Application of optimal stopping to model sales in financial markets: Examination and analysis

High-precision prediction of financial prices is still a deemed long-term challenge that constantly calls for state-of-the-art approaches. Thus, the purpose of the current study was to examine the efficacy of optimal stopping and use its connection with branching processes to predict several financial markets. For this purpose, the S&P ۵۰۰ index and dollar, gold, oil and bitcoin markets are predicted at ۵-, ۱۰-, ۳۰-, ۵۰-, and ۱۰۰-day forecast horizons, for each of which the optimal buying and selling point was determined. Closing price data for at least ۲۲۰۰ trading days during the period ۲۰۱۳-۲۰۲۱ was used for the purposes of this study. Moreover, given that any prediction and decision in the financial markets is highly based on probability, and hence risk, two strategies are devised for examination, namely (۱) high risk (success rate of at least ۵۰%) and (۲) low risk (success rate of at least ۷۰%). The findings indicated that the estimations on all the indices and prices were relevant for the high-risk scenario (that is a success rate of at least ۵۰%), while only those on the S&P ۵۰۰ index and price of gold were relevant for the low-risk scenario (success rate of at least ۷۰%).

Keywords:

pplied probability , optimal stopping , branching processes , simulation modeling , financial markets forecasting