Experimental financial markets with heterogeneous traders

Which role does information play in financial markets? How useful is information and how is it valued by traders? How do heterogeneously informed traders interact in a market, which trading strategies do they use in their pursuit for profit? How is the interaction of traders in the market reflected in empirically observed stylized facts? These are some of the most critical questions in finance. They have been the focus of my research of the last years and they are also the topic of this research project. I will try to find answers to these questions by conducting and analyzing several laboratory markets. In the last three years I (together with several co-authors) have developed a model of a financial market that can be examined in laboratory experiments. The most essential innovations of our approach are: First, our model is dynamic. It thereby overcomes one of the major weaknesses of earlier research relying only on static environments. Second, we use several information levels instead of only two used in most of the literature on the topic. This is critical to go beyond the straightforward (and not surprising) result that insiders are able to outperform uninformed investors. In our experiments so far we consistently find that information is not always useful for traders in a market: while insiders are the most successful traders we find that the worst informed regularly perform better (or at least not worse) than the average informed. This implies that gathering information is only useful if a trader can acquire insider information, else it is useless, probably even harmful. In the proposed project I continue to improve our market model by making the process for the arrival of new information more realistic. Consequently, in the new setting there will no longer be trading periods, but trading will be continuous and information will no longer arrive at fixed times but following a Poisson-process. In addition traders can decide themselves how much information they buy, i.e. the choice of the information level is endogenous. With the proposed project I want to tackle four research questions: (i) I want to continue the exploration of the value of information in markets with heterogeneously informed traders. (ii) In our markets so far we observed the most common stylized facts. I want to see whether the same stylized facts can also be found in the new market and how they are influenced by the specific setting. (iii) The proposed experiments should deliver evidence on the valuation of information by traders. This will be tested by making the choice of the information level endogenous. (iv) We observe that participants chose different strategies in our markets: most notably we find fundamentalists, chartists, and arbitrageurs. Their interactions play an important role for market dynamics and will be another major issue in this project. Being able to conduct this project at Yale University offers me the opportunity to discuss results and open questions with some of the most outstanding researchers in finance and experimental economics. This should lead to significant improvements of the model that would not be possible in Austria in the same time span.