When Banks Punch Back: Macrofinancial Feedback Loops in Stress Tests

IMF Working Papers, May 29, 2020

Source details

Canonical URL: When Banks Punch Back: Macrofinancial Feedback Loops in Stress Tests

Title: When Banks Punch Back: Macrofinancial Feedback Loops in Stress Tests
Authors: Mario Catalan, Alexander W. Hoffmaister
Date: May 29, 2020
Core contribution: Develops a credit response and externalities analysis model (CREAM) that integrates a disaggregated banking sector into a standard macroeconomic structural vector autoregressive model to analyze macro-financial feedback loops.

Model name: credit response and externalities analysis model (CREAM).
Integration: disaggregated banking sector embedded in a macroeconomic structural vector autoregressive framework.
Focus: banking-sector heterogeneity in lending responses and the propagation of externalities through macro-financial feedback loops.

Simultaneous capital losses across multiple banks can prompt banks to contract their balance sheets, generating externalities that propagate as macro-financial feedback loops.
Accounting for macro-financial feedback loops can significantly affect:
- macroeconomic outcomes, and
- bank-specific stress test results.
Heterogeneity in bank lending responses matters:
- It determines how each bank fares under adverse conditions.
- It determines the external effects that banks impose on each other and on economic activity.
The model can be used to assess contributions of individual banks to systemic risk along the time dimension.

Stress tests that omit macro-financial feedback loops may misstate both macroeconomic impacts and individual bank vulnerabilities.
Disaggregated bank-level analysis of lending responses is essential to:
- understand bank-specific outcomes under adverse shocks, and
- identify the externalities banks impose on the broader economy and on each other.
CREAM provides a tool to attribute systemic risk contributions to individual banks over time.