Same old story with a new twist
The NYT has an interesting piece on Merrill and the synthetic CDO’s that have caused so much trouble in the world. It’s the same old story about what happens when they start giving away free money, but this time around the secret ingredient in the patent medicine was the computer modeling that was supposed to make everything really, really safe:
The fire that Merrill was playing with was an arcane instrument known as a synthetic collateralized debt obligation. The product was an amalgam of collateralized debt obligations (the pools of loans that it bundled for investors) and credit-default swaps (which essentially are insurance that bondholders buy to protect themselves against possible defaults)…
The synthetic C.D.O. grew out of a structure that an elite team of J. P. Morgan bankers invented in 1997. Their goal was to reduce the risk that Morgan would lose money when it made loans to top-tier corporate borrowers like I.B.M., General Electric and Procter & Gamble.
Regular C.D.O.’s contain hundreds or thousands of actual loans or bonds. Synthetics, on the other hand, replace those physical bonds with a computer-generated group of credit-default swaps. Synthetics could be slapped together faster, and they generated fatter fees than regular C.D.O.’s, making them especially attractive to Wall Street…
as often occurs with Wall Street alchemy, a good idea started to be misused — and a product initially devised to insulate against risk soon morphed into a device that actually concentrated dangers. This shift began in 2002, when low interest rates pushed investors to seek higher returns.
“Investors said, ‘I don’t want to be in equities anymore and I’m not getting any return in my bond positions,’ ” said William T. Winters, co-chief executive of JPMorgan’s investment bank and a colleague of Ms. Masters on the team that invented the first synthetic. “Two things happened. They took more and more leverage, and they reached for riskier asset classes. Give me yield, give me leverage, give me return.”
It is amazing in retrospect that these potentially dangerous instruments grew into a $60 trillion market. The idea that you could buy or sell an “insurance policy” on anything — without any insurable interest — and then not put aside any reserves in case the thing faltered, is amazing, and nuts.
Oh, by the way, the cost of the AIG bailout just went from $85 billion to $150 billion, in part because of the $400 billion in credit default swaps guaranteed by the company.
November 11th, 2008 at 1:09 am
Like someone said, “To err is human, but you need a computer to really screw up.”
Ever wonder why the IPCC screws up on the “reality of man made global warming”? Computer models.
November 11th, 2008 at 2:39 am
The tone of the NYT piece is consistent with this blog post by a somewhat repentant former derivatives salesman.
November 11th, 2008 at 5:46 am
gs, not repentant, just disclosing. i just followed up with a post titled The Case for Derivatives http://www.informationarbitrage.com/2008/11/the-case-for-derivatives.html.
November 11th, 2008 at 7:48 am
Roger, your posts indicate that you made every effort to educate your clients. No personal imputation was intended by ‘somewhat repentant’; ‘somewhat rueful’ would have been a better phrase.
I am all for transparency, and I need no convincing that there is a case for derivatives. ‘The Case for Derivatives’ ends with a call not to throw the baby out with the bathwater. Speaking of which, I wonder how financial innovation will proceed after the bulk of derivatives transactions is transferred onto exchanges.
November 12th, 2008 at 8:05 am
OT but in the ballpark:
A few months ago I speculated that foreigners might form their own rating agencies because they no longer trust the American ones.
Indeed the Europeans are considering that option. They’re also realizing that a simpler alternative may be to take over the American ones: if European regulation of rating agencies is stricter than US regulation, the American entities will have to comply. And the Yanks still take the blame when something goes wrong.