Research On Online Investment Strategy Based On Kalman Filter And Its Empirical Application

After 27 years of rapid development,China's A-share market has grown into the world's second largest stock market with more than 3,500 listed companies.With the development of the Chinese stock market and the improvement of the quality of investors,traditional investment strategies have increasingly failed to meet the needs of investors.In this situation,many foreign investment theories and investment strategies have also been introduced into the Chinese market.However,due to the huge differences in the Chinese and foreign stock markets,many investment strategies that are effective abroad have not performed well in China.In this paper,the Anticor algorithm based on the stock reversal effect proposed by Canadian scholar Allan Borodin and the K-AC-M algorithm proposed by South African scholar Raphael Nkomo which introduces the Kalman filter and the momentum effect based on the Anticor algorithm,have been deeply stutied.In the empirical part,a multi-dimensional analysis of the actual application of the K-AC-M algorithm in the Chinese stock market has been conducted.The K-AC-M algorithm strategy was implemented using MATLAB software,and the effects of investment window factors such as investment window period,portfolio stock capacity,number of shares outstanding,industry classification,and overall market trend were compared and analyzed.Finally,it is concluded that different investment portfolios have an optimal window period;income and risk resistance capacity increase with increasing stock capacity;K-AC-M algorithm for large-cap stocks,small-cap stocks and index stocks and ST stocks are effective;algorithms can amplify stock returns in bull market,and bear market will amplify stock losses.Finally,according to the characteristic of K-AC-M algorithm with amplification effect,the trend judgment parameters are introduced to improve the algorithm,and the T_MA-KACM and T_MACD-KACM algorithms are obtained.The test results show that the new algorithm can effectively reduce the loss of K-AC-M algorithm and improve the anti-risk ability of the algorithm.