflow controls from 1995 onward; older vintages backcast.
  - Limitations: enforcement intensity not captured; measures coarse.

### VII. Stylized Facts on Policy Behavior During Episodes
- Median tendencies during episodes:
  - Resistance index tends to increase during an episode; statistically significant for episodes completed before 1998.
  - Sterilization does not tend to increase during an episode relative to the two years beforehand (temporary sterilization efforts often unsustainable).
  - Real government expenditures tend to increase strongly as capital inflows surge (fiscal policy generally procyclical).
  - Controls on inward capital flows: tightened during episodes completed before 1998 (not always significantly); during more recent episodes controls appear to have been eased.
  - Controls on outflows: no relaxation coinciding with surges for completed episodes; restrictions appear less strict during ongoing episodes.
- Regional patterns:
  - Emerging Europe and the CIS moved furthest toward liberalization; emerging Asia remained relatively restrictive.
  - Controls on outflows progressively loosened in emerging Europe and the CIS and other emerging markets; more recently in emerging Asia and Latin America.
- Intervention asymmetry:
  - Foreign reserves increased over the 1990s, indicating tendency to intervene to prevent appreciation more than depreciation (except during crises with large reserve reductions).

### VIII. Empirical Evidence — Outcomes and Policy Links
- Stylized macro facts:
  1. Large inflows associated with acceleration of GDP growth; afterwards growth often drops significantly.
     - For episodes that end abruptly: average GDP growth in the two years after the end tends to be about 3 percentage points lower than during the episode, and about 1 percentage point lower than during the two years before the episode.
  2. GDP growth fluctuations accompanied by large swings in aggregate demand and current account; strong deterioration during inflow period and sharp reversal at the end.
  3. End of inflow episodes typically entails sharp reversal of non-FDI flows; FDI more resilient.
  4. Surge associated with real effective exchange rate appreciation; median differences not always statistically significant; appreciation has not on average been generated by higher inflation.
- Hard landing correlates:
  - Faster acceleration in domestic demand, sharper rise in inflation, larger real appreciation during inflow period.
  - Longer episodes with much higher cumulative inflows (examples cited: Thailand 1988–96; Argentina 1992–94 and 1997–99; Mexico 1990–94).
  - Hard landings associated with strong increases in government spending during inflow period; expenditure restraint helps reduce upward pressures.
- Resistance and sterilization:
  - Higher resistance and greater sterilization were generally unable to prevent real appreciation and often unsuccessful in achieving soft landing; sterilization can induce higher nominal interest rates and perpetuate inflows.

### IX. Regression Results (selected coefficients and inference)
- Post-inflow GDP Growth regressions (Observations: 69; adjusted R-squared: 0.133–0.188 across specifications):
  - Real Government Expenditure: coefficients −0.109 (0.015)**; −0.111 (0.014)**; −0.111 (0.014)**; −0.099 (0.027)**; −0.093 (0.040)**.
  - Index of Resistance to Exchange Market Pressures: coefficients −1.812 (0.114); −2.090 (0.085)*; −2.086 (0.088)*; −2.147 (0.080)*; −2.282 (0.059)*.
  - Post-inflow World Output Growth: coefficients 1.023 (0.017)**; 0.836 (0.056)*; 0.858 (0.071)*; 0.875 (0.063)*; 0.844 (0.076)*.
  - Sterilization Index: −0.981 (0.262) in one specification.
  - Cumulative Size of Capital Inflow: −0.049 (0.148) in one specification.
  - Robust p values in parentheses; * significant at 10 percent; ** significant at 5 percent.
- Real Effective Exchange Rate Appreciation regressions (Observations: 106–107; adjusted R-squared: 0.115–0.227):
  - Real Government Expenditure: coefficients 0.544 (0.003)***; 0.396 (0.029)**; 0.321 (0.071)*; 0.307 (0.112).
  - Output gap: 0.954 (0.050)**; 0.715 (0.094)*; 0.654 (0.130).
  - Real U.S. Fed Funds Rate: 0.492 (0.604); 1.606 (0.100)*; 1.755 (0.078)*.
  - Cumulative Size of Capital Inflow: 0.241 (0.083)*; 0.249 (0.074)*.
  - Sterilization Index included in one specification: 2.562 (0.289).
  - Robust p values in parentheses; * significant at 10 percent; ** significant at 5 percent; *** significant at 1 percent.
- Interpretation:
  - Greater real government expenditure growth during inflow episodes is associated with worse post-inflow GDP growth and larger real appreciation.
  - Resistance to exchange market pressures shows negative coefficients for post-inflow growth (often significant at 10 percent) but does not robustly explain real appreciation.
  - Sterilization not robustly associated with better post-inflow outcomes.

### X. Role and Effectiveness of Capital Controls
- Empirical patterns:
  - Episodes with tighter controls on inflows are associated with narrower current account deficits and lower net private inflows, including lower net FDI.
  - Stricter inflow controls accompanied by lower post-inflow growth and larger appreciation, but distinctions not statistically significant; inflation rates significantly higher in episodes with tighter controls.
  - Episodes ending in abrupt reversals do not seem associated with lower capital controls.
- Regression evidence:
  - More restrictive controls on inflows may help avoid hard landings in some specifications; none of the capital control measures is statistically significant in the real exchange rate equations.
  - Using Edwards (2005) measure: only statistically significant for more parsimonious post-inflow growth regressions.
- Microeconomic costs and practical experience:
  - Controls often did not affect volume of net flows; could temporarily tilt composition toward longer maturities in a few cases (example: Chile after 1991).
  - Stricter outflow controls helped in Malaysia after 1998 but little support elsewhere.
  - Sustained capital controls associated with microeconomic costs: harder/expensive for small firms to raise capital; higher local borrowing costs for multinationals; distortions and resources spent to circumvent controls; increased trade costs due to inspection/reporting.
- Overall assessment: mixed evidence; capital controls not a substitute for sound macroeconomic policies (prudent fiscal stance; supporting exchange rate and monetary framework; appropriate prudential measures).

### XI. Persistence, Initial Imbalances, and Regional Factors
- Duration effects:
  - Episodes lasting less than two years display larger resistance to exchange rate changes, less real appreciation, and better post-inflow GDP growth.
  - However, higher resistance in short episodes was not associated with significantly smaller real appreciation or better post-inflow growth.
  - Resisting exchange market pressures may be more feasible for transitory inflows but does not guarantee better outcomes and ex ante identification of transitory episodes is difficult.
- Balance-of-payments pressures:
  - Fiscal restraint less decisive in episodes with high BoP pressures (above-median sum of current account and net private capital inflows): lower government spending growth not associated with significantly better outcomes.
  - Fiscal spending restraint associated with significantly better outcomes when episodes have low BoP pressures.
- Regional patterns:
  - Regions with stronger real appreciation during episodes (Latin America; emerging Europe and the CIS) also had larger increases in public expenditure.
  - Advanced economies followed more countercyclical fiscal policies and refrained from resisting exchange market pressures; they experienced less real appreciation and smaller GDP growth fluctuations.

### XII. Conclusions — Policy Implications and Guidance
- Four general patterns:
  1. Countries with relatively high current account deficits more vulnerable to sharp reversals of capital inflows.
  2. Public expenditure restraint during inflow episodes contributes to lower real exchange rate appreciation and better post-inflow GDP growth.
  3. Resistance to nominal exchange rate appreciation has not generally prevented real appreciation and has often been followed by sharper reversals, especially for persistent inflows.
  4. Restrictions on capital inflows have generally not been associated with lower real appreciation or softer landings.
- Policy guidance:
  - Stabilization challenges most serious for countries with substantial current account imbalances.
  - In contexts of relatively inflexible exchange rate policies, fiscal restraint is likely the most effective tool to avoid overheating and output instability.
  - If central banks initially resist nominal appreciation, that stance should be progressively relaxed if inflows persist to reduce the likelihood of eventual real appreciation and costly reversals.
  - Policy responses should consider country-specific circumstances: nature and persistence of inflows; stage of business cycle; fiscal situation; quality of domestic financial markets.
  - Other relevant tools not systematically analyzed here: financial supervision and regulation, labor and product market reforms.

*Source: _wp0940 - References..............................................................................................................*

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

### _wp0940 - References..............................................................................................................

### I. Introduction
- Two waves of large capital inflows to emerging market economies over the past two decades:
  - First wave: commenced in the early 1990s and ended with the Asian crisis in 1997.
  - Second wave: started in 2003, ebbed in 2008 in the wake of the global financial crisis.
- Benefits and challenges of capital inflows:
  - Benefits: deliver economic gains from increased financial integration.
  - Challenges: potential to generate over-heating, loss of competitiveness, and increased vulnerability to crisis.
- Common policy responses to large inflows:
  - Allow exchange rates to appreciate, or intervene heavily to resist appreciation.
  - Sterilization to neutralize monetary impact of intervention.
  - Introduce or tighten controls on capital inflows; ease controls on outflows.
  - Fiscal responses varied: stronger revenue growth sometimes harnessed to improve fiscal outcomes; in many cases, rising revenues led to higher government spending.
- Paper objectives:
  - Review experience with large capital inflows across many emerging market and advanced economies.
  - Characterize policy responses and assess macroeconomic implications.
  - Address three questions: macroeconomic implications; policy challenges; which policy measures were adopted and whether they worked (including mitigation of abrupt reversals).

### II. Contributions and High-Level Findings
- Sample and identification:
  - Methodology identifies 109 episodes of large net private capital inflows to 52 countries over 1987–2007, of which 87 episodes were completed by 2006.
- Macro outcomes associated with inflow episodes:
  - Episodes associated with acceleration of GDP growth, followed by frequently significant declines after inflows abated.
  - Post-inflow decline in GDP growth is significantly larger for episodes that end abruptly.
  - Of the 87 completed episodes, 34 ended with a sudden stop and 13 with a currency crisis.
  - Large swings in aggregate demand and current account balances: strong deterioration of the current account during inflow period and sharp reversal at the end.
  - End of episodes typically involved a sharp reversal of non-FDI flows; FDI proved more resilient.
  - Surge in inflows associated with real effective exchange rate appreciation.
- Four policy-related results:
  1. Countries with more volatile macroeconomic fluctuations (including sharp reversals) tend to have higher current account deficits and stronger increases in aggregate demand and real currency value during inflows.
  2. Episodes with more moderate declines in GDP growth post-inflow are associated with greater fiscal restraint during the inflow period—containing aggregate demand and limiting real appreciation. Specifically, keeping public expenditure growth steady rather than ratcheting up spending can help contain currency appreciation and foster better outcomes.
  3. Countries resisting nominal exchange rate appreciation through intervention were generally not able to moderate real appreciation when inflows persisted, and faced more serious adverse macroeconomic consequences when surges stopped.
  4. Tightening capital controls generally has not been associated with lower real appreciation nor with reduced vulnerability to sharp reversal of inflows.

### III. Database
- Coverage:
  - Annual data over 1985–2007 for 52 countries: 8 advanced and 44 developing.
  - Developing group includes many emerging market economies; industrial countries correspond to a sub-sample of small open OECD economies.
- Data sources:
  - IMF’s International Financial Statistics, supplemented by IMF’s World Economic Outlook and Balance of Payments databases.
- Definition of net private capital inflows:
  - Based on nature of recipient sector: changes in foreign assets and liabilities of the domestic private sector recorded in the IMF’s BOP database.
  - Excludes sovereign borrowing (changes in government assets and liabilities vis-à-vis the foreign private sector) and includes private borrowing from external official sources.
- Construction of net private capital inflows series (five steps):
  1. Calculate (net) FDI: direct investments into the recipient country minus direct investments abroad.
  2. Strip out assets classified under monetary authority and general government for remaining categories: portfolio investments, financial derivatives, and other investments.
  3. Net the series of private assets and liabilities to yield net inflows for the three categories.
  4. Add FDI to net private portfolio investment, financial derivative, and other investment categories → definition of net private capital inflows.
  5. Scale total net private capital inflows by GDP → net private capital inflows-to-GDP ratio (NPCIR).

### IV. Methodology for Identifying Episodes
- Two complementary criteria capturing country- and region-specific dimensions:
  - Country-specific criterion: NPCIR must be significantly larger than the trend for that country (large relative to historical experience).
  - Regional criterion: NPCIR must exceed the 75th percentile of the regional NPCIR distribution, recognizing that regional thresholds capture large flows relative to peers.
- Trend extraction:
  - Rolling, backward-looking Hodrick-Prescott (HP) filter applied to each country’s NPCIR series to capture real-time nature of policy decisions.
  - Initial five years used to determine trend; subsequent years added on a rolling basis.
  - Smoothing parameter: 1000 for rolling HP trend (robust to other values). For series with insufficient observations (mostly Central and Eastern Europe), HP filter applied to entire series with smoothing parameter of 100.
- Episode identification:
  - Year t is an episode year if either:
    - Deviation of NPCIR from its trend in year t > one historical standard deviation and NPCIR exceeds 1 percent of GDP, or
    - NPCIR exceeds the 75th percentile of the regional distribution for region j over the whole sample.
  - An episode begins in the first year meeting these criteria and continues as long as subsequent years keep meeting them.
- Handling clustered episodes:
  - No episodes identified in the two years prior to each large inflow episode.
  - If an episode’s end-year immediately precedes first year of another episode → combine into single episode.
  - If one year separates two identified episodes, combine them including the middle year only if NPCIR in that middle year is positive; if NPCIR is negative, exclude the first episode (usually lower average NPCIR).
- Defining abrupt endings:
  - An episode ends "abruptly" if NPCIR in the year after the episode terminates is more than 5 percentage point of GDP lower than at the end of the episode (following definition of "sudden stops").
  - Also abrupt if the end coincides with a currency crisis (Frankel and Rose (1996) definition): depreciation of at least 25 percent cumulative over 12 months and at least 10 percentage points greater than in the preceding 12 months.

### V. Illustrative Case: Mexico (method mechanics)
- Rolling trend and regional NPCIR threshold for Mexico:
  - Example regional threshold: 4.8 percent of GDP.
- Episode identification in Mexico:
  - 1990 identified as episode year: inflows exceeded trend by more than one standard deviation.
  - 1994 identified as episode year: inflows were 4.9 percent of GDP, marginally exceeding regional threshold of 4.8 percent.
  - Years 1990–1994 comprise a single multi-year episode; 1997 and 2000 are single-year episodes.
  - 1995 NPCIR declines by only 3 percent relative to 1994, but collapse of the peso meets currency crisis definition and counts as an abrupt ending.

### VI. Stylized Facts on Capital Inflows Over Time
- Two large waves shown by NPCIR:
  - First wave: early 1990s, ending abruptly with 1997–98 Asian crisis.
  - Second wave: building since 2002, accelerated in 2007 with inflows far exceeding prior experience.

### VII. Structure of the Paper (organization)
- Section II: Database and methodology used to identify large capital inflows.
- Documentation of stylized facts associated with surges in capital inflows.
- Discussion of possible policy responses to cope with large capital inflows.
- Section V: Main features of policy responses to capital inflows.
- Section VI: Empirical links between macroeconomic outcomes and policy responses.
- Section VII: Conclusion.

*Source: _wp0940 - References..............................................................................................................*

### 2006. The second wave has started abated in 2008 as the flows of international capital have

### _wp0940 - 2006. The second wave has started abated in 2008 as the flows of international capital have

### Characteristics of the “second wave” of capital inflows
- Occurred in a context of generally stronger current account positions for most (but not all) emerging market countries and a substantial acceleration in the accumulation of foreign reserves.
- Accompanied by a sharp increase in outflows, reflecting the global trend toward increasing diversification of international portfolios.
- Predominance of net foreign direct investment (FDI) flows relative to net “financial” flows (portfolio and other flows) in all four regions studied.
- Differentiators from the 1990s wave:
  - Larger set of countries involved.
  - Underpinned by generally more solid current account positions (notable exception: emerging European countries).
  - Took place in a more financially integrated world economy where significant financial outflows partially offset inflows.

### Episodes of large capital inflows — empirical patterns
- Identified 109 episodes of large net private capital inflows since 1987 across 52 countries.
- Temporal and regional patterns:
  - Two waves since the late 1980s: one in the mid-1990s and the recent one starting in 2002.
  - Episodes completed between 1987 and 1998 generally involved a smaller volume of flows relative to GDP but lasted longer than episodes that ended between 1999 and 2006.
  - Emerging Asian and Latin American countries dominated the first wave; more recent episodes concentrated in emerging Europe and other emerging market countries.
- Abrupt endings:
  - Of the 87 completed episodes, 34 ended with a sudden stop and 13 with a currency crisis.
  - In 7 episodes, a sudden stop coincided with a currency crisis.
- Late and ongoing episodes are characterized by larger FDI flows relative to episodes completed in the 1990s.

### Policy responses to large capital inflows — overview
- Policymakers used a variety of measures to prevent overheating, real currency appreciation, and reduce vulnerability to reversals: exchange rate intervention, sterilization, fiscal policy, and capital controls.
- Key tradeoff: extent to which authorities resist pressures for currency appreciation by intervening in foreign exchange markets versus allowing nominal appreciation or tolerating inflation.
- The “impossible trinity” framework: targeting the exchange rate, running independent monetary policy, and allowing full capital mobility cannot be achieved simultaneously; however, imperfect capital mobility grants some scope for intermediate options.
- Common practical issues:
  - Accumulation of foreign reserves to prevent appreciation can lead to excessively loose monetary conditions, overheating, and financial vulnerabilities, with real appreciation occurring via higher inflation rather than nominal appreciation.
  - Sterilization can maintain incentives for capital inflows and entail quasi-fiscal costs (central bank exchanging high-yield domestic assets for low-yield reserves). If implemented via increased unremunerated reserve requirements, costs shift to banking system and promote disintermediation.
  - Fiscal policy in inflow-receiving emerging markets is often procyclical; greater restraint on expenditure growth has three benefits:
    - Dampens aggregate demand during high inflows, allowing lower interest rates and reducing incentives for inflows.
    - Alleviates appreciating pressures on the exchange rate because public spending is biased toward nontraded goods.
    - Helps address debt sustainability concerns and provides scope for countercyclical fiscal responses when inflows stop.
  - Fiscal rules based on cyclically adjusted balances can help resist pressures for additional spending (example policy objective described: Chile aims at a cyclically-adjusted fiscal surplus with an adjuster to save excess copper revenues).
- Capital controls objectives: discourage inflows to reduce appreciation pressures, reduce risk of sudden reversals, and maintain some monetary policy independence.

### Exchange rate policy: measurement and resistance index
- Objective: measure authorities’ tendency to resist exchange rate appreciations using an exchange market pressure (EMP) index and a derived resistance index.
- EMP index construction:
  - Components:
    - Year-over-year percentage change of the nominal bilateral exchange rate Δ%er_{i,t} (increase implies appreciation).
    - Year-over-year change in foreign reserves Δres_{i,t}, where change in net foreign assets (NFA) is scaled by lagged monetary base (MB).
  - Standard deviations computed using monthly data; regional averages denoted σ_{Δ%er_{i,t}} and σ_{Δres_{i,t}} are used to scale components.
  - EMP definition: EMP_{i,t} = Δ%er_{i,t}/σ_{Δ%er_{i,t}} + Δres_{i,t}/σ_{Δres_{i,t}}.
  - Rationale for regional (rather than country-specific) standard deviations: avoids spurious dominance of exchange-rate term when country-specific volatility is tiny.
  - Interest rate term excluded from EMP to avoid contaminating the index used to detect exchange market pressures; short-term interest rate fluctuations are used separately to gauge monetary policy stance.
- Resistance index:
  - Defined as Resistance Index_{i,t} = 1 − ( (Δ%er_{i,t}/σ_{Δ%er_{i,t}}) / EMP_{i,t} ).
  - Standardized to be bounded in the unit interval using discrete mapping:
    - If raw index ≤ 0, standardized = 0.
    - If raw index in (0, 0.25], standardized = 0.2.
    - If raw index in (0.25, 0.5], standardized = 0.4.
    - If raw index in (0.5, 0.75], standardized = 0.6.
    - If raw index in (0.75, 1), standardized = 0.8.
    - If raw index ≥ 1, standardized = 1.
  - Interpretation:
    - Index = 0: no resistance to exchange market pressures (float or “leaning with the wind”).
    - Index = 1: maximum resistance (no movement in exchange rate or extreme “leaning against the wind”).
    - Values between 0 and 1: proportion of EMP relieved by intervention; values closer to 1 imply greater resistance.

### Sterilization policy: index and interpretation
- Sterilization defined as central bank operations offsetting a rise in net foreign assets (NFA) by a decrease in net domestic assets (NDA), keeping monetary base constant.
- Sterilization index construction:
  - Compute monthly changes ΔNDA_{i,t} and ΔNFA_{i,t} for each country and year.
  - Run regression ΔNDA_{i,t} = α_{i,t} + β_{i,t} ΔNFA_{i,t} + u_{i,t} using 12 monthly observations per year; sterilization index is –β_{i,t}.
  - Interpretation:
    - Sterilization index = 1 implies full sterilization (central bank completely offsets increase in NFA by decreasing NDA).
    - Sterilization index = 0 implies no sterilization.
    - Sterilization index can exceed unity (over-sterilization) in some cases.
- Broader sterilization index (alternative):
  - Replace NDA with change in M2 in regression ΔM2_{i,t} = α + δ ΔNFA_{i,t} + u; here δ = 0 implies full monetary sterilization (NFA fluctuations not transmitted to M2), δ = 1 implies no sterilization.
  - Authors focus on NDA-based index because it better matches country experiences.
- Methodological considerations:
  - Alternatives in literature: simple ratios (ΔM2/ΔNFA, ΔNDA/ΔNFA), VARs addressing endogeneity, simultaneous-equation systems to allow contemporaneous endogeneity.
  - Chosen index balances technical sophistication and cross-country data consistency.
- Sterilization and interest rates:
  - Aggressive sterilization tends to raise domestic interest rates because decreasing NDA requires higher yields on short-term paper.
  - Movements in short-term interest rates are therefore considered as counterparts of central bank sterilization efforts and as an alternative gauge of monetary policy stance.
- Limitations:
  - Sterilization index averages behavior over entire episodes and may miss initial aggressive sterilization typically used at onset of inflow surges.
  - Demand for money instability in some countries complicates using sterilization index as a pure measure of monetary policy stance.

### Fiscal policy measures and indicators
- Fiscal stance represented by change in growth of real noninterest government expenditure.
  - Rationale: defines policy cyclicality in terms of actual instruments, avoids reliance on potentially misleading ratios to GDP, and data on tax rates is not systematic.
- Complementary measure: deviation of real government spending from trend using Hodrick-Prescott filter to provide a structurally adjusted characterization.
- Data source and transformation:
  - IMF’s World Economic Outlook (WEO) database used.
  - Interest expenditures subtracted from government expenditures (WEO codes: GGENL – GGEI).
  - Nominal noninterest expenditures scaled by CPI (WEO code PCPI) to yield real noninterest government expenditure series.
- Policy guidance:
  - Greater fiscal restraint during inflow surges can dampen aggregate demand, alleviate exchange rate appreciation pressures, and improve debt sustainability to enable countercyclical responses when inflows reverse.
  - Discretionary fiscal tightening can be politically and operationally difficult; holding the line on spending and fiscal rules based on cyclically adjusted balances are practical recommendations.

### Capital controls — conceptual and practical considerations
- Capital controls used temporarily during inflow surges for countries with fairly liberalized capital accounts; objectives include discouraging inflows, lowering risk of sudden reversals, and maintaining monetary policy autonomy.
- Theoretical perspectives:
  - Neoclassical model: FDI increases physical capital and supports growth; if supply of foreign capital is upward sloping, marginal cost of capital arguments can justify taxes or quantitative restrictions (Cardoso and Goldfajn, 1998).
  - Controls may be welfare-reducing unless they correct another market failure (second-best); cited market failures include myopic private speculation (Dooley, 1995) and volatility from increased financial market integration (Stiglitz, 2004).
  - Rodrik and Velasco (1999) and Rodrik and Subramanian (2009) make theoretical and empirical arguments supporting discouraging short-term inflows.
- Types of controls:
  - Direct (administrative): prohibitions and explicit quantitative limits (example: Malaysia’s 1998 measures to dampen large net capital outflows).
  - Indirect (market-based): taxes and other market mechanisms (example: Brazil’s ‘entrance tax’ facility for fixed income investments cited as an illustration of direct taxation of inflows).
- Focus of analysis: implications of temporary use of capital controls during inflow surges in countries with liberalized capital accounts.

*Source: IMF working paper content unit _wp0940 (PDF chapter/section).*

### 1994. Indirect capital controls include explicit or implicit taxation of financial flows and

### _wp0940 - 1994. Indirect capital controls include explicit or implicit taxation of financial flows and 

### Measurement of capital controls
- Capital controls are measured using an index based on the IMF’s Annual Report on Exchange Arrangements and Exchange Restrictions (AREAER).
- AREAER indices cover nine dimensions of controls (capital and money market instruments; credit operations; derivatives and other instruments; direct investment; personal capital movements; real estate transactions; provisions specific to commercial banks and other credit institutions; institutional investors; surrender requirements).
- The paper averages across these nine dimensions to form indices for controls on inflows and controls on outflows.
- The 1996 AREAER allows distinguishing controls on inflows from outflows beginning in 1995; older vintages are backcast in proportion to pre-1996 AREAER vintages.
- Limitations noted:
  - Enforcement intensity is not directly captured.
  - Measures are “inherently coarse” and may fail to detect some policy changes.

### V. Policy responses — basic stylized facts
- Second wave of capital inflows: associated with strong exchange market pressures resisted through:
  - Accumulation of foreign reserves.
  - Allowing some upward movements in exchange rates.
- Contrast with earlier wave (1990s): pressures on exchange rates were negative for most emerging markets, reflecting large current account deficits and typical depreciation.
- Evidence on intervention asymmetry:
  - Foreign exchange reserves increased over the 1990s, suggesting tendency to intervene to prevent appreciation but not depreciation (except during extreme crises: large reserve reductions in 1997 in emerging Asia and in 2001 in Latin America and other emerging markets).
- Resistance index: remained relatively high across emerging market regions in recent past (desire to limit exchange rate appreciation).
- Sterilization index: high in early 1990s and early 2000s (beginnings of two waves of large inflows), tapering off around 2006.
- Real government expenditure: accelerated starting in 2005 in emerging markets, especially in Latin America and emerging Europe and the CIS.
- Capital controls indices (Figure 9):
  - Controls on inflows relaxed since late 1990s, but changes slow in aggregate.
  - Emerging Europe and CIS moved furthest toward liberalization; emerging Asia remained relatively restrictive.
  - Controls on outflows progressively loosened in emerging Europe and the CIS and other emerging markets; more recently in emerging Asia and Latin America.

### A. Policy responses during episodes of large capital inflows
- Episodes defined; analysis uses medians across episode averages and nonparametric median tests (Wilcoxon / Mann-Whitney rank-based median test).
- General trends for episodes (see Figure 10):
  - Resistance index tends to increase during an episode; statistically significant increase for episodes completed before 1998.
  - Sterilization does not tend to increase during an episode relative to the two years beforehand (temporary sterilization efforts often unsustainable due to quasi-fiscal costs).
  - Real government expenditures tend to increase strongly as capital inflows surge (fiscal policy generally procyclical).
  - Controls on inward capital flows: tightened during episodes completed before 1998 (not always significantly); during more recent episodes controls appear to have been eased.
  - For completed episodes, surge in inflows has not coincided with relaxation of controls on capital outflows; restrictions appear less strict during ongoing episodes.

### VI. Linking macroeconomic outcomes and policy responses — stylized facts
- Four major stylized facts (Figure 11):
  1. Large inflows associated with acceleration of GDP growth; afterwards growth often drops significantly.
     - For episodes that end abruptly: average GDP growth in the two years after the end of the episodes tends to be about 3 percentage points lower than during the episode, and about 1 percentage point lower than during the two years before the episode.
  2. GDP growth fluctuations accompanied by large swings in aggregate demand and current account; strong deterioration of current account during inflow period and sharp reversal at the end.
  3. End of inflow episodes typically entails sharp reversal of non-FDI flows; FDI more resilient.
  4. Surge associated with real effective exchange rate appreciation; median differences not always statistically significant; appreciation has not on average been generated by higher inflation (many episodes occurred during inflation stabilization plans).

### B. Avoiding a hard landing
- Hard landing correlates:
  - Faster acceleration in domestic demand, sharper rise in inflation, larger real appreciation during inflow period.
  - Longer episodes with much higher cumulative size of inflows (examples: Thailand 1988–96; Argentina 1992–94 and 1997–99; Mexico 1990–94).
- Fiscal policy:
  - “Hard landings” associated with strong increases in government spending during inflow period.
  - Expenditure restraint helps reduce upward pressures on aggregate demand and real exchange rate and facilitates soft landing.
- Resistance to exchange market pressures and sterilization:
  - Higher resistance and greater sterilization were unable to prevent real appreciation and generally unsuccessful in achieving soft landing.
  - Sterilization can induce higher nominal interest rates and perpetuate inflows.

### Regression evidence — post-inflow GDP growth (Table 3) (robust p values in parentheses; * significant at 10 percent; ** significant at 5 percent)
- Dependent variable: Post-inflow GDP Growth.
- Selected coefficients (columns reported 1–5):
  - Real Government Expenditure: -0.109 (0.015)**; -0.111 (0.014)**; -0.111 (0.014)**; -0.099 (0.027)**; -0.093 (0.040)**.
  - Index of Resistance to Exchange Market Pressures: -1.812 (0.114); -2.090 (0.085)*; -2.086 (0.088)*; -2.147 (0.080)*; -2.282 (0.059)*.
  - Post-inflow World Output Growth: 1.023 (0.017)**; 0.836 (0.056)*; 0.858 (0.071)*; 0.875 (0.063)*; 0.844 (0.076)*.
  - Real US Fed Funds Rate: 0.279 (0.165); 0.279 (0.170); 0.209 (0.294); 0.240 (0.226).
  - Cumulative Size of Capital Inflow: -0.049 (0.148); -0.048 (0.157) (in later specifications).
  - Sterilization Index: -0.981 (0.262) (in one specification).
- Observations: 69 (in each reported column).
- Adjusted R-squared ranges: 0.133, 0.138, 0.125, 0.187, 0.188 across specifications.

### Regression evidence — real exchange rate appreciation (Table 4) (robust p values in parentheses; * significant at 10 percent; ** significant at 5 percent; *** significant at 1 percent)
- Dependent variable: Real Effective Exchange Rate Appreciation.
- Selected coefficients (columns 1–4):
  - Real Government Expenditure: 0.544 (0.003)***; 0.396 (0.029)**; 0.321 (0.071)*; 0.307 (0.112).
  - Index of Resistance to Exchange Market Pressures: -0.239 (0.953); -0.256 (0.949); -0.107 (0.979).
  - Output gap: 0.954 (0.050)**; 0.715 (0.094)*; 0.654 (0.130).
  - Real U.S. Fed Funds Rate: 0.492 (0.604); 1.606 (0.100)*; 1.755 (0.078)*.
  - Cumulative Size of Capital Inflow: 0.241 (0.083)*; 0.249 (0.074)*.
  - Sterilization Index: 2.562 (0.289) (included in one specification).
- Observations: 107 (columns 1–3), 106 (column 4).
- Adjusted R-squared ranges: 0.115, 0.138, 0.227, 0.222.

### D. Any role for capital controls?
- Episodes with tighter controls on inflows are associated with:
  - Narrower current account deficits and lower net private inflows, including lower net FDI (Figure 14, top panel).
  - Stricter inflow controls accompanied by lower post-inflow growth and larger appreciation, but these distinctions are not statistically significant; inflation rates significantly higher in episodes with tighter controls.
- Episodes ending in abrupt reversals do not seem associated with lower capital controls (Figure 14, bottom panel).
- Using capital controls measures in regressions:
  - More restrictive controls on inflows may help avoid hard landings in some specifications.
  - None of the capital control measures is statistically significant in the real exchange rate equations.
  - Using Edwards (2005) measure: only statistically significant for more parsimonious post-inflow growth regressions (first three columns of Table 3).
- Literature and microeconomic costs:
  - Controls on inflows often did not affect volume of net flows; could temporarily tilt composition toward longer maturities in a few cases (e.g., Chile after 1991).
  - Stricter controls on outflows helped in Malaysia after 1998 but little support elsewhere.
  - Sustained capital controls associated with microeconomic costs: harder and more expensive for small firms to raise capital; higher local borrowing costs for multinationals; distortions and resources spent to circumvent controls; increased trade costs due to inspection/reporting.
- Overall assessment: mixed evidence; capital controls not a substitute for sound macroeconomic policies (prudent fiscal stance; supporting exchange rate and monetary framework; appropriate prudential measures).

### E. Persistence of inflows, initial imbalances, and regional factors
- Duration matters:
  - Episodes lasting less than two years display larger resistance to exchange rate changes, less real appreciation, and better post-inflow GDP growth (Figure 15, upper panel).
  - However, higher resistance in short episodes was not associated with significantly smaller real appreciation or better post-inflow growth (Figure 15, lower panel).
  - Resisting exchange market pressures may be more feasible for transitory inflows but does not guarantee better outcomes and ex ante identification of transitory episodes is difficult.
- Balance-of-payment pressures:
  - Fiscal restraint less decisive in episodes with high BoP pressures (above-median sum of current account and net private capital inflows): lower government spending growth not associated with significantly better outcomes (Figure 16, upper panel).
  - Fiscal spending restraint associated with significantly better outcomes when episodes have low BoP pressures (Figure 16, lower panel).
- Regional patterns:
  - Regions with stronger real appreciation during episodes (Latin America; emerging Europe and the CIS) also had larger increases in public expenditure.
  - Advanced economies followed more countercyclical fiscal policies and refrained from resisting exchange market pressures; they experienced less real appreciation and smaller GDP growth fluctuations.

### VII. Conclusions — policy implications and guidance
- Four general patterns from the review:
  1. Countries with relatively high current account deficits more vulnerable to sharp reversals of capital inflows.
  2. Public expenditure restraint during inflow episodes contributes to lower real exchange rate appreciation and better post-inflow GDP growth.
  3. Resistance to nominal exchange rate appreciation has not generally prevented real appreciation and has often been followed by sharper reversals, especially for persistent inflows.
  4. Restrictions on capital inflows have generally not been associated with lower real appreciation or softer landings.
- Policy guidance:
  - Stabilization challenges most serious for countries with substantial current account imbalances.
  - In contexts of relatively inflexible exchange rate policies, fiscal restraint is likely the most effective tool to avoid overheating and output instability.
  - If central banks initially resist nominal appreciation, that stance should be progressively relaxed if inflows persist to reduce the likelihood of eventual real appreciation and costly reversals.
  - Policy responses should consider country-specific circumstances (nature and persistence of inflows; stage of business cycle; fiscal situation; quality of domestic financial markets).
  - Other tools not systematically analyzed here but relevant: financial supervision and regulation, labor and product market reforms.

*Source: Authors’ calculations.*

### References

### _wp0940 - References

### Capital Controls and Capital Flows
- Ariyoshi, A., K. Habermeier, B. Laurens, I. Otker-Robe, J. Canales-Kriljenko, and A. Kirilenko, 2000, Capital Controls: Country Experiences with Their Use and Liberalization, IMF Occasional Paper No. 190 (Washington: International Monetary Fund).
- Calvo, G. A., L. Leiderman, and C. M. Reinhart, 1993 “Capital Inflows and Real Exchange Rate Appreciation in Latin America: The Role of External Factors,” IMF Staff Papers, Vol. 40 No. 1, 108–51.
- Calvo, G. A., L. Leiderman, and C. M. Reinhart, 1994, “The Capital Inflows Problem: Concepts and Issues,” Contemporary Economic Policy, Vol. XII No. 3 (July), 54–66.
- Calvo, G. A., L. Leiderman, and C. M. Reinhart, 1996, “Inflows of Capital to Developing Countries in the 1990s,” Journal of Economic Perspectives, Vol. 10 (Spring), 123–39.
- Miniane, J. 2004, “A New Set of Measures on Capital Account Restrictions,” IMF Staff Papers, Vol. 51, No. 2, pp. 276–308 (Washington: International Monetary Fund).
- Cowan, K, and J. De Gregorio, 2005, “International Borrowing, Capital Controls, and the Exchange Rate Lessons from Chile,” NBER Working Paper No. 11382 (Cambridge, Massachusetts: National Bureau of Economic Research).
- Desai, M., C. F. Foley, and J. R. Hines Jr., 2004, “Capital Controls, Liberalizations and Foreign Direct Investment,” NBER Working Paper No. 10337 (Cambridge, Massachusetts: National Bureau of Economic Research).
- Forbes, K., 2005, “The Microeconomic Evidence on Capital Controls: No Free Lunch,” NBER Working Paper No. 11372 (Cambridge, Massachusetts: National Bureau of Economic Research).
- Magud, N. and C. Reinhart, 2007, “Some Lessons for Policy Makers Who Deal With the Mixed Blessing of Capital Inflows,” in Capital Controls and Capital Flows in Emerging Economies, ed. by Sebastian Edwards, 645–74 (Chicago, Illinois: The University of Chicago Press).
- Wei, S.-J., and Z. Zhang, 2006, “Collateral Damage: Exchange Controls and International Trade,” IMF Working Paper 07/08 (Washington: International Monetary Fund).

### Exchange Rate Regimes, Exchange Market Pressure, and Fear of Floating
- Baig, T., 2001, “Characterizing Exchange Rate Regimes in Post-Crisis East Asia,” IMF Working Paper 01/152 (Washington: International Monetary Fund).
- Calvo, G. A., and C. M. Reinhart, 2002, “Fear of Floating,” Quarterly Journal of Economics, Vol. 117, No. 2, pp. 379–408.
- Bertoli, S., G. M. Gallo, and G. Ricchiuti, 2006, “Exchange Market Pressure: Some Caveats in Empirical Applications,” Working Paper 2006/17 (Florence, Italy: Dipartimento di Statistica “Giuseppe Parenti”).
- Girton, L., and D. Roper, 1977, “A Monetary Model of Exchange Market Pressure Applied to the Postwar Canadian Experience,” American Economic Review, Vol. 67, No. 4, pp. 537–48.
- Pentecost, E. J., C. Van Hooydonk, and A. Van Poeck, 2001, “Measuring and Estimating Exchange Market Pressure in the EU,” Journal of International Money and Finance, Vol.20, pp.401–18.
- Van Poeck, A., J. Vanneste, and M. Veiner, 2007, “Exchange Rate Regimes and Exchange Market Pressure in the New EU Member States,” Journal of Common Market Studies, Vol. 45, No.l 2, pp. 459–85.
- Levy Yeyati, E., and F. Sturzenegger, 2005, “Classifying Exchange Rate Regimes: Deeds vs. Words,” European Economic Review, Vol. 49, No. 6 (August), pp. 1603–35.

### Monetary Policy Responses, Sterilization, and Capital Inflow Management
- Cavoli, T., and R. S. Rajan, 2006, “Capital Inflows Problem in Selected Asian Economies in the 1990s Revisited: The Role of Monetary Sterilization,” Asian Economic Journal Vol. 20, No. 4, pp. 409–23.
- Christensen, J., 2004, “Capital Inflows, Sterilization, and Commercial Bank Speculation: The Case of the Czech Republic in the Mid-1990s,” IMF Working Paper 04/218 (Washington: International Monetary Fund).
- Moreno, R., 1996, “Intervention, Sterilization, and Monetary Control in Korea and Taiwan,” FRBSF Economic Review, Federal Reserve Board of San Franciso, Vol. 3, pp. 23–33.
- Ouyang, A. Y., R. S. Rajan, and T. D. Willet, 2007, “China as a Reserve Sink: The Evidence from Offset and Sterilization Coefficients,” HKIMR Working Paper No. 10/2007 (Hong Kong SAR: Hong Kong Institute for Monetary Research).
- Takagi, S. and T. Esaka, 1999, “Sterilization and the Capital Inflow Problem in East Asia,” Discussion Paper No. 86 (Tokyo: Economic Planning Agency).

### Capital Inflows, Crises, Contagion, and Adjustment
- Calvo, G. A., and C. A. Vegh, 1999, “Inflation Stabilization and Balance of Payment Crises in Developing Countries,” in Handbook of Macroeconomics, Ed. 1, Vol. 1, pp. 1531-614, ed. by J. B. Taylor and M. Woodford (Elsevier).
- Eichengreen, B., A. K. Rose, and Charles Wyplosz, 1996, “Contagious Currency Crises,” NBER Working Paper No. 5681 (Cambridge, Massachusetts: National Bureau of Economic Research).
- Frankel, J., and A. Rose, 1996, “Currency Crashes in Emerging Markets: An Empirical Treatment,” Journal of International Economics, Vol. 41, No. 3/4, pp. 351–66.
- Gourinchas, P., R. Valdes, and O. Landerretche, 2001, “Lending Booms: Latin America and the World,” Economia, Vol. 1, No. 2 (Spring).
- Milesi-Ferreti, G. M., and A. Razin, 1998, “Sharp Reductions in Current Account Deficits: An Empirical Analysis,” European Economic Review, Vol. 42 (April), pp. 897–908.
- Edwards, S., 2005, “Capital Controls, Sudden Stops, and Current Account Reversals,” NBER Working Paper No. 11170 (Cambridge, Massachusetts: National Bureau of Economic Research).
- Reinhart, C. M., and V. Reinhart, 1998, “Some Lessons for Policymakers Who Deal With the Mixed Blessing of Capital Inflows,” in Capital Flows and Financial Crises, pp. 93-127, ed. by Miles Kahler (Ithaca, New York: Cornell University Press).
- Reinhart, C. M., 2008, “Capital Flow Bonanzas: An Encompassing View of the Past and Present,” NBER Working Paper No. 14321 (Cambridge, Massachusetts: National Bureau of Economic Research).

### Fiscal Policy, Growth, and Financial Globalization
- Braun, M., 2001, “Why is Fiscal Policy Procyclical in Developing Countries?” Mimeo. (Cambridge, Massachusetts: Harvard University).
- Gavin, M., and R. Perotti, 1997, “Fiscal Policy in Latin America,” NBER Macroeconomics Annual, pp. 11–61 (Cambridge, Massachusetts: MIT Press).
- Lane, P., 2003, “The Cyclical Behavior of Fiscal Policy: Evidence from the OECD,” Journal of Public Economics, Vol. 86, pp. 2661–75.
- Kose, M. A., E. S. Prasad, K. Rogoff, S. J. Wie, 2009, “Financial Globalization: A Reappraisal,” forthcoming in IMF Staff Papers.
- Rodrik, D., and A. Subramanian, 2008, “Why Did Financial Globalization Disappoint?” Manuscript, Harvard University. Forthcoming in IMF Staff Papers.
- Obstfeld, M., 2009, “International Finance and Growth in Developing Countries: What Have We Learned?” Forthcoming in IMF Staff Papers.

### Empirical Methods, Datasets, and Statistical Tools
- Conover, W.J., 1980, Practical Nonparametric Statistics, Second Edition (New York: John Wiley and Sons).
- Mann, H.B., and D.R. Whitney, 1947, “On a Test of Whether One of Two Random Variables is Stochastically Larger Than the Other,” Annals of Mathematical Statistics, Vol. 18, pp. 50–60.
- Wilcoxon, F., 1945, “Individual Comparisons by Ranking Methods,” Biometrics Bulletin, Vol. 1, pp. 80–83.
- Hodrick, R., and E. C. Prescott, 1997, “Postwar U.S. Business Cycles: An Empirical Investigation,” Journal of Money, Credit, and Banking, Vol. 29, No. 1 (February), pp. 1–16.
- Schindler, M., 2009, “Measuring Financial Integration: A New Dataset,” forthcoming in IMF Staff Papers.
- Miniane, J. 2004, “A New Set of Measures on Capital Account Restrictions,” IMF Staff Papers, Vol. 51, No. 2, pp. 276–308 (Washington: International Monetary Fund).

*References list as provided in the source PDF*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2009/_wp0940.pdf_