Integration, Kointegration und die Langzeitprognose von Kreditausfallzyklen

Summary: This paper examines the longterm forecast performance of cointegrated systems relative to forecast performance of comparable VAR that fails to recognize that the system is characterized by cointegration. I use Monte Carlo simulation, real data sets, and multi-step-ahead forecasts to study this question. The cointegrated system I examine is composed of six vectors, five macoreconomic variables, and a credit-default-cycle. The forecasts produced by the vector error correction modell associated with this system are compared with those obtained from a corresponding differenced vector autoregression, as well as a vector autoregression based upon the levels of the data. Alternative measures of forecast accuracy (full-system) are discussed. My findings suggest that selective forecast performance improvement may be observed by incorporating knowledge of cointegration rank. Furthermore the results indicate that a cointegration modeling of credit risk should be favored against the prevalent level or differenced estimation.