| dc.description.abstract | In recent years, fintech has risen rapidly. In the financial investment, with the popularization of personal computers and the Internet, it is easier to obtain financial data, and investment strategies based on data and algorithms are becoming more and more popular.
This means that investors no longer rely on professional financial institutions to provide investment advice, but choose their favorite algorithmic trading strategies for algorithmic trading, and can design their own trading strategies through programs or algorithmic trading development platforms. Thus, they become empowered investors.
However, each strategy model has different design reasons and performance evaluation metrics. It is difficult for investors to understand the meaning and effects of model parameters, and it is also difficult for investors to compare different models and choose investment strategies that meet their preferences.
In response to this problem, this study will design a meta-process model that integrates different trading strategies based on the back-testing process of algorithmic trading. Through this meta-process model, investors will be able to align the performance of different trading strategies, so that different trading strategies is comparable.
At the same time, this research will also use this meta-process model to design a visual performance comparison platform. This platform can import different trading strategy models, adjust model parameters through web forms, then call the model for back-testing, and report the results of the back-testing. The performance of different models is finally displayed using interactive chart overlays such as line charts and bar charts, thus investors will be able to easily analyze the performance of different models.
This research also provides pivot analysis and a filter for parameters and performance metrices. Furthermore, it provides the suggested analysis process of the model. Investors can have a better understanding of the parameters and performance of the model from the process, and will be able to find out what model suits their preferences from this platform.
In the future, this research platform can expand more trading models, and add more performance evaluation metrics. Comparing trading models from different perspective, so that investors can quickly and accurately find their preferred strategy. | en_US |