This archive is distributed in association with the INFORMS Journal on Computing under the MIT License.
This repository contains supporting material for the paper An efficient scenario reduction method for problems with higher moment coherent risk measures by Xiaolei He and Weiguo Zhang.
The software and data in this repository are a snapshot of the software and data that were used in the research reported in the paper.
To cite the contents of this repository, please cite both the paper and this repo, using their respective DOIs.
https://doi.org/10.1287/ijoc.2022.0375
https://doi.org/10.1287/ijoc.2022.0375.cd
Below is the BibTex for citing this snapshot of the respoitory.
@misc{HeZhang2024,
author = {Xiaolei He and Weiguo Zhang},
publisher = {INFORMS Journal on Computing},
title = {{An efficient scenario reduction method for problems with higher moment coherent risk measures}},
year = {2024},
doi = {10.1287/ijoc.2022.0375.cd},
url = {https://github.com/INFORMSJoC/2022.0375},
note = {Available for download at https://github.com/INFORMSJoC/2022.0375},
}
The goal of this software is to demonstrate the effectiveness of the scenario reduction method, as proposed in the paper, comparing to other reduction methods on both a simple portfolio optimization problem and a realistic one.
In the folder data, stock_index1 represents the weekly price of nine stock indices from different countries, downloaded from https://finance.yahoo.com/. The horizontal heading of the table displays the stock index names.
stockprice represents the weekly price of fifty stocks selected randomly from the US S&P 500 index, downloaded from https://finance.yahoo.com/. The horizontal heading of the table displays the symbol of each stock in the website https://finance.yahoo.com/.
The folder simple-samples and realistic-samples contain the 20 original scenario sets generated based on the stock_index1 and stockprice respectively.
The folder contains the Matlab implementations of the numerical experiments. The codes are split into the following two folders:
-
"simple-portfolio" contains codes for numerical experiments on the simple portfolio optimization problem
-
"realistic-portfolio" contains codes for numerical experiments on the realistic portfolio optimization problem.
The code file rcode.R in folders simple-portfolio and realistic-portfolio is to generate the original scenario set using the build-in functions in "rmgarch" package in R software.
The results are presented in the numerical experiments section of the paper.
To replicate the results in Tables 1-5 and Figures 2-3, put the data folder simple-samples and the code folder simple-portfolio under the same folder. Then, execute the main and result_descrip code files sequentially.
To replicate the results in Tables 6-10 and Figures 4-5, put the data folder realistic-samples and the code folder realistic-portfolio under the same folder. Then, execute the main and result_descrip code files sequentially.