世界发展银行-Banking-Sector-Performance-During-the-COVID-19-Crisis_51页_2mb

Summary of "Banking Sector Performance During the COVID-19 Crisis"

Core Content

This working paper by Asli Demirguc-Kunt, Alvaro Pedraza, and Claudia Ruiz-Ortega examines the performance of bank stocks during the early stages of the COVID-19 pandemic and evaluates the impact of financial sector policy interventions. The study uses a global database of policy responses and bank stock data from 53 countries, covering 896 commercial banks and 2,147 non-bank financial institutions. The analysis focuses on the market reaction to different types of policy measures, including liquidity support, prudential measures, borrower assistance, and monetary policy.

Main Findings

1. Bank Stock Underperformance

• Bank stocks underperformed both domestic markets and other non-bank financial firms during the early months of the pandemic (March to April 2020).
• The underperformance was more pronounced for less liquid banks, banks with higher oil sector exposure, and public banks.
• Larger banks showed relatively better performance compared to smaller banks, especially after the initial shock.

2. Policy Impact on Bank Stock Returns

• Liquidity support measures had a positive effect on bank stock prices, reducing the liquidity risk premium.
• Borrower assistance also had a strong immediate impact, particularly in developed countries. However, in developing countries, these measures had no significant effect or even a negative impact.
• Prudential measures had no significant aggregate effect on bank stock returns, as markets priced in the downside risk from capital buffer depletion and riskier lending.
• Monetary policy (such as rate cuts and asset purchases) had mixed results, with no overall increase in stock prices, but a reduction in liquidity premium.

3. Regional and Institutional Differences

• Public banks faced more severe underperformance compared to private banks.
• Oil-exposed banks showed a negative correlation with oil price movements, with those in the 75th percentile of oil exposure experiencing lower returns.
• There was large heterogeneity in public bank ownership across regions, with India having 53% of public banks, while high-income countries had only 2%.

4. Policy Interventions Overview

• Policy measures were categorized into four types:
  • Liquidity support: Central bank refinancing, collateral frameworks, and swap agreements.
  • Prudential measures: Relaxation of regulatory and supervisory requirements, including capital buffers and stress tests.
  • Borrower assistance: Government guarantees, direct credit lines, and interest rate subsidies.
  • Monetary policy: Rate cuts, quantitative easing, and credit easing.
• 429 policy announcements were analyzed across 44 countries (16 developed, 28 developing), with prudential measures being the most common (41%), followed by liquidity support (23%) and borrower assistance (21%).

Key Insights

• Banks were expected to absorb the economic shock, but this role came at the cost of increased stress on the banking sector.
• Market expectations of profit losses and capital risk played a crucial role in shaping the risk premium for banks.
• Policy effectiveness varied depending on the type of measure, the country's economic context, and the bank's characteristics (e.g., size, liquidity, ownership).
• The study highlights the importance of liquidity support in stabilizing bank stocks, while borrower assistance had limited impact in developing economies due to fiscal constraints.
• Prudential measures had a mixed reception, with developing countries experiencing negative market responses.

Methodology and Data

• The authors used event study methodology to assess the market response to policy announcements.
• Abnormal returns were calculated using a market model:
  $$
  A R e t _ {b, t} = R _ {b, t} - \hat {\alpha} _ {b} - \hat {\beta} _ {b} R M _ {t}
  $$
• Ex-ante liquidity was measured by the cash-to-total assets ratio, and oil exposure was calculated using OLS regression on stock returns and oil prices.
• Bank size was measured using the Inverse Hyperbolic Sine (IHS) transformation of total assets.
• The data spanned from May 2018 to May 2020, and included daily stock prices, quarterly financial statements, and state ownership.

Conclusion

The paper concludes that while policy interventions helped mitigate the adverse effects of the pandemic on the banking sector, their effectiveness was not uniform. Liquidity support and monetary easing had a positive impact, while prudential measures and borrower assistance had mixed or negative effects, depending on the country and bank type. The study emphasizes the need for careful monitoring of bank vulnerabilities, particularly in developing economies, where fiscal space was limited. It also highlights the importance of liquidity buffers and the role of market expectations in shaping the risk premium for banks during the crisis.