5 Conclusion
Our main result states that, as long as every agent is not constrained in
borrowing, a future albeit uncertain drop in endowments that is currently
anticipated with high enough probability by every agent will trigger a crash
if the drop actually occurs. Moreover, our simulations show that the crash
magnitude is positively correlated with the commonly agreed anticipation
level. Section 4 shows that those conditions are tight; that is, an endowment
drop may not trigger a crash if the anticipation level is not high enough.
The basic insight is that, when expecting future low endowments, agents
will increase their demand for securities to hedge against this event. This, in
turn, will raise the purchasing price of those securities and therefore will lower
their returns. In particular, to arbitrarily increase their holdings and thus
to induce such crashes, agents must not be constrained in their borrowing
capabilities.
The psychological factors that we put forth in our study are two-fold.
First, any factor leading an individual to believe in the occurrence of a drop
is relevant because such beliefs will act as self-fulfilling prophecies. Those
beliefs can stem from instance from herding, market rumors, fear of contagion
or panic (or possibly all those issues together). We do not sort which one
seems most likely, but rather we point out that they are all relevant because
they lead to the same phenomena: a crash anticipation. Second, it must be
true that all the agents in the economy agree on the anticipation (herding or
rumors have reached the whole market for instance). This second point must
18