A Comparative Study between the KMV and the Zero-Price Probability for Default Prediction

Abstract: This master’s thesis is a comparative study between a structural model and a simulation based model for predicting default-probabilities. The structural model used is the KMV model and the simulation based is a new model called Zero-Price Probability. The main focus is on the new simulation based approach rather than the older established models. The comparison is done to get an implicit idea of the power of both models. This was done by calculating the Zero-Price Probability using Monte Carlo simulations and also calculating the KMV, both models are presented for 500 observations and on a forecast horizon of one year. They are presented for several firms, both defaulted and of good financial health, from different industries and regions. The results in this paper show that the Zero-Price Probability seems to be better at forecasting default for firms that have defaulted, however it also overestimates the default-probability of firms that haven’t defaulted.