Using Moving Window for Verifying Program Trading’s Robustness Profit and Risk to Design and Implement an Analyzing Platform.

• Main Authors: Cheng-Yu Huang, 黃晟宇
• Other Authors: 許智誠
• Format: Others
• Language: zh-TW
• Published: 2016
• Online Access: http://ndltd.ncl.edu.tw/handle/rnzczb

碩士 === 國立中央大學 === 資訊管理學系 === 105 === In the field of investment, the use of fundamentals or based on the results of the transaction program to invest is a very common behavior. Over the last few years, the public pay more attention to making investment through the transaction program because of FinTech. However, in the past investment-related areas, there is no complete verification process for the strategic robustness of the transaction program. In most of the literature, only discusses and focuses on the changes of the trading strategy parameters, the comparison of different technical indicators, the selection of different samples, and so on. Although these have some help in investment, it uses all the samples to do the analysis, whether it is similar in the future profit is a question. This research hopes to combine the moving window, moving average method and the evaluation method of moving window to establish a verification process of the profit robustness for evaluating transaction strategy. According to the process, this research will design and implement a platform for analysis. As a method for investor in using the program trading in the actual market before can verify trading strategy is good or bad. This research verifies if the use of the moving window feedback parameters is different, it has an impact on future performance between different trading strategies. Bollinger Bands and the double moving average trading strategies are also verified that the Bollinger Bands is better than double moving average strategy. Bollinger Bands test in Yuanta/P-shares Taiwan Top 50 ETF Index and Yuanta/P-shares Taiwan Mid-Cap 100 ETF of the 67 constituent stocks, show that 55% of the companies their Bollinger Bands are WFE qualified state, which represents the Bollinger Bands are profitable in future. At the end of this research, this process improves the entire program transactions in the moving windows verification process.