Dynamic graph reinforcement learning algorithm for portfolio management: A novel time–frequency correlated model

Cong Ma and Shijing Nan

Finance Research Letters, 2024, vol. 63, issue C

Abstract: Revealing the dynamic correlations among various assets is crucial for portfolio management. In this study, we build a novel Multi-graphs representation based on wavelet coherence to capture and learn their dynamic time–frequency correlations. Then, a novel portfolio management strategy is proposed by integrating the Multi-graphs representation with the deep reinforcement learning algorithm, referred to as the Dynamic Wavelet Coherence Graph Convolutional Reinforcement Learning (WCG-RL) algorithm. Several numerical experiments fully illustrate the performance of our proposed WCG-RL algorithm is applicable to stocks with different market capitalization, and its performance surpasses that of the state-of-the-art algorithms in the Chinese stock market.

Keywords: Wavelet coherence; Time–frequency relationship; Deep reinforcement learning; Portfolio management; Graph Convolutional Neural Networks (search for similar items in EconPapers)
Date: 2024
References: View references in EconPapers View complete reference list from CitEc
Citations:

Downloads: (external link)
http://www.sciencedirect.com/science/article/pii/S1544612324004033
Full text for ScienceDirect subscribers only

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:eee:finlet:v:63:y:2024:i:c:s1544612324004033

DOI: 10.1016/j.frl.2024.105373

Access Statistics for this article

Finance Research Letters is currently edited by R. Gençay

More articles in Finance Research Letters from Elsevier
Bibliographic data for series maintained by Catherine Liu ().