Research of the Kazakhstan Financial Market with Long Memory Models

The main goal of this work is to identify significant patterns in the stock price market, which are traded on the Kazakhstan market, which will help with the conclusion that you should use a model with a long memory, or indicate problems with which we cannot use a model with a long memory in work. See if forecast accuracy is improved using this model. Thus, the following tasks are set to achieve the goal: 1. Familiarize yourself with the term “long memory”, define and review the literature. 2. Show frequently encountered processes with long memory, and describe their properties. 3. Build ARIMA and ARFIMA models on the uploaded data, build a forecast, and evaluate the quality of forecasting of these models. 4. Final results from point 3 The first chapter talks about the history of the term “long memory”, about common processes / models with long memory and about their properties, about the criterion for selecting the “best” model among several. The second chapter, respectively, is devoted to the study of the price of shares traded on the financial market of Kazakhstan, the unloading of data, the search for patterns in improving the accuracy of the forecast