Published in:IMF Working Papers, Vol. , pp. 1-52, 2007
Extent:
1 Online-Ressource (54 p)
Language:
Not determined
DOI:
10.2139/ssrn.1078795
Identifier:
Origination:
Footnote:
Nach Informationen von SSRN wurde die ursprüngliche Fassung des Dokuments December 2007 erstellt
Description:
Using monthly data for a set of variables, we examine the out-of-sample performance of various variance/covariance models and find that no model has consistently outperformed the others. We also show that it is possible to increase the probability mass toward the tails and to match reasonably well the historical evolution of volatilities by changing a decay factor appropriately. Finally, we implement a simple stochastic volatility model and simulate the credit transition matrix for two large Brazilian banks and show that this methodology has the potential to improve simulated transition probabilities as compared to the constant volatility case. In particular, it can shift CTM probabilities towards lower credit risk categories