Abstract
Aimed at studying the Multi-Commodity Exchange of Kazakhstan through the use of stochastic models of financial forecasting in the context of globalization. Today, nonlinear forecasting models, including artificial neural networks, are widely used in financial forecasting. Thus nonlinear models, in contrast to linear forecasting models, require higher computing systems and data for practical training and data processing. The purpose of the article is to assess the state of the Commodity Exchange of Kazakhstan in financial forecasting using stochastic models, thus studying the importance of financial forecasting for the correct adoption of investment decisions on the KASE Commodity Exchange. The article shows a systematic approach to creating a simplified and effective forecasting model that allows you to make optimal decisions using minimal calculation systems in the process of making optimal investment decisions. Natural gas futures traded on the commodity and raw materials exchange of Kazakhstan were identified as a suitable object for research, given the need for massive changes in the energy structure of Kazakhstan. In the article, we used data analysis and statistical methods to determine optimal forecasting strategies and features, as a result of which we studied and determined an accurate linear forecasting model. The analysis of the data also determined the context of the study, identifying a positive change in relation to the fuel and energy base of market participants, definitions were made using a significant change in the assessment of this research object. In the article, when developing a financial forecasting model, several areas were studied, including the identification of weekly and annual patterns, studies of the introduction of seasonal and exogenous variables. The article concludes that attempts to reduce external dependence and separate data on the Shum increase the performance of a model with a minimum of computational data and systems. Key words: Kazakhstan, KASE, forecasting, data analysis, time series analysis, commodity markets, natural gas