SELECTION OF MEDIUM-TERM INVESTMENT STRATEGIESWITH THE USE OF GENETIC ALGORITHM

Abstract

This work examines the problem of selection of investment strategies for an individual investor over the course of a middle-length period. A new approach is proposed for investment portfolio selection using the genetic algorithm that is working based on the financial indicators of stock price movement over the course of the previous year, which is then used to select an investment portfolio from all stocks matching certain values of those indicators.. An analysis of existing approaches to investing using genetic algorithms was carried out. The selection of investment strategies for the medium term (one year) period in the conditions of the modern stock market was carried out on the basis of historical data using a genetic algorithm. The results are compared with other common investment strategies and it is shown that the approach to the selection of investment portfolio positions using a genetic algorithm can outperform both the overall market growth and other popular investment strategies. The development and optimization of the latest genetic algorithm for medium-term investment problems of an individual investor using financial indicators of positions on the stock market, such as a moving average for analyzing the movement of the share price and a price oscillator for analyzing the volatility of price changes, was carried out. The regularities between the financial indicators of a position on the stock market and the results of changes in its price in the future are characterized. The results of the algorithm in different situations of the stock market are analyzed and potential future approaches to expand the research into this area are proposed. The developed algorithm allows selection of positions on the stock market depending on the results of changes in their prices for previous periods.