Distributionally robust optimal allocation of financial assets under the uncertainty and irrationality

The essence of portfolio selection is to allocate funds to different financial products reasonably to achieve the goal of risk avoidance and asset accretion. However, the inherent uncertainty of financial markets, coupled with the irrational behavior of investors, complicates the development of an efficient portfolio strategy. In view of this, we propose a distributionally robust optimal allocation strategy for financial assets that accounts for both uncertainty and irrationality. The proposed model used the Wasserstein-based distributionally robust optimization to deal with the financial market uncertainty, while the constructed smooth and S-shaped utility function is utilized to portray investor’s irrationality. Additionally, we also provide methods for determining the hyperparameters of the proposed model and reformulate the proposed model into a tractable problem. The effectiveness of the proposed model is verified with almost all S&P 500 components. Experimental results show that, relative to other benchmarks, the proposed model improves investment returns while eliminates risks. In particular, the proposed model is better than a portfolio model that only considers market uncertainty or investor irrationality. In summary, the proposed model simulates the investment decision-making behavior of investors in real life by simultaneously considering the financial market uncertainty and investor irrationality. Benefiting from this design, the proposed model is promising in increasing investment returns and eliminating risks.