EBA欧洲银行-Session-4-Semi-Structural-Credit-Gap-Estimation-J.-H.-Lang2C20P.-Welz_41页_1mb

Summary of "Semi-Structural Credit Gap Estimation"

Core Content

This paper introduces a semi-structural approach to estimate household credit gaps, aiming to identify excessive credit developments that may signal financial crises. The method is based on a structural economic model that incorporates economic fundamentals, such as demographic structure, institutional quality, and potential GDP, to derive a normative trend level for household credit. The credit gap is then defined as the deviation of actual household credit from this trend, offering a theory-based interpretation rather than relying solely on statistical filters.

The model is derived from the overlapping generations framework by Eggertsson and Mehrotra (2014), which is modified to include institutional quality and demographic factors. The model assumes that borrowing constraints are determined by a non-linear function of institutional quality, reflecting the idea that institutional development can significantly influence the level of credit. The equilibrium real interest rate is also a key determinant of the credit trend, and the credit cycle is modeled as a residual statistical process with an AR(2) structure.

The paper estimates the model for 12 EU countries (Belgium, Denmark, Finland, France, Germany, Ireland, Italy, Netherlands, Portugal, Spain, Sweden, and Great Britain) using quarterly data from 1980 to 2015. The results show that credit cycles last between 15 to 25 years with amplitudes of around 20%, and that the semi-structural credit gaps have superior early warning properties for financial crises compared to purely statistical measures like the Basel credit-to-GDP gap. This is because the semi-structural model accounts for economic fundamentals, avoiding the pitfalls of statistical methods that may overemphasize long-term trends and understate cyclical risks.

Main Points

• Excessive credit growth is a key driver of financial instability, including the global financial crisis.
• The semi-structural approach uses economic theory to derive a normative credit trend.
• The model incorporates demographic structure, income inequality, institutional quality, and potential GDP as fundamental determinants of the credit trend.
• Borrowing constraints are modeled as a function of institutional quality, with a logistic transformation used to reflect non-linear relationships.
• The credit cycle is modeled as a residual AR(2) process, allowing for flexible and dynamic analysis.
• The semi-structural credit gaps are estimated for 12 EU countries and show significant amplitude and predictive power for financial crises.
• The model avoids the issues of purely statistical credit gaps, such as over-persistence and underestimation of cyclical risks.
• The early warning properties of the semi-structural credit gap are superior to those of the Basel credit-to-GDP gap.

Key Information

• Model Basis: Overlapping generations model (Eggertsson and Mehrotra, 2014), modified to include demographic and institutional factors.
• Variables Used:
  • Real household credit
  • Potential GDP
  • Equilibrium real interest rate
  • Institutional quality (non-linear transformation)
  • Population ratio (young/middle-aged to all income-receiving individuals)
• Methodology:
  • Unobserved components system
  • Maximum likelihood estimation with Kalman filter
  • Incorporates economic theory in a semi-structural manner
• Empirical Findings:
  • Credit cycles last 15–25 years
  • Amplitude of cycles is around ±20%
  • Semi-structural credit gaps increase before financial crises and decrease after
  • The model avoids excessive persistence and overestimation of risk seen in statistical credit gaps
• Policy Relevance:
  • The semi-structural credit gap can provide useful information for countercyclical macroprudential policy
  • It improves the early warning capabilities for financial crises

Conclusion

The paper contributes to the macroprudential analysis by proposing a theory-based method for estimating credit gaps. This approach offers a more accurate and interpretable measure of credit excesses, which can help policymakers better anticipate and respond to financial instability. The model is robust to structural changes and captures the interaction between financial and business cycles, making it a valuable tool for systemic risk assessment and policy formulation.