Detection of outliers in the financial time series using ARIMA models

The subject of this work is a comparative analysis of selected models used to describe the volatility of financial time series with attention to the problem of exceptions. According to the definition provided by Moore and McCabe in 1999 [3], in statistics, outliers are the observations that fall outside the overall pattern of a distribution. This paper describes common approaches to detecting outliers, modelling dataset describing real-world processes in financial area and making forecast. Based on the approach indicated by A. J. Fox in 1972, this subject focuses on using the selected ARIMA models for forecasting the future value of data set which containing outliers. The studies showed that the typical features of financial time series are the so- called grouped variances. Therefore, using ARIMA models for forecasting was insufficient, as data with effect of grouping variances requires models considering these effects, like ARCH, GARCH.