Quantifying the Meltdown
The complexity of quantitative analysis is accelerating financial turmoil.
The chaos engulfing the world’s financial markets is remarkable
in its severity and complexity. The latest stage has seen the collapse of Lehman
Brothers, the sale of
Merrill Lynch, and the Fed stepping in to rescue
insurance giant AIG with an $85 billion loan. Today, the world’s central
banks pumped
in $180 billion in cash in an effort to resuscitate the global money
markets.
As financial turmoil accelerates, it is worth rereading The Blow-Up
(requires free registration), an article that we ran in November 2007 examining
the way that quantitative
analysis contributed to the credit crisis that has gradually deepened ever
since.
Much of what’s happening currently connects back to this:
the application of incredibly complex mathematical and statistical techniques
to financial markets. An article
in yesterday’s Financial Times highlights
how the failure of mathematical modeling to accurately foresee market behavior
is now exposing even seemingly safe institutions such as AIG to the wider credit mess:
This story is only available to subscribers.
Don’t settle for half the story.
Get paywall-free access to technology news for the here and now.
Subscribe now
Already a subscriber?
Sign in
On a wider level, AIG
failed to see how the fate of supersenior [pools of debt previously
considered safe] could be linked to
behaviour in other parts of the financial world. For what has made the price
falls so vicious this year is that all the institutions that had previously
piled this “boring” supersenior on their books have needed to sell at once. Hence
the development of a vicious, downward spiral.
These institutions can hardly be blamed. This morning I
spoke with Jiang
Wang, a professor at the Laboratory
for Financial Engineering at MIT’s Sloan School of Management. He says that
the models used by big financial institutions simply aren’t engineered to cope
with the kind of severe conditions we are now seeing:
“Quantitative models/tools have served finance well at the
micro level, such as valuation techniques, trading strategies, and specific
risk analysis and product design. However, they are not at the level of capturing system wide risks and
dynamics, and not intended to be. Much
more work and data are needed here.”
Unfortunately, as the
situation worsens, it becomes even harder to predict what will happen next.
