An Empirical Study On A Multi Factor Quantitative Investment Model

The financial market has the characteristics of high returns and high risks,so how to reduce risks as much as possible in the pursuit of high returns has become the goal pursued by institutions in recent years.Quantitative investment was born with this background.In recent years,there have been more and more domestic institutions and individuals using quantitative investment strategies,and their proportion in funds has been increasing.Traditional funds will also use quantitative methods to improve their fund efficiency.This article mainly introduces a three-factor quantitative investment model,which is based on the current comprehensive application of the common types of mainstream multi-factor models,and has made corresponding innovations in the factor selection process.The first chapter of this article mainly introduces the background of quantitative trading.History of development at home and abroad.The second chapter mainly introduces the theoretical basis and commonly used models of quantitative investment.The multi-factor strategy is a common model in the stock market quantitative strategy model.Its flexible combination is used by many institutions and is also widely studied in many academic circles.This article details the history of multi-factor model formation,the theory formation process,and the principles of each stage in more detail.In addition,the multi-factor model is also classified,which is mainly divided into regression model,scoring model and ranking model,and the model characteristics are introduced in more detail.The regression model is mainly used for prediction,while the scoring model focuses more on stock selection,The scoring model is more flexible,and it mainly explores factors that can increase returns,but does not care whether the factors strictly exist in theory.The third chapter mainly introduces the types of factors in the multi-factor model,systematically introduces the common factors of the multi-factor model,which can be divided into fundamental factors,statistical factors,public opinion big data category factors,and commonly used quantitative analysis The language used is introduced,and the pros and cons of each language are analyzed,which factors are selected,and the comparison between the factors.The fourth chapter mainly introduces the verification of the multi-factor model factors on the weight three factors,mainly to extract the individual Alpha factor for later modeling use,and at the same time verify the effectiveness of the Alpha factor.The three-factor method is a common factor in quantitative modeling.In this paper,it is regressed to obtain the Alpha value,and the preferred stock pool is sorted for Alpha.The method of periodic circulation is used to backtest the data of the last 4 years.,Observe the backtest results,evaluate the strategy,and judge the strategy from the perspectives of Sharpe ratio,annualized return rate,maximum retracement rate,winning rate,etc.In the fifth chapter,the abstracted emotional factors and quantitative energy factors are brought into the multi-factor model to test the effectiveness of the factors,including the effectiveness of single factor and the effectiveness of multi-factor combination.In order to maintain the stability of the investment portfolio,this article introduces stock index futures for hedging.The target is the Shanghai and Shenzhen 300 stock index futures,which can better hedge our stock pool and ensure the effectiveness of the hedging.Since the Shanghai and Shenzhen 300 stock index futures have 4 contracts,we usually choose the main contract when doing trading.The article uses a weighted average of the transaction prices of the 4 contracts to obtain a continuous stock index futures value,which is convenient for backtesting.It fits the actual transaction to the greatest extent.In order to deal with large-scale stock backtesting calculations,this paper builds a stock matrix based on a time series,forms a stock matrix by daily tradable stocks,and constructs a matrix of stock factors in the matrix to achieve multiple factors for a large number of calculations.It is processed at a very fast speed,and the backtest convenience of the model is extremely high.