2025-05-13-美联储-量化宽松_银行流动性风险管理和非银行融资_来自美国行政数据的证据（英）_62页_1mb

Summary of "QE, Bank Liquidity Risk Management, and Non-Bank Funding: Evidence from U.S. Administrative Data"

Abstract and Objective

This study examines the effects of quantitative easing (QE) on bank liquidity risk management and non-bank funding during the COVID-19 pandemic. Using granular U.S. administrative data, the analysis explores how banks adjust their deposit composition and credit supply in response to increased fragility from uninsured deposits sourced from non-bank financial institutions (NBFIs). The research identifies QE as a key driver of funding fragility and assesses its unintended consequences for corporate liquidity and investment.

Key Findings

• Bank Responses to QE: Banks with higher pre-pandemic exposure to NBFIs experienced a significant surge in uninsured deposits during QE. They actively managed liquidity risk by:

  • Reducing rates on uninsured deposits and increasing rates on insured deposits.
  • Shifting deposit composition toward more stable funding sources.
  • Cutting credit line commitments, particularly undrawn credit lines, to firms.

• Real-Effects: The reduction in liquidity insurance from banks constrained firms' access to contingent liquidity, leading to lower investment. While total credit supply was unaffected, the decline in liquidity-providing credit lines amplified the real impact on firms reliant on exposed banks.

• Theoretical Relevance: The paper extends the Kashyap et al. (2002) model to incorporate runnable deposits, demonstrating that the complementarity between deposit-taking and credit line issuance breaks down under heightened liquidity risk, offering a new explanation for the limitations of QE transmission mechanisms.

• Aggregate Effects: Firms with greater reliance on exposed banks saw reduced liquidity support, contributing to a decline in investment amid diminished capacity to hedge future liquidity shocks.

Methodology

• Data: Utilizes FR 2052a (granular deposit data) and FR Y-14Q (loan-level data) matched with Call Reports and RateWatch. The sample comprises 29 large U.S. banks from 2016-Q1 to 2022-Q4.
• Identification: Employs a difference-in-differences approach, leveraging pre-pandemic exposure to NBFIs as an exogenous shock to isolate QE effects.
• Approach: A multi-stage empirical strategy using bank- and month-fixed effects regressions, controlling for bank characteristics and policy changes like SLR adjustments.

Conclusions and Implications

This research underscores that while QE injects liquidity into the banking system, it exacerbates funding fragility through uninsured NBFIs. Banks' proactive liquidity management, while rationalizing risks, inadvertently reduces contingent liquidity for firms, constraining investment. These unintended consequences highlight the need for policymakers to consider the interplay between central bank policies and banks' incentives, emphasizing that QE's transmission may be more nuanced than anticipated. The findings call for further research on the role of NBFIs and regulatory frameworks to mitigate liquidity risks in monetary policy implementation.