2015年-IMF国际货币组织全球_Soft_Power_and_Exchange_Rate_Volatility_35页_515kb

Summary of "Soft Power and Exchange Rate Volatility"

Core Content

This working paper by Serhan Cevik, Richard Harris, and Fatih Yilmaz explores the role of "soft power" factors in determining exchange rate volatility across countries. The paper challenges the traditional view that exchange rate volatility is solely driven by macro-financial variables, introducing a new set of variables that reflect a country's demographic, institutional, political, and social characteristics.

The study utilizes a balanced panel dataset of 115 countries from 1996 to 2011, analyzing the relationship between exchange rate volatility and "soft power" variables. The authors employ multiple econometric approaches, including pooled OLS, IV, and GMM, to ensure robustness of their findings. They also conduct variable reduction techniques to address collinearity among the explanatory variables.

Main Points

• Exchange Rate Volatility: The paper finds that exchange rate volatility is highly persistent over time, particularly in emerging market economies. It also notes that the global financial crisis in 2008 increased volatility, but it later reverted to pre-crisis levels by 2012.

• Soft Power Variables: These include governance indicators, demographic data, education levels, financial sector metrics, and development indicators. The authors identify a set of nine key "soft power" variables after applying a variable reduction technique based on the Variance Inflation Factor (VIF):

  • Political stability (POLS)
  • Voice and accountability (VCEA)
  • Life expectancy (LIFE)
  • Average years of primary schooling (PRIS)
  • Average years of tertiary schooling (TERS)
  • Bank z-score (ZSCR)
  • Bank concentration (BCNC)
  • Financial openness (FOPN)
  • Share of agriculture relative to services (AGMS)

• Control Variables: The paper includes 10 conventional macro-financial control variables:

  • Consumer price inflation (CPII)
  • Terms of trade volatility (TOTV)
  • Labor productivity growth volatility (PRDV)
  • Government consumption to GDP ratio volatility (GOVV)
  • Current account balance to GDP ratio (CBAL)
  • Trade openness (OPEN)
  • Export concentration (CONC)
  • Exchange rate regime (REGM)
  • Credit to GDP ratio (CRED)
  • Stock market capitalization (SCAP)

• Empirical Findings:

  • Voice and Accountability (VCEA) and Life Expectancy (LIFE) have a dampening effect on exchange rate volatility.
  • Financial Openness (FOPN) and Bank Z-Score (ZSCR) also reduce exchange rate volatility.
  • Primary schooling (PRIS) increases volatility, while Tertiary schooling (TERS) has a dampening effect, especially in developing countries.
  • The results are robust across different estimation methods and sub-samples, indicating that "soft power" variables significantly influence exchange rate volatility.

Key Variables and Methodology

Dependent Variable

• VOL: Annual volatility of the real effective exchange rate (REER), calculated as the natural logarithm of the realized variance of monthly REER returns.

Estimation Techniques

• Pooled Model (OLS and IV): Used to estimate the relationship between exchange rate volatility and control variables.
• Fixed Effects Model (GMM): Applied to account for unobserved heterogeneity and endogeneity, using first differences and levels of variables as instruments.

Variable Reduction

• The authors use the Variance Inflation Factor (VIF) to reduce collinearity, retaining variables with VIF below 5.0.
• Alternative VIF thresholds (2.5) were tested, but the results remained largely consistent.

Conclusion

The paper concludes that "soft power" factors play a crucial role in explaining exchange rate volatility, especially in emerging market economies. These factors, which include governance, demographics, education, financial sector health, and development indicators, influence the behavior of exchange rates through their impact on macroeconomic fundamentals and risk premia. The findings suggest that traditional models based on macro-financial variables are incomplete and that incorporating "soft power" variables can help improve the understanding and prediction of exchange rate volatility.

Key Information

• Time Period: 1996–2011
• Sample Size: 115 countries
• Methodology: Pooled OLS, IV, and GMM
• Variables: 20 "soft power" variables and 10 control variables
• Main Findings:
  • "Soft power" variables significantly affect exchange rate volatility.
  • Volatility is more persistent in emerging markets.
  • Education levels, particularly tertiary schooling, have a dampening effect in developing countries.
  • Financial openness and institutional quality reduce exchange rate volatility.

Figures and Tables

• Figure 1: Shows the trend in REER volatility from 1995 to 2012, highlighting a decline in the 1990s, a spike in 2008, and a return to pre-crisis levels by 2012.
• Tables 1–5: Present the results of various econometric models, including pooled OLS, IV, and GMM, with both exogenous and endogenous control variables. The results are generally robust across different specifications.

References

• The paper references a wide range of studies, including Meese and Rogoff (1983), Acemoglu et al. (2001), and others, to support the theoretical and empirical background of the analysis.