BIS research uncovers non-linear link between collateral and default

Traditional economic theories suggest that pledging collateral reduces default likelihood.

However, this paper, using comprehensive loan-level data from the Eurozone, documents a non-linear relationship.

The study finds that increasing the ex-ante collateral-to-loan ratio initially lowers the default probability, consistent with theory.

Yet, as loans become overcollateralized, further increases in the ratio correlate with higher default probabilities.

This surprising U-shaped pattern is driven by the riskiness of the collateral itself.

The research estimates that a higher collateral-to-loan ratio correlates with greater variance in the underlying collateral's market value after loan origination, making collateral an intrinsic and measurable feature of secured lending.

This holds true across different collateral types and regardless of eventual default, highlighting that collateral is far from risk-free.

The study resolves long-standing discrepancies in empirical findings, where previous research often showed riskier borrowers pledging more collateral or secured loans having worse repayment outcomes.

The authors attribute this to economists' informational disadvantage and the common practice of classifying loans as simply secured or unsecured, which imposed a linear relationship.

By accounting for time-varying bank- and firm-specific risk factors and using granular data that captures non-linear dynamics, the paper reveals correlations between collateral and default consistent with theory.

The novel dataset, including initial appraised value and monthly market values, enables precise measurement of value fluctuations, offering a more nuanced understanding of credit risk.