Abstract:In recent years, deep or reinforcement learning approaches have been applied to optimise investment portfolios through learning the spatial and temporal information under the dynamic financial market. Yet in most cases, the existing approaches may produce biased trading signals based on the conventional price data due to a lot of market noises, which possibly fails to balance the investment returns and risks. Accordingly, a multi-agent and self-adaptive portfolio optimisation framework integrated with attention mechanisms and time series, namely the MASAAT, is proposed in this work in which multiple trading agents are created to observe and analyse the price series and directional change data that recognises the significant changes of asset prices at different levels of granularity for enhancing the signal-to-noise ratio of price series. Afterwards, by reconstructing the tokens of financial data in a sequence, the attention-based cross-sectional analysis module and temporal analysis module of each agent can effectively capture the correlations between assets and the dependencies between time points. Besides, a portfolio generator is integrated into the proposed framework to fuse the spatial-temporal information and then summarise the portfolios suggested by all trading agents to produce a newly ensemble portfolio for reducing biased trading actions and balancing the overall returns and risks. The experimental results clearly demonstrate that the MASAAT framework achieves impressive enhancement when compared with many well-known portfolio optimsation approaches on three challenging data sets of DJIA, S&P 500 and CSI 300. More importantly, our proposal has potential strengths in many possible applications for future study.

Ensemble learning for portfolio valuation and risk management

Developing An Attention-Based Ensemble Learning Framework for Financial Portfolio Optimisation

SABER: Stochastic-Aware Bootstrap Ensemble Ranking for portfolio management

Online Learning of Portfolio Ensembles with Sector Exposure Regularization

Risk-averse reinforcement learning for portfolio optimization

Portfolio Optimization Managing Value at Risk under Heavy Tail Return, using Stochastic Maximum Principle

Uncertainty-Aware Reinforcement Learning for Portfolio Optimization

A Deep Reinforcement Learning Framework For Financial Portfolio Management

Statistical Learning of Value-at-Risk and Expected Shortfall

Option Dynamic Hedging Using Reinforcement Learning

Efficient Nested Estimation of CoVaR: A Decoupled Approach

Pricing Bermudan options using regression trees/random forests

Portfolio Construction Based on Stochastic Dominance and Empirical Likelihood

Sub-sampling and other considerations for efficient risk estimation in large portfolios

Estimation of Large Financial Covariances: A Cross-Validation Approach

Learning short-option valuation in the presence of rare events

A new deep reinforcement learning model for dynamic portfolio optimization

Deeply Learning Derivatives

A Two-Level Reinforcement Learning Algorithm for Ambiguous Mean-Variance Portfolio Selection Problem

Portfolio optimization using local linear regression ensembles in RapidMiner

Deep learning for CVA computations of large portfolios of financial derivatives