Research On Stock Market Trend Tracking Algorithm Based On MACDHL And Sentiment Analysis

Focusing on economic construction is the key to rejuvenating the country and the foundation of building a nation.The operation of the economic system and the market conditions of all walks of life across the country can be reflected through the barometer of the stock market.Therefore,more and more people have begun to study the stock market,hoping to predict the market conditions of companies and enterprises in advance,and carry out corresponding operations to obtain profits or stop losses in time.However,there are many factors that affect the stock market,including the overall economy,industry competition,stockholder sentiment,etc.Therefore,how to find a suitable method for predicting the stock market has been a hot topic of research by experts and scholars in recent years.This thesis mainly considers the stock market situation from the technical and fundamental aspects of the stock market and puts forward a new trend tracking algorithm to predict the timing of stock trading.Firstly,in the technical aspect of the stock market,the thesis studies the financial indicators of the stock market,and according to the correlation and limitations of the Moving Average Convergent Divergent Histogram Line(MACDHL)and the stock price trend,a trend tracking algorithm based on MACDHL is proposed.This algorithm establishes a model to determine the buying and selling time by judging the weak rising and falling trends of stocks,the rising and plunging trends as well as the situation near the MACDHL center line,and explains the reasons for the improvement of the method and the influence brought by each step of the improvement.This algorithm uses MACDHL index to comprehensively consider the short-term and long-term stock market,improves the method of determining the trend of stock boom or slump to improve the volatility of the stock,and improves the value range of the threshold corresponding to the center line to reduce losses.Secondly,in the fundamentals of the stock market,this thesis studies the financial news related to stock companies,and proposes a trend tracking algorithm based on sentiment analysis according to the correlation and limitations of news sentiment analysis results and news content.The algorithm established a model to determine the buying and selling time by comprehensively considering the sentiment tendency,classification tendency and correlation of stock news in a certain period of time.The reasons for the improvement of the method and the impact of each improvement of the method were described step by step.This algorithm improves the incomplete problem of considering news in the timing of stock buying and selling and makes a comprehensive analysis of related news from multiple dimensions.Finally,considering the technical and fundamental aspects of the stock market,combined with the specific training results of the above two models,this thesis finally proposes a trend tracking algorithm based on MACDHL and sentiment analysis.The algorithm not only considers the weak rising and falling trend,surging and plunging trend of stocks in the technical surface and the situation near the MACDHL center line,but also considers the sentiment tendency,classification tendency and correlation of stock news in a certain period of time in the fundamentals.The experimental analysis shows that compared with the previous static trend tracking algorithm,adaptive trend tracking algorithm and trend tracking algorithm based on moving average convergence and divergence,the success rate of profit and accumulated return on investment has achieved better returns,and has better practical application value.