| dc.description.abstract | In recent years, fintech has risen rapidly. In the financial investment, we will often see the Robo-Advisor. The Robo-Advisor will recommend an investment portfolio based on investors′ preferences. In the investment portfolio, the selection and allocation of investment components are the keys to determine the excellent and stable investment portfolio. Modern portfolio theory is a common method of portfolio weight allocation, but this method ignores the concept of a hierarchical structure of commodities, resulting in a portfolio that is very unstable when verified outside the sample. Most of the general program trading research is mostly for the parameters of a single trading strategy optimization, or optimization of the entire sample of the investment portfolio, so it is very complicated to properly verify the composition of the investment portfolio.
To response this problem, this study combines the portfolio with the moving window method and is committed to building a portfolio risk assessment tool that allows investors to import the daily profit performance files of the investment through this tool. The investors can set performance ranking methods, weight allocation model and other parameters. The research tool will use the moving window method to select investment components and allocate weights inside the sample, and then simulate investment and rebalance the components and weights of investment portfolio outside the sample. Automatically calculate the performance of different window configurations, and present the window performance through a heat map to find out the window of the best performance and the plateau parameters of stable performance.
In this study, we have set up six verification cases. We want to use this research tool to achieve the purpose of each verification case and recommend the investment portfolio parameters of individual verification cases. Through the verification results, we can know that each case can observe the plateau type and maximum plateau area through the performance heat map obtained by this research tool, and recommend the most suitable window size setting and stable portfolio plateau parameters.
In the future, this research tool can add more performance ranking factors, weight allocation models, and use reinvestment method to fit the verification to the real investment situation, so that the tool can effectively assist investors to build the most suitable investment portfolio. | en_US |