Multi-armed bandit approach to portfolio choice problem

Ozkaya, Guneykan; Wang, Yaping

Multi-armed bandit approach to portfolio choice problem

License

Citation

Ozkaya, G, Wang, Y. Multi-armed bandit approach to portfolio choice problem. 2020. handle: http://hdl.handle.net/10230/46374

This citation was generated automatically.

Abstract

In this paper, we combine several Multi-Armed bandit algorithms with methodologies from finance literature and apply it to portfolio choice problem. Our results show that when we combine bandit algorithms with methodologies that take account of the non-normal distribution of returns, portfolio performance improves. Our results show that if contextual bandit algorithms applied to portfolio choice problem, given enough context information about the financial environment, they can consistently obtain higher Sharpe ratios compared to classical methodologies, which translates to fully automated portfolio allocation framework.