2011年-IMF国际货币组织全球_A_Quantitative_Assessment_of_Financial_Conditions_in_Asia_22页_1mb

报告摘要

Summary of "A Quantitative Assessment of Financial Conditions in Asia"

Core Content

This working paper introduces a new Financial Condition Index (FCI) for 13 Asian economies, developed using two methodologies: the VAR model and the Dynamic Factor Model (GDFM). The FCI aims to capture the exogenous changes in financial conditions that influence future economic activity, rather than changes that reflect the business cycle.

The FCI is designed to serve as a leading indicator of GDP growth. The paper evaluates its predictive power through both in-sample and out-of-sample tests, showing that it improves the accuracy of GDP forecasts.

Main Methodologies

A. Weighted-Sum Approach

The FCI is calculated as a weighted average of financial variables (e.g., equity prices, exchange rates, credit growth, and interest rate spreads).
Weights are derived from generalized impulse responses in a VAR model, which estimate the impact of one-unit shocks to financial variables on GDP growth over 4–6 quarters.
The index is constructed as:

$$
F C I _ {t} = \sum_ {j = 1} ^ {n} w _ {j} \left(x _ {j, t} - \tilde {x} _ {j}\right)
$$

Where $w_j$ is the weight, and $\tilde{x}_j$ is the average of the financial variable over the sample period.

B. Principal Component Approach

The FCI is constructed using a Generalized Dynamic Factor Model (GDFM).
A common factor is estimated from a group of financial variables and interpreted as an unobserved common variable that drives the variation in financial conditions.
The common factor is then regressed on economic activity variables (GDP growth and inflation) to remove endogenous components.
The resulting residual is the FCI, which captures only exogenous changes in financial conditions.

C. Combined Index

The combined FCI is the simple average of the VAR-based and GDFM-based indices.
This approach aims to combine the strengths of both methodologies.
The combined index generally outperforms the individual indices in both in-sample and out-of-sample predictive tests.

Key Findings

The FCI has predictive power for GDP growth and can be used as a leading indicator.
Financial conditions in Asia tightened significantly in late 2008, reflecting stock market losses and tighter credit conditions.
By early 2010, financial conditions recovered rapidly, reaching pre-crisis levels, supported by accommodative monetary policies and a strong rebound in equity markets.
The contemporaneous correlation between FCI and GDP growth is 0.6 for emerging Asia and 0.9 for advanced Asia.
Partial $R^2$ values for the FCI in in-sample tests are above 0.7 for most economies, indicating that the index can explain a large portion of GDP growth variation.
Out-of-sample RMSE values show that the FCI improves forecast accuracy, especially when combined with other variables.
The real-time correlation between GDP outturns and forecasts increases from 0.68 to 0.87 when the FCI is included.

Financial Variables and Their Contributions

The relative contributions of financial variables differ across economies.
In China and the Philippines, credit growth plays a larger role in the FCI, reflecting the importance of banking intermediation.
In export-dependent economies like Hong Kong SAR and Taiwan Province of China, exchange rates have a greater contribution to the FCI.
Stock prices and exchange rates are associated with greater GDP volatility, while interest rates and credit are linked to less volatile GDP growth.

Policy Implications

The FCI provides useful insights for policymakers on the drivers of financial conditions.
It helps identify which financial variables are excessively accommodative or tight, guiding the appropriate policy response.
The index is particularly useful in real-time forecasting, where it enhances the accuracy of GDP forecasts.

Structure of the Paper

Introduction: Highlights the importance of financial conditions in the real economy and introduces the FCI.
Methodology: Describes the development of the FCI using the weighted-sum and principal-component approaches.
Evaluating the Financial Condition Index: Presents statistical tests to assess the FCI’s predictive power.
Developments of Financial Conditions in Asia: Analyzes the evolution of financial conditions over the global crisis, recovery, and normalization phases.
Conclusions: Summarizes the effectiveness of the FCI in capturing financial conditions and its utility in economic forecasting.

Figures and Tables

Figure 1: Shows the FCI and GDP growth for selected Asian economies, illustrating the tightening and recovery phases.
Figure 2: Reports the relative contribution of financial variables to the FCI and GDP growth volatility.
Figure 3: Displays the average contribution of financial variables to the overall FCI over the last 10 years.
Figure 4: Highlights the impact of exchange rates and stock prices on GDP growth volatility.
Table 1: In-sample predictive tests of the FCI for GDP growth.
Table 2: Out-of-sample predictive tests of the FCI for GDP growth.
Table 3: Correlation of real GDP outturns and forecasts with and without the FCI.

Conclusion

The FCI is a robust and informative indicator of financial conditions in Asia, capable of predicting GDP growth and guiding monetary policy decisions. It is particularly effective in capturing the exogenous shifts in financial conditions and has shown strong performance in both in-sample and out-of-sample forecasting. The combined FCI offers the best predictive power, suggesting that integrating different methodologies can enhance the index’s utility for economic analysis and policy-making.

展开完整摘要