Estimating of the default risk premium and non-normality

Purpose: To highlight the difficulties in calculating the default risk premium given irreconcilable results in the literature which fail to agree on the sign or magnitude of the expected return differential between distress risk portfolios of relatively high and low exposure. Originality: In contrast to previous work, we study the pricing of the default risk premium based on measures of central tendency derived from the multivariate-t distribution - a class of distribution that accommodates important empirical features of monthly equity returns on distress-risk based portfolios. Findings: Our findings suggest that estimates of the default risk premium are sensitive to the measure of central tendency and show that the time series of returns on distress-risk sorted portfolios provides very little information about the magnitude of the true distress premium under a data generating process that is consistent with observed returns.