Research On Quantitative Investment Strategy Of Sector Rotation Based On Machine Learning

In recent years,automatic trading based on the quantitative algorithm and the artificial intelligence has become an efficient trading method with the rapid development of fintech.The quantitative trading in China is still in its infancy compared with the western mature capital market,it provides favorable conditions for quantitative trading according to fewer institutional investors and less efficient capital market.So this article use the AI machine learning algorithm to construct industry choice strategy,in the hope to get a steady benefits beyond the market average level.Machine learning has been used in many fields such as figure recognition,natural language processing on account of its excellent capacity in dealing with fuzzy nonlinear data.Capital market is a complex nonlinear system with a low signal-to-noise ratio,so we apply the machine learning to quantitative investment field to assist us to explore those unknown nonlinear characteristic in capital market,and to help investors to make better investment decisions.As a part of asset allocation,Industry strategy is of great significance for investors to make investment decisions.Firstly,this paper researches the basic characteristics of industry changes and analyzes its differences with the stock market.Through four categories of factor research tests,we selected seven factors with relatively effective and low correlation as alternative input variables.Then,through theoretical analysis and empirical test,this paper constructs three relatively effective supervised machine learning models: support vector regression,fully connected neural network and multi-factor model,we find that all of them can well explain the industry return rate.Finally,in order to obtain more excess returns with more information,we integrated the three sub-models into a complex and effective machine learning model.Through performance analysis,the integrated model shows a better performance than the sub-model in terms of both profit and risk.This paper,on the one hand,studied the basic characteristics of the industry changes,and built an effective factor library,so that investors can understand the capital market better from the perspective of technology and fundamentals.On the other hand,in the process of constructing the strategy,this paper used the machine learning algorithm to show the excellent results through performance analysis,which provides guidance for the investors.