Credit Suisse Traders Made The Unusual Mistake Of Committing CDO Fraud On Themselves

Author:
Publish date:
Updated on

I've been thinking a lot about financial industry compensation recently, and probably so have you, for different reasons. As a non-recipient of said compensation, I've been waxing philosophical about how your bonus can incentivize you either to put on low-risk trades that are unlikely to blow up your firm or to go instead with high-risk overlevered bets that look good in December but will leave the place a smoking ruin in March, by which point you'll be out of there with your pile of bonus CLOs. But if you don't take kindly to other people telling you what to do / "incentivizing" you to do it, there's always the do-it-yourself bonus, either in the traditional form (write checks to self) or in the slightly more complicated form of writing down the amount of money that you would like your trades to make, then getting a bonus based on the number you wrote down:

David Higgs, a U.K. citizen, pleaded guilty to conspiracy to falsify books and records and to commit wire fraud at a hearing before U.S. District Judge Alison J. Nathan in Manhattan federal court on Wednesday. ... "Rather than mark these securities down to market as we were required to do, at the direction of my superior Kareem Serageldin and with the help of other traders at Credit Suisse, I, with the agreement and assistance of Kareem Serageldin and others, manipulated and inflated the cash bond position markings of a trading book referred to as 'ABN1' in order to hide losses in this book and to achieve daily and month-end profit and loss objectives," Mr. Higgs said.

"I did this because I wanted to remain in good favor with my boss, Kareem Serageldin, and enhance my job performance."

When asked how he hoped to enhance his job performance, Mr. Higgs said he received a year-end bonus.

The SEC's complaint (against Higgs, Serageldin, and traders Faisal Siddiqui and Salmaan Siddiqui) is amazing all the way through. Basically their group marked their positions by (1) downloading trading data on ABX and other comps, (2) gulping at how much money they'd lost, and (3) replacing the comparable trading data with made up numbers to hit P&L targets. A sampling:

48. On August 31, 2008 [sic, probably should be 2007], the trading assistant told Salmaan Siddiqui that once again, they were "going to need quite a bit more P&L . . . Overall, in ABN we're currently down about 25 now . . . David [Higgs] said we can't be more than a couple down in ABN."

49. The trading assistant and Salmaan Siddiqui worked together to re-mark the book in order to falsely hit the desired P&L target.

And if just making stuff up wasn't enough, they would get their buddies to make stuff up:

55. On August 31, 2007, Salmaan Siddiqui sent a list of four AAA bonds to his bond salesman contact at Dealer C and requested month-end prices for the bonds. At approximately the same time, Salmaan Siddiqui communicated to his contact the desired prices on the bonds. In the space of less than one minute, the bond salesman at Dealer C sent the bonds and prices back to Salmaan Siddiqui, making them look like genuine third-party prices when they were not.

While the complaint is thoroughly entertaining/shocking, the basic story is only sort-of news, because it happened four years ago, CS announced it then, they marked down the mis-marked positions (in mortgage CDOs), and everyone forgot about it for four years until the indictments and pleas today. You could be cynical about this; as Yves Smith says:

Wow! Charges! Better yet, criminal charges! Finally the Administration is getting tough on crime.

Right. It’s tough on crime against banks. ... [H]ere we have an almost four year lapse in filing charges against ... two low level employee-miscreants that allegedly produced $2.8 billion in losses.

Here on the other hand is someone who buys it:

The allegations get right at the heart of what many have been so angry about: that Wall Street got rich by selling what they allegedly fully knew were bad mortgage-backed securities. And because they sold those mortgage securities so well, their machine needed feeding, so more and more people wound up with mortgages they couldn’t afford.

Well, no, that's perfectly wrong: what happened here is that Credit Suisse got poor by not selling what it didn't know were bad mortgage-backed securities. If Credit Suisse had gotten rich by selling these CDOs for more than they were "worth"* to unsuspecting dupes, then it would be ... well, it'd be Goldman Sachs. (Or at least Citi.) But instead they just hung on to their CDOs, not because they were taking an informed market view, or because they were just too ethical to dump the stuff on clients, but because they didn't know they were losing money.

A thing that I've thought about in the past, often in relation to Swiss investment banks, is whether you should be more worried about conflicts within banks (where there should be full transparency, but where everyone has the same goal of increasing profits and so might not be all that into fair client dealings or mark-to-market discipline) or between banks and counterparties (where there is an arm's-length market, but where it's maybe easier to, y'know, lie). Obviously most of the hits on the CDO business have been about banks somehow misleading counterparties about their CDOs, and it looks like those hits are going to keep coming. That is something you could worry about, but I've not worried about it that much. One reason not to get that worried about those CDO cases is that they aren't about fraud relating to the actual things that were being sold - in all the CDO cases, the underlying securities were fully and accurately disclosed, and the only fraud was about who had picked those securities. Investors on the other side could, if they so chose, do their own credit work and decide whether or not to buy.

Here, the Credit Suisse traders clearly thought they'd do better in the internal bonus market than the customer market. They could have sold these bonds - but they'd have been unlikely to find a customer who was enough of a sucker to buy them at valuations anywhere near to where they were carrying them. Higgs et al. correctly reasoned that they'd have a better chance of fooling their bosses than fooling the market, because their bosses were not armed with particularly impressive tools to check their work. As the complaint says:

51. Although the Price Testing unit was theoretically responsible for reviewing the prices assigned by traders, in practice the unit lacked the expertise to adequately challenge pricing by traders, and Price Testing personnel were often deferential to the views of traders.

Seems like something maybe worth looking into, no? It makes sense that you'd want your best CDO traders trading CDOs, rather than checking their prices, but surely a "price testing unit" can be trained to ... suspect that if the AAA ABX goes down, day after day, a long AAA MBS position should also be going down? Actually call independent third parties for quotes? Notice that, if you're long MBS and short ABX as a hedge, and you make a huge profit on the hedge, your long position shouldn't also be going up? (That's paragraph 76 of the complaint.)

I guess it's not great that other banks were defrauding their customers by dressing up mortgage CDOs a bit before selling them. But those CDOs were sold at "market prices" - you can quibble with the marketing, but no one paid par to buy bonds that were trading elsewhere at 70. There's a limit to how much you can get away with in an actual market with arm's-length third party customers. Turns out it's pretty important for banks to replicate that limit in their internal marking process. The problem here is not just that Credit Suisse had some traders who wanted their P&L to go up every day, regardless of what the market did - it's that Credit Suisse didn't have any way to measure P&L except as "whatever the traders say it is."

Former Credit Suisse Trader Pleads Guilty in Bond Probe [WSJ]

SEC Charges Former Credit Suisse Investment Bankers in Subprime Bond Pricing Scheme During Credit Crisis [SEC]

* Ugh. Here the story is that the traders marked the bonds on their books at higher than their market prices. Those market prices subsequently went lower, in at least some cases, on default experience etc. So they were definitely marked at more than they were "worth," but if they'd sold them at market prices, would they have been sold for more than they were "worth"? I guess, if they later suffered defaults, and someone was betting against them, ugh, sure, whatever.

Related