9.2 Bubble-Crash Mechanisms
117
stock value and starting to sell stocks then precedes a crash. Note, however, that
sometimes bubbles fizzle out rather than end in a full-blown crash.
• Often a new technology or an equivalent development, such as expected riches
from the Americas or the South Seas drives the bubble. The hype for novel Internet
stocks in the dot.com bubble or the excitement about increasing values of homes
before 2008 are based on the same mechanism.
• Bubbles sometimes follow a dire period, such as a war and then really take off
when optimism in the future reappears. Think of the “roaring twenties” preceeding
the 1929 crash!
• It is remarkable that often no specific exogenous cause for a crash can be identified.
It appears to stem from an intrinsic instability of the economic system, resembling
a saturated vapor that spontaneously precipitates, stimulated by a random inhomogeneity.
It is obvious that the concepts such as “optimism” or “faith” are difficult to handle in a
physics framework and describe mental states that are better handled in a framework
using psychology or other behavioral sciences.
9.3 Behavioral Economics
The field of behavioral economics considers the driving forces deep seated in our
human psyche and how we make decisions and value their outcome. The measure of
how we value such an outcome is called utility. Historically economists assumed that
the agents of trade behave rational, which is also what we assumed in the first few
chapters. Such rational agents are so-called Econs according to Thaler [5]. Econs
calculate the expected utility or value of an action V by the well-defined expectation value V =
i p i O i , where p i are the probabilities and O i the outcomes.
In contrast, the vast majority of the population behaves non-rational at times and
are called Humans by Thaler. They calculate the expected utility by applying some
non-linear function of the p i and O i . As an example, consider the following bet:
• win 80 Euros with certainty, or
• win 100 Euros with a 90% chance and nothing with a 10% chance.
Econs would pick the second bet, because the expected return is 10 Euros higher,
but I would pick the first choice. I value the prospect of certainly winning 80 Euros
higher than the prospect of actually winning nothing with a 10% chance. Apparently
close to certainty the linear calculation of expectation values does not work for us
Humans.
Likely the most influential scientists, who analyzed the psychology of decision
making, are A. Tversky and D. Kahneman. They describe a theory [6] of how Humans
make decisions based on what outcomes can be expected, hence the name prospect
theory. Their work is based on a large number of psychological experiments, or bets,
similar to the one above, that they evaluated. They found a number of key concepts
Précédent

- 126/292

Suivant