AIG...And The Quants.

AIG...The New Disaster Dujour Standard.

With the news that the Government is plowing another $30 Billion into AIG and looking to break up the 90 year old Insurance Company and selling off the pieces prompted this post.

As you know the Federal Reserve Bank Of NY currently owns about 80% of the equity in AIG, in return AIG received a 24 month $85 Billion credit facility loan to finance and meet collateral and other cash obligations. Later on the US Government increased the investment to $150 Billion. I am not going to get into the specifics of the transaction as its more confusing just to read it off the various press reports. To explain it, would take me a month. But this is certain, AIG lost a whopping $99.29 Billion in 2008. An incredible figure that is just mind boggling, and those losses will just continue as AIG is holding the entire planet hostage. Moreover, just breaking up and selling off AIG is not going to mitigate the losses, as the US Taxpayer will be saddled with CDS and CDO counter party risk for years to come. It cost $5.2 Billion to cover/settle AIG's Lehman related CDS counter party claims, which was cheap considering the estimate claims were close to $100 Million. There are potentially thousands of these type of counter party claims yet to be settled.

But the questions remains, how and what happened to AIG? It can't be all incompetence and lack of supervision with a huge dash of greed sprinkled in... can it? How can an insurance company that is in the business of managing risk totally blow it? How can AIG traders take the type of positions (CDO Squared/CDS) without raising "red flags" within the compliance departments? How can these positions then dwarf to the type of losses that management weren't aware of? Did this just happen overnight? Who is to blame?

William Shakespeare once said the following in Henry VI:

"The first thing we do, let's kill all the lawyers".

I am here to say 419 years later:

"The first thing we do, lets kill all of the Quants".

The quantitative analysts are not totally to blame, there is enough obscene behavior to go around from greedy salespeople, turbo charged traders, clueless CEO's, incompetent risk advisers, and ill equipped regulators, but the trouble started with the financial models that were created by the quantitative analysts. From my years on Wall Street, these guys ate, slept, drank, and dreamt MS Excel Models. All they ever did was run regressions and run all types of models for all type of situations, they had everything covered, there was never a chance in hell that things would become unglued. How could they? These models were created by MIT, Stanford, and Harvard Phd's for gods sake! The gold standard of academia! They were stress and back tested over and over again in every conceivable fashion. Noble prize winners created these things. These guys were paid a vulgar amount of money and they were considered Rock Stars in their professions. So what happened?

Pure and simple - Models are only good in the sense that they predict normal behavior in normal times. There are only as good as the people creating them. Even brilliant people are fallible..they are not perfect. Most importantly, there is no way for a mathematical model no matter who creates them to properly hedge against something that is random or unknown. Quant Models do a great job of predicting, managing, gauging probability, but probability gets thrown out of whack say in the case of the Kobe Earthquake that leveled Japan and Bearings Bank (other factors contributed, but an example) in 1995. This is how Long Term Capital blew up and was eventually taken over by several Wall Street Firms. LTCM's Nobel Prize winning professors never modeled for the unknown or random, in that case - Russia defaulting on their debt obligations. In boring normal times, these models work like clock work, but when a "Black Swan" like event happens, get a helmet and head for the hills. This is the nasty little secret that the Quants know about but cant speak or write about, this is their game changer. This is their Waterloo...Totally random distributions. If they knew how to model "Black Swans", they can just pick the lottery numbers every week and save us the damage. They can model the next tsunami, hurricane, earthquake, and terrorist strike, buy options and derivatives, and make trillions!

But in the case of AIG...it was so easy to see that the models that AIG was employing were vastly limited and dangerous. They knew half way into the crisis that the models were blowing up...why didn't they change course? Simply...once the devil is out of the bag...there is no way to put him back in. The positions take on a life of its own. In this case the CDO/CDS positions that AIG had taken, had taken over the company. The fallibility of humans, these quants didn't want to take ownership or responsibility for their flawed models...after all they worked 99% of the time! The quants just alerted the traders to take on more leverage and deeper purchases of derivatives, because at the end of the day the market will always bail you out correct? It didn't happen, and AIG just collapsed under the leverage that was created. Listen...If AIG had quants with different models, guess what?... they would have used them. They had no such alternative as they had to continue to fall prey to the same flawed thesis. The AIG models just didn't take into account very simply...marked to market adjustments to their CDO/CDS positions....its this that leveled and murdered the insurance giant. The company had to come up with tens of billions of dollars just to settle mark to market every day. They lied in filings, lied to shareholders, lied to the press, most importantly they lied to themselves.

The central quant figure at AIG, was a widely known and quoted Yale University Professor who was praised by academics and Ben Bernanke himself in many speeches. AIG used his models to figure out which swap contracts would be profitable. What AIG didn't anticipate was how market forces and complicated contract terms not weighted by these models would effect these swaps in the short term. They knew that his models were limited and that his models didn't take into account mark to market adjustments. The sheer complexity of the swaps took over and it became near impossible to access the risks. The CEO even came out and stated no position at AIG is taken, if they don't pass the quantitative model test. They took out swap contracts from auto loans, credit card receivables, corporate loans, to you guessed it sub prime mortgages. They basically were guaranteeing losses in these asset classes. They would make good on losses, but most importantly and dangerously, if the positions went against them, they would have to make collateral payments to ther counter parties. Also AIG would have to mark these positions on their books at current mkt value. If the value falls...they have to take write downs. The quant model that AIG was using took into account total defaul, but not future mark to market or collateral adjustments, whats worse they never even properly hedged against this, this is what devastated AIG in the end. The model that was used took tens of years to develop, compiled years and years of data, it had almost everything covered. Now the U.S. Tax Payer is covering the losses...

Quants keep creating models that are flawed in nature, because that is what they do. They cant model the unknown, so they just skip that part. They will do it again when this crisis is over, explaining that we really have it covered this time. If only Shakespeare was still around.

Random Thoughts-

The market looks completely trashed at the moment. Total liquidation every where. Total break down of all that is holy. Can we rally? Do we rally? Is it capable of rallying? The mood is so bad that a counter rally can happen, and it could be a huge one.

This entry was a little longer then expected, I hope you get something out of it.

Please look forward to posts on the following soon:

1- What really happened to Bear Stearns? Lehman? Wamu?
2- Part 2 of CDS
3- How Lobbyists and PACS are making it difficult for Obama/Geithner to do the right thing.
4- Leverage and the End of Wall Street
5-The Art Of Securitization.
6- ETF's and Structured Finance..The devil is out of the bag.
7- The Disgrace of "Slumdog Millionaire"
8- VAR...Value at Risk
9- Whats the Deal with J.P. Morgan?


Good Night and please Keep The Faith!