A Globally Optimal Portfolio for m-Sparse Sharpe Ratio Maximization

Authors: Yizun Lin, Zhao-Rong Lai, Cheng Li

NeurIPS 2024 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Extensive experiments with real-world financial data sets are conducted to evaluate the performance of the proposed m SSRM-PGA.
Researcher Affiliation Academia Yizun Lin1 Zhao-Rong Lai1 Cheng Li1 1Department of Mathematics College of Information Science and Technology Jinan University, Guangzhou, China
Pseudocode Yes Algorithm A1 m SSRM-PGA
Open Source Code Yes The codes of m SSRM-PGA are accessible via the link: https://github.com/linyizun2024/m SSRM/tree/main/Codes_for_Experiments_in_Paper.
Open Datasets Yes These data sets are collected from the baseline and commonly-used Kenneth R. French s Real-world Data Library.2 2http://mba.tuck.dartmouth.edu/pages/faculty/ken.french/data_library.html
Dataset Splits No The paper describes using a "moving-window trading framework" with time window sizes T=60 and T=120 for model updating, which serves as a rolling training period. However, it does not specify a distinct validation split (e.g., fixed percentage or sample count) separate from the training and testing phases.
Hardware Specification No The paper does not provide specific details about the hardware (e.g., CPU, GPU models, memory) used for running the experiments.
Software Dependencies No The paper does not provide specific version numbers for any software dependencies or libraries used in the implementation or experiments (e.g., Python, PyTorch, specific solvers).
Experiment Setup Yes As for m SSRM-PGA, we examine three levels of sparsity m = 10, m = 15, m = 20 and set ϵ = 10 3. The setting of other parameters are presented in Appendix A.9. ... Initialization: Set v(0) = p, tol = 10 5, Max Iter = 104