[FoRK] Computational Complexity and Information Asymmetry in Financial Products (Working paper)

Russell Turpin russell.turpin at gmail.com
Sat Oct 17 14:23:00 PDT 2009


Interesting. One of the criticisms of the derivative markets, relative
to the financial crisis, is that the counterparties who sold
guarantees against risk took on more than they could cover given
systemic events the risk of which their models inadequately accounted.
This paper suggests that such evaluation may be computationally
intractable.

Of course, there's another reason why such events might be
inadequately accounted. If a financial firm's future is screwed,
regardless, in a meltdown of certain scope, then there is a certain
sense in which it is rational for the firm to completely discount the
risk of such an event. It's much like an individual agreeing, in
return for $10/day, and besides everything else in a contract, to pay
someone a million dollars if the individual keels over dead tomorrow.
Sure, I'll take that $10. Ten times over. Or a thousand times over.
There's not even a need for me to take out life insurance. From my
viewpoint. ;-)

My taking such contracts doesn't increase the chance of me keeling
over dead tomorrow. At least, not from natural cause. But large
financial firms discounting that risk might, in aggregate, increase
the risk of such a meltdown. And as this paper points out, domino
effects are not always easy to evaluate. Which means it may be damn
hard for future regulatory agencies to tell that risk-assuming
counterparties are adequately backed.


More information about the FoRK mailing list