The Effect Of Expert Opinions On Price Bubbles:Experimental And Empirical Study

Since the1960s, the new emerging experimental economics introduced laboratory methods into economics, providing a new method for studying the economies. Experimental economics test the traditional economic theories and the policy implementation effect in the controllable laboratory situations. Compared with empirical method, experimental method has its own advantages and is an effective supplement to empirical method when studying and explaining the operation of economy. Through experiment by simulating the real market in the laboratory, economists can study and explain the operation mechanism of economy, the decision-making process of people and the mechanism which would improve the market operating efficiency. Especially after Vernon Smith discovered bubbles in the laboratory in the1980s, experimental research of asset bubbles has been a very useful area and economists have been studying the effect of different factors or institutions on price bubbles.Many experimental results show that asset bubbles in the laboratory is robust, and market mechanisms such as price limit, short-selling, transaction costs, etc., do not help price converge to fundamental values. Many research literatures proved that trader experience is the only factor to make price converge to fundamental values. Furthermore, experience has been extended to advice from experienced traders, that is, advice also helps price converge to fundamental values.In view of this, we considered the opinions of experts or commonly called as security analysts are more professional and have wider influence. Our security market has achieved considerable development since its establishment in the early1990s. In spite of this, the information efficiency of the security market is low and far from the strong efficient market. Accordingly, security analysts have been playing an important role in our market. In order to make a deeper awareness of the role of information intermediary, we study the effect of experts’opinions on prices and trade volumes in laboratory security market. We designed three different sessions, which are market with dishonest, market with honest expert and mixed market with dishonest experts. According to our experimental results, false prediction results released by dishonest experts destroyed the efficiency of the market, and led to more serious bubbles and higher trade volumes.Furthermore, why there are still so many demands when price goes far above fundamental value? Who push up price far beyond intrinsic values? Will the sophisticated traders trade against bubbles or ride on bubbles to? According to efficient market theory, price will not deviate from fundamental values for long times since sophisticated traders (arbitrageurs) will quickly trade against irrational agents to eliminate the mis-pricings. However, the real market operation results are not always consistent with theoretical prediction. So we divide all the traders into informed traders and uninformed traders and we also study the strategies of different kinds of traders experimentally. Our experimental results show that the informed traders’ ride-on-bubbles strategy push price higher and higher.In addition, we use data in the security market in our country to empirically examine the effect of heterogeneous beliefs (we use differences of experts’opinions as its proxy indicator) on stocks’ cross-section returns. The preliminary conclusions we draw from the empirical study is consistent with former research findings:Given that short-selling is constrained, heterogeneous beliefs will lead to contemporaneous over-pricing of stocks, and lower future returns.This paper makes several innovations in research object and experimental design. As to the former aspect, we examine the effect of expert prediction on market operation results including price deviations and trade volumes, and also pioneer the study of traders’ behaviors in laboratory. As to the latter aspect, we adopt indefinite period’s design, which is closer to the real market and rarely implemented in experimental research with real people. In addition, we do not fix the buy-out value once the experiment came to an end. Such experimental design mainly aims to avoid the end-play effect. Our research enriches the experimental study about asset price bubbles from the point of information intermediacy.Limited by the author’s research ability and other objective reasons, there are several inadequacies in this paper. Firstly, we simply divide traders into two classes by whether they can receive the private information, and neither do we study the strategies of uninformed traders in detail. Secondly, this paper does not study the bursting of bubbles and the strategies of different traders before and during the burst. Successive research may take these problems into attention.