General Stationary AR and MA Time Series Models with Min and Max-Min Structures

摘要

Autoregressive process other than additive structures have already attained the attention of several researchers. Process with minimum and maximum as components have important applications in hydrology, oceanographic studies and financial markets. In this paper, a general class of autoregressive process is introduced and condition for strict stationarity is studied. Using this general structure of the process one can develop autoregressive models of given marginal distribution. Several examples and sample path properties of the process are discussed.