The Relationship between Risk-Neutral and Actual Default Probabilities : The Credit Risk Premium

The relationship between the risk-neutral measure Q and the actual or real-world measure P, and the corresponding credit risk premium, are investigated in this paper. Quantifying and understanding the long-term average risk premium is important for a variety of financial applications and investment decision-making. This study develops an empirical analysis of this relationship, using CDS spreads of European corporates for estimating risk-neutral probabilities, and Moody's historical transition matrices to derive the corresponding actual values. Special attention is given to the recent financial crises and our study allows us to quantify its impact on risk premia. In line with some research based on pre-crisis data, we find that the ratio between the risk-neutral and actual default intensities, which we call the coverage ratio, is a convex and decreasing function of the actual default intensities. We are able to further differentiate between different time-horizons and conclude that current risk premia levels are still above their initial levels and this could indicate a permanent upward shift in risk premia. Finally, we link our results with the concept of Real Economic Value and its role in the bail-out of several European financial institutions