2011年-IMF国际货币组织全球_Risk_Sharing_and_Financial_Contagion_in_Asia_An_Asset_Price_Perspective_42页_1003kb

Summary of "Risk Sharing and Financial Contagion in Asia: An Asset Price Perspective"

Core Content

This paper explores the dual effects of financial integration in Asia, focusing on risk sharing and financial contagion from an asset price perspective. It evaluates whether the benefits of financial integration in terms of risk sharing outweigh the costs associated with financial contagion, using a novel method based on stochastic discount factors and quantile regressions.

The key findings suggest that while risk sharing is low in Asia, contagion risks are more significant within the region. The paper argues that Asia has the potential to enhance risk sharing without increasing contagion risk, and emphasizes the importance of policy coordination and financial market development in achieving this.

Main Points

• Financial Integration Benefits:

  • Financial integration can help transfer resources, diversify risks, and promote financial development.
  • Risk sharing is a core benefit, but it is not always perfect, especially in Asia.

• Financial Contagion Risks:

  • Financial contagion refers to the spread of financial shocks across markets.
  • It is more significant intra-regionally in Asia, suggesting that financial linkages can amplify regional vulnerabilities.

• Tradeoff Between Risk Sharing and Contagion:

  • There is an overall tradeoff between the benefits of risk sharing and the costs of financial contagion.
  • The terms of the tradeoff vary across countries, depending on economic fluctuations and inflation differentials.
  • Some countries have more successful risk sharing mechanisms than others due to contextual factors.

Key Information

Risk Sharing Measurement

• Perfect Risk Sharing: Occurs when the stochastic discount factors (SDFs) of two countries are equal, i.e., $ M_{t+1}^i = M_{t+1}^j $.
• BCS Index: A metric to quantify the degree of risk sharing between country pairs:
  $$
  BCS_{i,j} = 1 - \frac{\text{var}(\log M_{t+1}^i - \log M_{t+1}^j)}{\text{var}(\log M_{t+1}^i) + \text{var}(\log M_{t+1}^j)}
  $$
  • Ranges from -1 to 1, with higher values indicating better risk sharing.
  • The BCS index is derived from bond yield curves and an affine term structure model, which allows for the estimation of SDFs directly.

Financial Contagion Measurement

• CoVaR Approach: Used to measure financial contagion by estimating the conditional Value-at-Risk of one market given the stress in another.
• Tail Correlation: Focuses on negative tail spillovers rather than simple correlation, which better captures the risk of contagion.
• Empirical Evidence: Financial contagion is more pronounced within Asia, and the degree of risk sharing is lower than the US or EU.

Estimation Methodology

• Affine Term Structure Model: Used to estimate the stochastic discount factors from government bond yields.
• Steps for Estimation:
  1. Interpolate yield curves using piecewise cubic Hermite polynomials.
  2. Estimate VAR models for the first three principal components of the yield curve.
  3. Apply the Adrian-Moench procedure to estimate the price of risk ($\Lambda_0$ and $\Lambda_1$).
  4. Refine estimates using maximum likelihood with initial conditions from the Adrian-Moench method.
• Results:
  • The refined estimates show a significant improvement in model fit compared to the initial estimates.
  • The volatility of SDFs is higher than that of exchange rates, consistent with prior findings.

Policy Implications

• Enhance Risk Sharing: Asia can improve risk sharing without increasing contagion risk.
• Policy Recommendations:
  • Promote financial market development.
  • Improve macroeconomic coordination.
  • Focus on contextual factors that influence the success of financial integration.

Key Figures and Tables

Figures

• Figure 1: Log Stochastic Discount Factors for selected countries.
  • (a) US and EU: Show high correlation and similar levels.
  • (b) NIEs: Lower correlation and higher volatility.
  • (c) ASEAN3: Similar to NIEs, with less risk sharing.

Tables

• Table 1: Affine Term Structure Model Fit

  • Compares RMSE (Root Mean Squared Error) for different countries.
  • Shows that refined estimates significantly improve model fit.

• Table 2: Summary Statistics of $ m_{t+1} $

  • Presents mean and standard deviation of log stochastic discount factors.
  • Highlights volatility and negative mean of SDFs across countries.

• Table 3a: BCS Indices for all pairs of 13 countries

  • Average BCS index is 0.10, indicating low risk sharing.
  • High BCS indices (e.g., US-Hong Kong SAR at 0.587) suggest better risk sharing.
  • Low BCS indices for most Asian country pairs indicate limited risk sharing.

Conclusion

• The paper concludes that Asia has not fully realized the benefits of financial integration.
• Financial integration can still be beneficial if accompanied by sound policy frameworks and effective risk management.
• Policy coordination and market development are essential to enhance risk sharing while mitigating contagion risks.