2015年-BIS国际清算银行_Bank_capital_shock_propagation_via_syndicated_interconnectedness_42页_7mb

Summary of "Bank Capital Shock Propagation via Syndicated Interconnectedness"

Core Content

This paper investigates how bank capital shocks propagate through the syndicated loan market, focusing on the role of interconnectedness and the implications for financial stability. The study is based on a micro-founded model that incorporates risk management tools such as value-at-risk (VaR) and explores the dynamics of bank behavior in response to equity capital shocks.

Main Points

1. Syndicated Lending and Interconnectedness

• Loan syndication increases bank interconnectedness through co-lending relationships.
• This interconnectedness can lead to contagion effects when banks face capital shocks, as a withdrawal from a syndicate can cause ripple effects across the market.
• Lead arrangers may either increase their own lending or dissolve the syndicate in response to such shocks, which can have significant systemic implications.

2. Efficiency-Stability Trade-off

• Syndicated lending allows banks to diversify credit risks and increase total lending.
• However, the complementarity in lending decisions due to reliance on syndicate partners can also act as a source of financial contagion.
• The paper highlights an efficiency-stability trade-off, where interconnectedness can both stabilize and destabilize the system depending on the shock scenario.

3. Model and Assumptions

• The model assumes risk-neutral banks that optimize their loan portfolios under a VaR constraint.
• Each project is financed by a lead arranger, which underwrites the loan.
• Banks can participate in syndicated loans and receive a return adjusted by a fee for the lead arranger's role.
• The VaR constraint is based on the bank's equity and the expected return and variance of its portfolio.

4. Simulation Results

• The model is calibrated to reflect real-world syndicated loan and bank characteristics.
• Simulations under different network structures and shock distributions show that:
  • A homogeneous-degree network (where each syndicate has the same number of participants) is least stable.
  • It exhibits a bifurcation in outcomes, where a small capital shock can lead to either zero dissolved loans or over 50% dissolution.
  • The empirical network (based on real data) shows a more stable core-periphery structure, suggesting that localized lending may help reduce systemic risk.
  • Rapid adjustments in the market during crises are observed, with most adjustments happening through reductions in the number of loans, not their size.

5. Policy Experiments

• A policy experiment where a highly active bank faces a large negative equity shock shows moderate systemic effects, indicating that the current network structure may limit the impact of such shocks.
• Another experiment imposes tighter VaR constraints, which leads to reduced risk-taking by banks ex-ante, suggesting that tighter regulation can help mitigate systemic risk.

Key Information

• JEL Classification: E44, E52, G12, G20, E32.
• Keywords: Syndicated lending, systemic risk, network externalities, value at risk, bank capital shocks, rare event risk.
• Empirical Evidence:
  • In 2008, the volume of syndicated loans contracted by about 43%.
  • Lead arrangers' loan share increased during the crisis, suggesting they compensated for withdrawals by other participants.
• Model Findings:
  • The core-periphery structure in the empirical network helps mitigate systemic risk.
  • The VaR constraint is a key mechanism for risk management and can influence the level of risk-taking and the probability of contagion.
• Implications:
  • The structure of the syndicated loan market plays a critical role in the propagation of capital shocks.
  • The interconnectedness of banks through syndicated lending can amplify market disruptions during crises.
  • Strategic complementarities in bank behavior are essential for understanding non-linear and threshold-based systemic events.

Conclusion

The paper contributes to the understanding of how bank capital shocks can propagate through the syndicated loan market, highlighting the role of network structure, VaR constraints, and risk-sharing mechanisms. It suggests that empirical network structures may provide greater stability compared to more uniform ones, and that policy interventions such as tighter VaR constraints can help reduce risk exposure and mitigate contagion. The study emphasizes the need for a deeper understanding of interconnectedness in financial systems to improve financial stability and systemic risk management.