Dynamic effects of financial spillovers on bank lending: evidence from local projection-based impulse response analysis

ABSTRACT

Using local projections, we find that a positive (detrimental) financial stress (FS) shock affects bank lending more quickly and is more persistent than a negative (benign) FS shock. The study confirms that there is timing asymmetry, but we do not find magnitude asymmetry between positive and negative shocks. Furthermore, we find that a bank balance sheet characterized by a higher level of liquidity and efficiency and a lower level of leverage tends to better insulate lending from the impact of FS spillovers. Our findings have important implications for policymakers who seek to develop more efficient and resilient banking sectors.

KEYWORDS:

• Financial stress
• spillovers
• bank lending
• local projections

JEL CLASSIFICATION:

• G15
• G21

Acknowledgments

The authors are grateful to the editor and anonymous referee for their helpful comments on an earlier version of the paper. Wang Chen is also grateful to Jinan University for the support by Ning-Jing-Zhi-Yuan program.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ The method is more robust to misspecification and easily accommodates non-linearities. Moreover, it can circumvent the need to impose dynamic restrictions, which is typically embedded in the traditional VAR.

² The KCFSI is a composite index designed to measure the level of stress in US financial markets.

³ The country list is available upon request. The sample period is determined on the basis of the availability of bank-level data.

⁴ We exclude foreign banks with US country ownership from our sample because these banks are directly exposed to US financial market conditions and, thus, are likely to behave differently than other foreign banks. In addition, banks with many missing values in the data are excluded from the sample. We also winsorize the bank-level explanatory variables at the 0.5% and 99.5% levels to exclude the disturbance of the outliers.

Additional information

Funding

This work was supported by the Social Sciences Funding Program of Guangdong Province, China  [GD16YYJ05]; Japan Society for the Promotion of Science [17H00983, 18K01610]; Humanities and Social Sciences Funding Program of Chinese Ministry of Education [18YJC790012]; Natural Science Foundation of Guangdong Province, China [2018030310396]; National Natural Science Foundation of China [71803063].