ESTIMATION AND TESTING IN MODELS CONTAINING BOTH JUMPS AND CONDITIONAL HETEROSCEDASTICITY

In this article we develop a test for the hypothesis that a series (ob served in discrete time) is generated by a diffusion process. This tes t is based on an overidentifying relation between variance and kurtosi s parameters that holds for generalized autoregressive conditional het eroscedastic diffusions. The proposed test is not specific to a partic ular data frequency and clearly indicates the presence of jumps in dol lar exchange rates. To assess the size and intensity of the jumps, we estimate a model containing both jumps and conditional heteroscedastic ity.