In the 12 months to July, Gordon Gekko was alive and well – running a fixed income business at a major bank. He was so busy scanning the CDO league tables and picking up awards at industry functions that he could no longer remember which of his many sports cars were in his underground garage and which were waiting at the car pound.
Dozens of glossy-haired salespeople worked for Gekko. They fanned out across the globe, selling AAA-rated paper to dodgy conduits and dopey money market funds. Gekko’s efforts were also supported by a “structurer” called Antoine. He selectively picked data from markets and elsewhere to give Gekko’s prices a veneer of mathematical respectability.
Meanwhile, at the unfashionable end of the trading floor near compliance, sat the real quant, Rasumovsky. Every day people came to this melancholy ex-physicist with complex bets, crafted from bits of interest rates, credit, foreign exchange, equities and other assets.
Rasumovsky was a professional sceptic. When presented with clever derivatives confections built using a model, he would pull them to pieces and reconstruct the derivative using simple plain vanilla components that traded regularly in the market. He’d then adjust the value downward if he thought it was too optimistic. Traders wouldn’t dare to bully Rasumovsky because his prices were fed directly to the compliance department and ultimately to the bank’s mark-to-market accounts.
But there was one part of the fixed income floor’s activity that was off-limits to Rasumovsky: Gekko’s AAA-rated paper factory. No matter how many billions of dollars of assets flowed through this factory, not a cent was supposed touch the bank’s balance sheet let alone its trading book. As a result, Rasumovsky wasn’t asked to check whether the valuations were valid.
One day, Rasumovsky was accidentally copied onto an email containing a detailed prospectus for one of Gekko’s products. Curiosity got the better of him and he decided to analyse its price like any other derivative. It was a CDO based on a pool of residential BB and BBB mortgage-backed securities (RMBS). The risks and returns were then sliced and diced up and one tranche was presented to investors as AAA paper.
Rasumovsky knew about the models used to price normal CDOs based on corporate bonds or loans, but over a series of evenings and weekends he realised how different this was. The underlying assets were adjustable-rate loans made to hundreds of thousands of US subprime borrowers.
If these borrowers managed to meet their repayments, everything was fine. Cash trickled down a waterfall with 20 separate steps, at the end of which was the CDO investor. If the borrowers couldn’t pay, their property was repossessed and sold, and these proceeds trickled down the waterfall instead.
Rasumovsky soon realised that the value of RMBS collateral amounted to an option to receive the cheapest of two things from the borrower: the stream of mortgage repayments or the property secured on them.
The fact that these loans were “adjustable-rate” also mattered hugely for the valuation. After all, many financially-challenged borrowers had been pulled in with low “teaser” loans. A jump in mortgage repayments could quickly make their properties vulnerable to repossession.
Indeed, US property prices actually had to rise to stop the value of the RMBS from falling if properties were repossessed. This was partly to cover transaction fees, partly to take account of the fact that properties might have to be sold at a discount and partly to provide enough of a cash buffer to give investors their promised return.
Rasumovsky wasn’t an expert on these things, but he still built a crude model attempting to show how US property prices fed into the CDO price. Unless they continued to rise at 10% per year, there was no way the 100 cents in the dollar valuations that Gekko’s salespeople were getting paid for could be justified.
Disturbed by this, Rasumovsky approached Antoine, who scoffed at him. Hadn’t Rasumovsky heard of the “savings glut”? These securities would never be marked to market on the bank’s balance sheet, because they were sold at par as soon as they were created – to institutions with an insatiable appetite for AAA-rated credit.
Rasumovsky thought investors were getting a raw deal and toyed with talking to compliance. But he bit his tongue. What, after all, did he know about mortgages? And it was his job to protect the bank’s balance sheet ““ not investors.
Then in August, the investors suddenly disappeared. Gekko was fired and the bank’s finance department called Rasumovsky. Could he find a market price for Gekko’s unsold CDO inventory? He dusted off his model and got to work. Things were easier now, as since early this year there existed a traded index ““ the ABX ““ which could be used to calibrate the pricing model.
This index showed that AAA RMBS was trading at 70 cents in the dollar. AA was trading at 40. Rasumovsky fed this information backwards through his model to derive a price for the BB and BBB subprime collateral Gekko had left on the bank’s books. The price consistent with the market information was zero.
Rasumovsky hit the send button. His email winged its way through compliance and the finance department. A week later, the bank reported huge write-downs and its chief executive resigned.
Today, Rasumovsky is busier than ever. He is cleaning up the mess of the credit crunch, applying the sceptical discipline of science to what was previously an enormous confidence trick. Some commentators think the quants were to blame. But the real damage was done by modern-day Gekkos and quant-lite structurers such as Antoine. The banks that listened to their heavy-duty quants early enough are the ones doing well today.
This article originally appeared on www.breakingviews.com. Â If you wish to take out a free trial to breakingviews, click here.
Copyright © breakingviews 2009
