Sectoral credit allocation and interest rate spreads worldwide

ABSTRACT

Using a comprehensive quarterly database covering 22 advanced economies and 18 emerging market economies spanning the period from 1990 to 2018, this study provides evidence indicating that a higher ratio of total credit to the non-financial sector relative to GDP is associated with wider interest rate spreads. Importantly, this relationship between the stock of debt and spreads remains robust even after controlling for measures of debt flows in both advanced and emerging countries, suggesting that the stock of debt carries more valuable information for capturing a country’s financial position in international markets compared to the flow of debt. Moreover, through a detailed exploration of credit allocation across different sectors, this research reveals that the provision of credit to households and the government plays a crucial role in understanding the dynamics of interest rate spreads in both advanced and emerging countries. These empirical findings contribute to addressing the existing gap in understanding the determinants of the risk premium associated with cross-border financing activities. Additionally, the observed correlations between credit measures and spreads offer important empirical moments that should be incorporated into international macro models and financial policy.

KEYWORDS:

• International finance
• credit
• capital flow
• interest rate spread

JEL CLASSIFICATION:

• F34
• F41
• F44
• G15

Acknowledgements

I am deeply grateful to valuable comments from the editor, Mark Taylor, and two anonymous referees.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

¹ See Kwak (2023) who establishes the causal relationship between non-financial corporate debt and sovereign interest rate spreads in the European countries, using idiosyncratic shocks to firms as an instrument. He also finds that the causal estimate is qualitatively similar to the correlation, although the causal estimate is larger that the ordinary least square (OLS) estimate.

² This standard error is robust to general forms of cross-sectional and temporal dependence.

Additional information

Funding

This work was supported by the Sogang University Research Grant of 2023 (202310033.01).