Credit risk measurement and pricing in performance pricing-based debt contracts

This dissertation examines how lenders (1) choose a credit risk proxy and (2) link changes in that proxy to the risk premium charged on the loan in performance pricing-based debt contracts. I focus on the choice between accounting ratios and debt ratings as the credit risk proxy. I analyze three attributes of these two measures: timeliness, comprehensiveness, and costliness. I find that performance pricing loans that use accounting ratios are significantly riskier and more likely to be secured than debt rating-based loans. I also find that debt rating-based performance pricing loans are more likely to be given to larger borrowers and borrowers with high absolute levels of accruals.; I find mixed results in comparing the actual pricing of performance pricing grids to a theoretical debt pricing model. As expected, the total change in risk premium across the pricing grid is positively associated with the total change in the credit risk proxy across the pricing grid, the average credit risk location of the grid, and the volatility of the borrower. In addition, as predicted, the risk premium is convex in the credit risk proxy, on average, within the performance pricing grid. However, many of the individual contracts do not satisfy the convexity prediction and are not well explained by existing debt pricing theory.