Stock Investment Decision Model Based On Analyst Recommendations Composition

Analysts play a key role in information gathering,information processing,andinformation dissemination in the stock market.Their published recommendation reports are important information resources for investors to make investment decisions.The recommendation reports issued by analysts are important information resources for investors to make investment decisions.Due to the large differences in individual characteristics of analysts,investors lack the effective analysis method to extract the valuable information from analyst recommendations for making their investment decisions.To this end,this paper proposes an stock investment decision model based on analyst recommendations composition,which focuses on the perspective of group effect and could assist investors in scientific decision-making.The main researches are as follows:Firstly,by exploring the corresponding real stock return sequence of analyst recommendations,we verify the utility of analyst recommendation.We find that positive analyst recommendations are accompanied by positive return sequence,whereas negative analyst recommendations are accompanied by negative return sequence.The higher the stock rating,the better the trend of the stock price,which proves the existence of effective investment information in the analyst recommendations.Secondly,we extract the key features of the cumulative abnormal return series and propose a performance measurement algorithm based on the generalized area of the yield curve.The performance index is calculated by applying the algorithm to the return series and unifies the performance measurement of the cumulative abnormal return series,which solves the problem that the analyst recommendations are difficult to unify quantitative evaluate and compare.Furthermore,we argue that the stock rating and the change of the rating in the recommendation report are the key factors affecting the performance index.But,it ' s not reliable to direct use the basic contents of the analyst recommendation for the prediction of stock returns.Finally,stock investment decision model based on analyst recommendations composition is proposed.We use bipartite networks to quantify the many-to-many connection between stocks and analysts and the concept of analyst recommendations composition of the recommended stock is defined.The analysts profile is constructed to quantify the individualcharacteristics of the analysts and the composition analysis is executed.Based on these above procedures,the investment value of stock is identified and the efficiency of investors' decision-making is improved.This paper promotes research in the area of analyst recommendation and provides new ideas and new methods for investors to use large scale data of analyst recommendations for investment decision-making.