Model of General Split-BREAK Process

Abstract

This paper presents a modification (and partly a generalization) of STOPBREAK process, which is the stochastic model of time series with permanent, emphatic fluctuations. The threshold regime of the process is obtained by using, so called, Noise indicator. We proceed to investigate the model which we named the General Split-BREAK (GSB) process. After brief recalling of its basic stochastic properties, we give some procedures of its parameters estimation. A Monte Carlo study of this process is also give, along with the application in the analysis of stock prices dynamics of several Serbian companies which were traded on Belgrade Stock Exchange.

Keywords

• GSB process
• STOPBREAK process
• noise indicator
• Split-MA process
• stationarity
• invertibility