Downloads: 10 | Views: 155 | Weekly Hits: ⮙1 | Monthly Hits: ⮙4
Research Paper | Mathematics and Statistics | North Macedonia | Volume 14 Issue 1, January 2025 | Rating: 5.6 / 10
Optimization of Ornstein-Uhlenbeck Models in Pair Trading: Heuristic and Metaheuristic Approaches
Nikola Basheski, Vesna Dimitrievska Ristovska
Abstract: Stochastic processes are widely used in describing time dependent sequences of random changes in both physical and social phenomena, including, relevantly to this paper, random market movements of stock prices. Optimization techniques are applied in finance, for example, in Banking processes such as minimizing losses from credit risks and changes of the value of collateral assets, and in trading - such as assessing optimal position sizes and buy/sell times. This study explores heuristic and metaheuristic optimization methods for financial modeling in pair trading. Using Genetic Algorithm and Simulated Annealing, the study aims to optimize parameters of the Ornstein-Uhlenbeck process, a stochastic model used in trading decisions. Performance is evaluated on simpler test cases before applying these methods to minimize the log-likelihood function of the process. Results show the effectiveness of these algorithms in enhancing trading strategies, providing robust insights for financial optimization.
Keywords: Genetic Algorithm (GA), Simulated Annealing (SA), Ornstein-Uhlenbeck (O-U) process
Edition: Volume 14 Issue 1, January 2025,
Pages: 93 - 98