A Study Of The Sentiment Asset Pricing With Information

The effect of economic cycle on the stock market cycle is intrinsic, long-term andfundamental, but the two cycles are not completely synchronized, even often presentsystematic divergence. In addition, China’s stock market cycle shows the characteristics of"ramp up sharp down". These realities seem to suggest that the non-fundamental factors playa significant role in the stock prices. Therefore, based on the concept of sentiment acceleratorof Las Tvede (2008), we explore the sentiment accelerator effect in China’s stock market. Themain characteristics of sentiment accelerator mechanism are systematic impact of investorsentiment on stock prices, and asymmetric effect of investor sentiment on stock prices.In fact, the main cause of the securities market to its own special phenomena is thefinancial asset pricing anomalies. Thus, based on the perspective of the behavioral financialasset pricing theory and combined with the main characteristics of sentiment acceleratormechanism, this paper constructs the sentiment asset pricing model with information.Furthermore, it gives the corresponding empirical tests. The main contents of this paperinclude the following six aspects:First, we present the static sentiment asset pricing model with information according tothe main characteristics of sentiment accelerator mechanism. The model shows that investorsentiment has an important systematic impact on asset prices, and has an asymmetric effect onthe asset pricing. In the model, sentiment is contrasted with information. The rational investortrades on information in the usual way, while the sentiment investor trades on his ownsentiment as if it were information. Therefore, by focusing on the interaction between therational investor and the uninformed sentiment investor, our model shows how this interactioncould sustain incorrect prices. Ultimately, we demonstrate how the financial asset is pricedwhen sentiment investors learn from prices.Second, we extend the static economic environment to the dynamic economyenvironment, and then set up a two-period trading sentiment asset pricing model withinformation. In addition, we expand the two-period trading equilibrium to the multi-periodtrading sentiment asset pricing model with information. We obtain the conclusions similar tothat of the static setting, namely, the investor sentiment is an important factor that affectsasset prices, and the effect on the equilibrium price is different under different market states.We give an analytical solution to the sentiment equilibrium price in the discrete tradingequilibrium, which could be decomposed to three terms. The first term is the previous pricewhich is the anchoring value of the current price. The second term is the rational investor’s insufficient adjustment for the current price. The third term is the sentiment investor’sadjustment for the current price. By solving the model and numerical simulation, we offer apartial explanation to some financial anomalies such as short-term momentum, long-runreversal and so on. Moreover, we describe a dynamic price path anchoring to the initial price.Third, in order to further dicuss the impacts of information and sentiment on the assetprices, we extend the discrete trading equilibrium to the continuous trading sentiment assetpricing model with information. In the continuous trading equilibrium, the sensitivitycoefficient of sentiment is a constant, while it gets smaller and smaller over time in thediscrete trading equilibrium, which shows that the speed of sentiment incorporated into pricesgets slower and slower. Moreover, in the continuous trading equilibrium, the information isgradually incorporated into prices, and all the information is incorporated into prices by theend of trading, while in the discrete trading equilibrium, the information parameter becomessmaller and smaller over time, which shows that the information is incorporated into pricesgradually, and the parameter becomes an arbitrarily small positive number by the end oftrading.Fourth, we mainly study the effect of individual stock sentiment on individual stockprices, and systematically comb its micro-foundation and the conduction mechanism. Weconstruct individual stock sentiment index, and use panel data model with dummy variables tomake empirical analyses that individual stock sentiment has a systematic and asymmetricimpacts on the stock prices. First of all, we use Fama-French three-factor model toempirically testify the scale effect and the book-market effect in China’s stock market. Then,we illustrate the cross-sectional differences of stock returns from the angle of investorsentiment. Further, combining with difference impact of investor sentiment on the stockreturns at different time intervals, we examine the double asymmetric effect of sentimentimpact in China’s stock market. The individual stock sentiment has a bigger impact on stockprices of small caps in the downlink period of the stock market because of informationasymmetry in credit market, external financing costs and other factors.Fifth, we construct the market sentiment index, and analyze the impact of marketsentiment on the stock index prices. We make use of MS-VAR model to quantitativelyanalyze the periodicity of the stock market, sentiment-price spiral effect and asymmetriceffect under different market situation. In a downward phase of the stock market, monetarypolicy changes, especially investor sentiment changes lead to larger and longer thefluctuations of the stock market. We could directly and clearly see the existence ofasymmetric effect of the stock market from the impulse response functions. Moreover, the response of the stock price to the shocks of investor sentiment shows immediate effect, whilethe response to the shocks of monetary policy variables exists lagging effect. Ultimately, weexplain the mystery of deviation of stock market cycle and the economic cycle and thecharacteristics of "ramp up sharp down" in China’s stock market.Sixth, we mainly analyze how the extreme sentiment affects the stock prices. Applyingthe non-parametric model, the empirical study finds that the correlation between the moderatechange of investor sentiment and the stock index return is positive, which shows an obviousmomentum effect. However, the stock index return presents significant reversal effect if thechange of investor sentiment is dramatic. At this time, the proportion of rational traders ismuch larger, and the rational investor arbitrage space appears, forcing prices of financial assetto the fundamental value. The empirical finds offer a partial explanation to financialanomalies such as the mean reversion of stock index return, the characteristics of "ramp upsharp down" in China’s stock market and so on.