Application Of American Stock Similarity Factor In Multivariate Model

With the continuous progress of science and technology and the continuous development of social productivity,the world economy is booming,while the degree of globalization is increasing,which promotes exchanges and cooperation among people of all countries and improves the efficiency of capital allocation.In this context,as the two largest economies in the world,exchanges between China and the United States are increasing,and economic ties are also strengthening.In recent years,the linkage between Chinese and American stock markets has been increasing.The change of US economic data and the trend of US stock market have a significant impact on China's stock market.Therefore,it is of practical significance to explore the internal relationship between Chinese and American stock markets and provide guidance for Chinese stock market investment.The research content of this paper is to use Python program to simulate trading.Through cross-market pairing trading and constructing similarity factor of American stock,we try to study pairing trading in A-share market,reduce the influence of short limit on pairing trading,realize statistical arbitrage idea in Chinese stock market,and improve the market performance and feasibility of pairing trading strategy.The sample data of empirical analysis are the closing prices of stocks listed on A-share market,New York Stock Exchange and Nasdaq Stock Exchange in the past three years.The core idea of quantitative trading is to replace subjective judgments with mathematical models and develop a set of trading signal system with a positive expectation by digging into historical data.With the popularization and development of quantitative trading,there are more and more quantitative strategies in the market,among which the typical strategies are paired trading and multi-factor stock selection strategy.The former is based on statistical arbitrage,which takes advantage of the short price deviation of two assets and hedged the risk to obtain the Alpha return of the two assets.Core assumption is that the spread of paired assets has the mean value recovery.The latter combines multiple factor strategies with logical background,selecting stocks with high comprehensive scores on each factor to construct portfolios,whose core is how to mine factors with logical background.Due to the high degree of economic correlation between China and the United States,there is interactivity between the stock markets of the two countries.Constructs US equities-A share paired factor according to the ideas of the paired trading,and then uses the factor to select stocks.Empirical analysis reveals that the factor shows better degree of differentiation for A share and lower correlation with traditional stock selection factors.Therefore,the factor can be used in the multi-factor model for stock selection.The contribution is to put forward the factor which has not been put forward in the previous research,and the validity of this factor is proved by the empirical research.