2023-07-12-IMF-外部冲击_政策和实际汇率的尾部变动_51页_797kb

报告摘要

Summary of "External Shocks, Policies, and Tail-Shifts in Real Exchange Rates"

Core Content

This IMF Working Paper by Nicolas E. Magud and Samuel Pienknagura examines the impact of external shocks on real exchange rate (REER) movements in small open economies (SOEs), with a particular focus on extreme (tail) movements. The study uses panel quantile regressions to analyze how shocks to global uncertainty (measured by the VIX) and global financial conditions (U.S. monetary policy) affect the distribution of REER changes, especially in the tails. It also investigates the effectiveness of foreign exchange intervention (FXI) and capital flow management (CFM) in mitigating these effects.

Main Findings

1. Impact of External Shocks on REER Distributions

Global uncertainty (VIX) and U.S. monetary policy shocks significantly affect the distribution of REER changes.
The impact is stronger in the tails, especially in the left tail (large depreciations).
VIX shocks lead to left shifts in the REER distribution, with the 10th percentile experiencing the largest depreciation (about 1.1% for a one standard deviation shock).
U.S. monetary policy shocks also cause left shifts, but with stronger effects on the median and 90th percentile (stronger currencies), reducing the probability of large appreciations.
Negative shocks (e.g., rising VIX, tightening U.S. monetary policy) are more impactful than positive shocks.

2. Effectiveness of FXI and CFM

FXI is more effective in mitigating the effects of external shocks in the left tail, especially in economies with shallower FX markets, lower central bank credibility, and higher credit risk (i.e., weaker macro fundamentals).
CFM is found to be ineffective in reducing the impact of external shocks on REER movements, regardless of whether they are used to restrict inflows or outflows.
FXI is more effective in capital outflow events than in inflow events, suggesting that policymakers are more concerned about excessive real depreciation than appreciation.
The effectiveness of FXI varies along the distribution of REER changes, highlighting the importance of using quantile regressions rather than standard averages.

3. Policy Implications

FXI is more effective for non-heavy users of the policy, suggesting that overreliance on FXI may reduce the incentive for building deep FX markets and central bank credibility.
Capital flow management may offer limited benefits in the immediate aftermath of shocks but is not a robust tool for long-term stability.
Initial conditions matter: net debtor countries are more affected by external shocks than net creditor countries, emphasizing the need for strong macro policy frameworks.

Key Information

4. Nonlinear Effects of Shocks

The nonlinear impact of shocks is more pronounced in the tails of the REER distribution.
The effectiveness of FXI is higher in weaker economies, due to base effects and market frictions.
Asymmetries exist in how positive and negative shocks affect different deciles of the REER distribution.

5. Magnitude and Cost of FXI

Reducing the 10th percentile depreciation by 10% after a VIX shock requires an FXI of about 2.8% of GDP.
This is equivalent to the decline in reserves observed during the Global Financial Crisis (GFC).
For each 1% of GDP of FXI used, the real exchange rate depreciation is mitigated by 3.5 percentage points, which is more than double the average impact found in other studies.

6. Policy Complementarities

The paper contributes to a growing literature on policy complementarities in response to different types of shocks.
FXI and CFM can be used in combination to mitigate welfare costs from external shocks.
Macroprudential policies and domestic monetary policy are also considered in the analysis.

Methodology

7. Quantile Regression Approach

The study uses panel quantile regressions to analyze the distributional impact of shocks.
It focuses on three key quantiles: the 10th, median, and 90th percentiles.
This method allows for a more nuanced understanding of how shocks affect different parts of the REER distribution, especially extreme movements.

8. Data Sources

REER data is from the IMF Information System Notice (INS).
Trade openness, international reserves, CPI inflation, terms of trade, and real GDP are from IMF’s IFS and WEO.
VIX data is from the Chicago Board of Exchanges.
Banking crises dummies are from Laeven and Valencia (2020).
U.S. monetary policy shocks are from Ilzetzki and Jin (2021), following the Gertler and Karadi (2015) approach.
FXI and CFM data are from Adler and others (2021) and Magud and others (2018), respectively.
Exchange rate shallowness is measured by the ask-bid spread from Bloomberg and Refinitiv.
Inflation anchoring is measured using Bems and others (2021).
Sovereign spreads are from Bloomberg.

Conclusion

The paper highlights the importance of tail shifts in the REER distribution when analyzing the impact of external shocks and policy responses. It argues that FXI is more effective in mitigating tail risks, especially in economies with weaker macro fundamentals, and that quantile regression is a more appropriate method for capturing these nonlinear effects. The study also emphasizes the need for deep FX markets and central bank credibility to reduce the moral hazard of relying too heavily on FXI. Finally, it suggests that capital flow management is ineffective in the long run, and that policy complementarities and initial conditions play a critical role in the effectiveness of interventions.