Font Size: a A A

Sentiment-based Consumption Capital Asset Pricing Model

Posted on:2014-01-28Degree:DoctorType:Dissertation
Country:ChinaCandidate:R G ZhangFull Text:PDF
GTID:1229330401960196Subject:Financial engineering and economic development
Abstract/Summary:PDF Full Text Request
Traditional financial theory suggests that rational arbitrage can cancel irrational behaviorout, and brings prices closer to fundamentals. It gives no role to investor sentiment. However,there are lots of abnormalies of investors’ behavior since1970s, and traditiaonal financialtheory cannot offer a satisfied explanation. More and more empirical researches and financialexperiments present evidences that investor sentiment plays a systematic role in asset pricing.Hence, it would be important to incorportating investor sentiment in asset pricing model, andhelp to improve our understanding of the behaviour of asset prices.According to the empirical results, we find that investor sentiment has an important impacton stock returns. We do researches on the influence mechanism of investor sentiment throughCCAPM and numerical simulations, and our results offer a partial explanation of somefinancial anomalies in the stock market: pricing bubble, high volatility, and survival.First, according to B-W method we set up market sentiment indexes at different frequencies,and examine market sentiment and return comovements. We consider six separate proxies,such as new fund accounts, new stock accounts, ShangZheng volume, and employ principalcomponent analysis to construct a composite market sentiment. OLS models are used fortesting the aggregate effect of market sentiment on stock returns, and panel data models areused for testing the individual effect. The empirical results show that the aggregate effect andindividual effect of market sentiment are statistically significant. Moreover, the term structurecharacter of the market sentiment is a monotonous decreasing function of the time term.Mixed-frequency market sentiment is more important than the low-frequency one, and can bemore important than the market premium.Second, according to B-W method we set up individual stock sentiment indexes at differentfrequencies, and examine individual stock sentiment and return comovements. We considerfour separate proxies: the buy-sell imbalance, adjusting turnover rate, the growth of amount,prior return, and employ principal component analysis to construct a composite individualstock sentiment. The effects of individual stock sentiment on the stock returns are significantat different frequencies. Moreover, we find that individual stock sentiment is almost asimportant as market premium. However, mixed-frequency individual stock sentiment is more important than the market premium. The empirical results also show that the term structurecharacter of individual stock sentiment is a monotonous decreasing function of the time term.Third, we present an asset pricing model by incorporating investor sentiment. Thesentiment equilibrium price could be decomposed to the rational term and the sentiment term,and the investor sentiment has a systematic and significant impact on the risky asset price. Inthe model with heterogeneous sentiment investors, the sentiment term has a wealth-weightedaverage structure and the investor’s wealth proportion could amplify the sentiment shock onthe asset price. The model could offer a partial explanation of some financial anomalies in thestock market: the phenomenon of savings transfer to the stock market, pricing bubble andhigh volatility.Fouth, to illustrate the impact from time varying sentiment the equilibrium, based oninvestor sentiment, this paper presents a dynamic asset pricing model. The model shows thatInvestor sentiment has a systematic and significant impact on the equilibrium, and timevarying sentiments can lead to various price changes. Moreover, the equilibrium stock price isthe wealth-share-weighted average of the stock prices that would prevail in an economy withone sentiment investor only. Finally, investor sentiment induces fluctuations in the wealthdistribution, which induces mean reversion in stock returns. The model offers a partialexplanation for the financial anomalies of mean reversion.Fifth, to illustrate time varying sentiment and the survival of sentiment investor, based oninvestor sentiment, this paper presents a continuous asset pricing model. The model showsthat Investor sentiment has a systematic and significant impact on the equilibrium. Moreover,consistent with the high volatility in the reality stock market, time varying sentiments can leadto various price changes. The numerical examples show that sentiment investors can surviveafter100years. The wealth of sentiment investors, whose sentiment is close to rationalsentiment and relative risk aversion coefficient is high, is slightly smaller than rationalinvestors. The model offers a partial explanation for the financial anomalies of survival,shows that investor sentiment has a significant and long-lasting impact on asset prices.Sixth, based on bounded rationality, we present a static asset pricing model. The modelshows that investor sentiment has no impact on stock price when investor sentiment is lowerthan a threshold. If investor sentiment is higer than the threshold, bounded rationality investors could change investment strategies, and then stock price is affected by investorsentiment. Based on investor sentiment, our model offers a partial explanation for thedecision-making hysteresis.Finally, these results show that investor sentiment has a significant impact on asset prices.Hence, incorporating investor sentiment into models and empirical researches might help toimprove our understanding of the behavior of asset prices.
Keywords/Search Tags:Asset Pricing, Mixed-frequency Sentiment, Time Varying Sentiment, Heterogeneous Sentiment, Bounded Rationality
PDF Full Text Request
Related items