The Design And Implementation Of Stock Backtesting And Computing System Based On Storm

With the economic soaring of China, more and more people are involved in the stock market. The investors, who have their own investing philosophy, wish to be able to predict their investment strategy performances in the future. Although the future is unknowable, but China’s stock market bas been opened for about 30 years. The performance of investment strategy in the history can provide data support now.A stock backtesting and computing system allows investors to specify the query sentence and strategy information, then simulate the stock strade process and calculate the income situation within the time interval of backtesting to guide investors’ investments. This system is the first one to provide such features of the product in the industry, and has features like consumption of resources, high reliability, and fast calculation.The system is based on the emerging distributed computing framework Storm. It takes the advantage of parallel computing, and divides complex computing tasks into multiple subtasks to improve computing performance. MongoDB is the main data storage carrier and MyBatis is the database persistence layer framework. Spring framework is also used in this system. This paper analyzes the project background and technical background of the system, focus on the requirements analysis and system design and elaborates implementation details of important module.In testing case, the system can finish any complex calculaton in 500ms. Currently, the system has been successfullyd deployed and running in the actual environment. It works very well and stable.