COMPARISON OF SYSTEMS OF FORECASTING THE DIRECTION OF CHANGES IN THE EXCHANGE RATE OF A FINANCIAL INSTRUMENT USING SIMPLE, EXPONENTIAL AND LINEAR WEIGHTED MOVING AVERAGES

Abstract

Several modern scientific studies point to the fact that technical analysis indicators have a predictive power of various types. Accordingly, trading systems for working in financial markets built using them may have some practical value. This work explores a specific range of topical issues of development, testing, and implementation of a trading system that generates an instruction to execute a stock transaction based on the signals of technical analysis indicators, particularly exponential, weighted, and simple moving averages. The works of modern scientists have been analyzed in which options for using these indicators are considered. In this context, the current research aims to analyze the influence of linearly weighted moving average settings and their combinations on the profitability of the trading system, as well as to compare the results of such a system with the results of a strategy built on a combination of simple and exponential moving averages. General scientific (analysis, synthesis, comparison, modeling) and special (testing, statistical analysis, graphical, tabular) research methods were used to achieve the goal. Based on this, the tasks solved using indicators of this kind were formed. A methodology for selecting moving averages and their settings is proposed when creating, testing, and implementing a trader's trading system. A methodology for selecting moving averages and their settings is proposed when creating, testing, and implementing a trader's trading system. Several approaches to forming and interpreting the signal regarding the change in the rate of a financial asset generated by the system are considered. The work also analyzes the criteria for comparing the results of strategies at the testing stage. The results of the application of various options of strategies were calculated and compared, and the optimal ones were selected according to the given selection criteria. The simulation of trading operations was performed for the currency pair EUR/USD, and weekly quotes from 1999 to 2023 were used, based on which the optimal combination of indicators for use in the trading strategy was determined. It is also stated that the system, built on moving averages, has shortcomings and needs additional optimization. Variants of possible optimization and the corresponding toolkit that can be analyzed are indicated. Based on the results of the research, it was concluded that the proposed approach to the development and use of the trader's trading system can be used to perform accurate exchange operations.