BIS国际清算银行-The-macro-financial-effects-of-international-bank-lending-on-emerging-markets_66页_1mb

Summary of BIS Working Paper No. 899: The macro-financial effects of international bank lending on emerging markets

Core Content

This working paper investigates the macro-financial effects of international bank lending on emerging market economies (EMEs), focusing on the causal impact of cross-border credit flows on domestic economic and financial variables. The study is based on a sample of 22 EMEs over the period 1990Q1–2018Q4 and employs a granular instrumental variable (GIV) approach to address endogeneity concerns.

Main Points

• Capital inflows are seen as both expansionary and volatile, potentially leading to boom-bust cycles in EMEs.
• International bank lending is a significant component of gross liabilities in EMEs, especially when foreign direct investment (FDI) is excluded.
• The causal effect of international bank lending is estimated using GIVs, which are constructed by exploiting heterogeneity in the size distribution of bilateral bank lending.
• The paper argues that capital controls can help mitigate the volatility of international lending and reduce the associated macro-financial risks.
• The GIV approach is a novel method in the context of international bank lending and is shown to be more effective than other commonly used instruments, such as the excess bond premium or US broker-dealer leverage.

Key Findings

• Positive causal effects: Cross-border bank credit leads to looser financial conditions, appreciation of exchange rates, narrowing of sovereign and corporate spreads, falling domestic interest rates, and increased housing prices in EMEs.
• Real domestic credit and GDP growth: The shock to international bank lending is associated with real domestic credit growth and real GDP expansion, primarily driven by investment.
• Robustness: The results are robust to various checks, including the role of crises, the sample of lending countries, and the method of constructing GIVs.
• Capital controls: Countries with higher capital inflow controls experience weaker effects of international bank lending, suggesting that such controls can dampen macro-financial vulnerabilities.

Methodology

• Empirical Strategy: The paper uses a two-stage least squares (2SLS) approach with local projections and instrumental variables.
• Granular Instrumental Variables (GIVs): These are constructed by:
  • Removing common shocks from bilateral lending data.
  • Using principal component analysis (PCA) to estimate idiosyncratic shocks.
  • Leveraging heterogeneity in the size distribution of bilateral claims to create exogenous shocks.
• Exclusion Restriction: The GIVs are orthogonal to global financial cycle indicators, ensuring they are valid instruments for identifying the causal effect of international bank lending.
• Model Extension: The paper extends the basic model to a panel structure, allowing for:
  • Heterogeneous exposures across lending sources.
  • Reverse causality between bank lending and domestic macro-financial variables.
  • Time-varying shares of lending sources and additional controls.

Contributions to Literature

• First application of GIVs to international bank lending and capital flows.
• Direct estimation of the macro-causal effect of capital flows on EMEs.
• Non-linear panel setting and dynamic effects of bank lending shocks.
• Robust comparison with existing instruments, highlighting the superiority of GIVs in terms of exogeneity and relevance.

Conclusion

The study concludes that international bank lending has significant macro-financial effects on EMEs, and capital controls can effectively mitigate these effects. The GIV approach provides a valid and robust identification strategy, improving the accuracy of causal inference in the context of international financial flows.

Key Information

• Sample Period: 1990Q1–2018Q4
• Sample Size: 22 EMEs
• Methodology: Granular instrumental variables (GIVs) using BIS locational banking data
• Main Variables: Financial condition indices, exchange rates, sovereign and corporate spreads, domestic interest rates, real GDP, and housing prices
• Capital Controls Index: Used to assess the effectiveness of capital inflow controls
• JEL Classification: E0, F0, F3
• Keywords: Granular instrumental variables, capital flows, emerging markets, cross-border claims, credit shocks, international banking, capital controls

Structure

• Introduction: Motivates the study of international bank lending on EMEs.
• Section 2: Provides a perspective on cross-border bank lending and its role in EMEs.
• Section 3: Details the empirical strategy and construction of GIVs.
• Section 4: Describes the data used in the analysis.
• Section 5: Presents and discusses the results.
• Section 6: Conducts robustness checks on the GIV construction.
• Section 7: Compares GIVs with other instruments.
• Section 8: Concludes with policy implications and methodological contributions.