Behavorial Models of Financial and Betting Markets

Daily people have to make decisions under uncertainty since they usually cannot control all factors influencing the consequences of their decisions. Also most significant decisions in economic terms have to be made under uncertainty. This is obvious for investments of the stock market or taking up insurance. But also many other cases like educational or entrepreneurial decisions involve uncertainty. All the examples mentioned have in common that they have a strong impact on economic growth and welfare. This is particularly true for developing countries as, due to the absence of financial institutions and sufficient health services, people here face higher risks than people in developed countries. Also the most severe economic problems at the global level like mitigating climate change, fighting poverty, or regulating speculation on financial markets cannot be properly analyzed without taking into account the involved uncertainties.

In view of these facts it seems evident that adequate models of individual decision making under uncertainty is crucial for positive economic analyses. A bulk of empirical studies has shown that the classical normative framework of expected utility theory is often not able to accommodate observed patterns of decision making. Well known examples from empirical finance are the equity premium puzzle or the disposition effect. The empirical evidence has motivated the development of numerous alternative theories which aim at giving a more realistic account of real choice behavior. Prospect theory which differs from expected utility theory be probability weighting and reference dependence, has become nowadays the most prominent alternative theory.

But also prospect theory is often violated in specific choice problems. Particularly, new experimental studies show that reference dependence is not adequately modeled in prospect theory. Additionally, there is evidence that individual behavior cannot be captured by a deterministic theory as prospect theory, since people often behave randomly or make errors. Therefore, stochastic theories have to be developed and tested.

The present project aims at contributing to an improved descriptive modeling of individual choice under uncertainty and application to the analysis of financial markets. We will also consider sport betting markets as they have many parallels to financial markets but are easier to analyze as for sport events only a small number of well-defined states of nature exist. Altogether the project has the following goals:

• Development of new theories of choice under risk which provide an improved accommodation of empirical evidence
• Application of these theories to financial and betting market
• Empirical, in particular experimental tests of the new theories
• Analysis of efficiency and regulation of betting markets and application of the results to financial markets