sion making. There are papers that build on overconfidence to explain the
amount of trading [Daniel, Hirshleifer and Subrahmanyam 1998, Odean 1998
b]. Another stream of research is based on the value function part of prospect
theory [Barberis, Huang and Santos 2000, Barberis and Huang 2000, Hong
and Stein 1999, Zuchel and Weber 2000] and tries to understand the ob-
served correlation in stock returns. We will base our work on yet another
body of knowledge from psychology, i.e., the understanding of what people
perceive when they talk about the riskiness of an asset. This enables us to
come up with a risk-value framework which is behaviorally based and better
describes peoples’ risk judgments than EU.
Risk is at the center of most theories about asset valuation. It is therefore
essential to understand what investors perceive the risk of an asset to be.
Investorst risk judgments should serve as a basis of asset valuation. Therefore
results of psychological research on risk perception and risk judgment are
essential for asset valuation.
This paper starts from risk-value models. In a risk-value model, risk and
value are taken as primitives. To evaluate an alternative, the decision maker
first separately derives value and risk of this alternative. Then value and
risk are combined into an overall preference. While value is always defined
as the expected outcome, the essential question relates to the measure of
risk. Markowitz also used a risk-value model with risk being measured by
the variance of the outcome.
Quite a number of attempts have been made to come to a more realistic
concept of risk.1
New results suggest that risk perception can be modelled as the expected
value of the transformed deviation of the outcome from a benchmark. The
1The inclusion of third moments of the return distributions, see Kraus and Litzenberger
(1976), or taking semi-variance as a measure of risk, see Bawa and Lindenberg (1977) are
examples of early attempts which, however, are not based on psychological research.