Research On Cross-sectional Return Predictability And Asset Pricing Based On Technical Trend Information

After more than 30 years of development and improvement,the Chinese stock market has initially established a multi-level market system,which plays an increasingly important role in promoting reform and development and optimizing resource allocation.However,the market is a dynamic and open system,in which various imperfect factors,such as market friction and investor behavior bias,interact with each other and evolve continuously under the constantly changing market conditions,so that information trends play an important role in price formation.Based on this,this dissertation systematically explores the prediction effect of technical trend information on stock cross-sectional return（i.e.technical trend effect）,analyzes the generation mechanism of technical trend effect,and optimizes the portfolio strategy based on trend information by exploring the impact of financial analysts and fundamental trend on the return predictability of technical trend information,and builds an asset pricing model including trend factors.The specific research contents and conclusions are as follows:Firstly,this dissertation examines the prediction effect of technical trend information on stock cross-sectional return from different investment periods.Firstly,by eliminating the common noise components in technical indicators,a new technical trading index,TECH^IWC,is constructed based on the technical trend information and an iterated weighted combination approach.We find that this index significantly predicts future stock returns from short-term to long-term,and this predictive ability cannot be explained by firm size,book-to-market ratio,liquidity and other firm characteristics,and market conditions.Secondly,based on the limited arbitrage theory,this dissertation explains the source and the reason for the persistence of the TECH^IWC effect from a new perspective,idiosyncratic volatility.We find that there is a significant positive correlation between the TECH^IWC effect and idiosyncratic volatility,which indicates that idiosyncratic volatility is an important arbitrage holding cost.Secondly,based on trading volume trend information,this dissertation examines the role of moving averages of trading volume on asset pricing.Firstly,this dissertation proposes a new technical index that can reflect the trend information of trading volume,that is,the moving average distance index based on trading volume（MAVD）.We find that MAVD has a significant negative predictive effect on stock cross-sectional return,and this predictive ability is not affected by other firm characteristics,well-known risk factors,market timing,and volume shock.Secondly,MAVD is better than MAVD（the moving average distance index based on price）in prediction and pricing power,which emphasizes the importance of trading volume trend information for the Chinese stock market.Finally,based on behavioral finance theory,we explain the MAVD effect from the perspective of mispricing,and find that the MAVD effect is related to limits to arbitrage,investor attention,investor overconfidence,investor sentiment,and speculative trading behavior.Thirdly,this dissertation explores the impact of financial analysts on the predictive performance of technical trend information on return（i.e.technical trend effect）.Firstly,we construct a comprehensive price trend index and comprehensive trading volume trend index,and explore the impact of analyst activities on technical trend effect.We find that the higher the analyst coverage,the lower the degree of mispricing,and the weaker the price or trading volume trend effect.Secondly,considering the information’s updates from analysts,we discuss the impact of analyst recommendation revision on technical trend effect.We find that compared with stocks with no revision,upgraded and downgraded stocks have specific information updates,and the price or trading volume trend effect is weaker for stocks with upgrade revision and downgrade revision.Thirdly,the empirical evidence provided in this dissertation shows that there are deviations when analysts make recommendation on stocks,and they do not take technical trend information into account.Finally,by constructing a joint information strategy based on technical trend information and analyst recommendation,we find that analyst recommendation（technical trend information）can be used as supplementary information of technical trend information（analyst recommendation）to improve the predictive ability of future stock returns.Fourthly,this dissertation analyzes the impact of fundamental trend information on the predictive performance of technical trend information on return,and constructs an asset pricing model based on trend information.Firstly,we construct the comprehensive fundamental trend information index and comprehensive technical trend information index,and prove that the two types of comprehensive trend information have significant predictive ability on return.Secondly,the prediction effect of joint trend information is tested.We find that the combination of technical trend information and fundamental trend information has better prediction performance.Finally,we explore the pricing power of technical and fundamental trend factors,and prove that the pricing model with trend factors has better pricing performance than the classical asset pricing model.This shows that trend factors are important pricing factors in the Chinese stock market.The main contributions of this dissertation are as follows:firstly,the existing research on technical trend information mostly focuses on its time-series return prediction of stocks,and the research on stock cross-section return prediction is insufficient.On this basis,we systematically explore the predictive effect of technical trend information on stock cross-sectional returns from different investment periods by constructing a better technical trading index,examine the existence and persistence of this trend effect,and explain the source and the reason of the persistence of this effect from a new perspective（i.e.idiosyncratic volatility）.Secondly,we are the first to propose the moving average trading volume distance index（MAVD）,confirm the incremental prediction effect of trading volume trend compared with price trend,explain the MAVD effect from multiple perspectives of investor behavior bias,and test the prediction performance of MAVD in different volatility portfolios for the first time.Thirdly,considering the price trend effect and trading volume trend effect at the same time,we examine the impact of financial analysts on the prediction performance of technical trend information on return.It is the first empirical study that analysts fail to incorporate technical trend information into their recommendation decisions.Fourth,we combine technical trend information and fundamental trend information to predict stock returns,introduce them into the portfolio management process at the same time,construct the joint trend information strategy.This makes up for the shortcomings of the previous strategies,and provides theoretical basis and practical reference for investors to carry out portfolio management and asset allocation in the dynamic market.On this basis,we construct the trend factors reflecting the short-term,medium-term and long-term technical and fundamental information,which improves the pricing power of the existing asset pricing models.