Resampling slashes credit risk VAR underestimates – research

Academics claim Vasicek model’s underestimation tendency can be slashed to near-zero

It’s a perennial dilemma for credit risk managers: how do you gauge an accurate picture of risk exposure on loan portfolios where data is thin or discontinuous? If available data only covers a few years, a lender could be significantly underestimating its credit risk value-at-risk. But, according to new research, judicious use of resampling techniques can remove most of this risk, where banks are applying one of the most commonly used interest rate risk models.

In research published in the Jour