Abstract
We consider autoregressive conditional heteroskedasticity (ARCH) processes in which the variance depends linearly on the absolute value of the random variable y as While for the standard model, where the corresponding probability distribution function (PDF) decays as a power law for in the linear case it decays exponentially as with We extend these results to the more general case with We find stretched exponential decay for and stretched Gaussian behavior for As an application, we consider the case as our starting scheme for modeling the PDF of daily (logarithmic) variations in the Dow Jones stock market index. When the history of the ARCH process is taken into account, the resulting PDF becomes a stretched exponential even for with a stretched exponent in a much better agreement with the empirical data.
- Received 13 November 2001
DOI:https://doi.org/10.1103/PhysRevE.65.046149
©2002 American Physical Society