Remember Lehman Brothers? Citigroup sure does. When Lehman went belly-up on September 15, 2008, Citi found itself on the other side of some 30,000 derivatives trades with a functionally deceased counterparty. With so many loose ends to tie up, Citi faced a business question: would the bank treat its fallen rival with businesslike compassion? Or tear into its chest cavity to feast on Lehman's still-beating heart?
Citi feels it did the former. But recent court testimony isn’t helping their case. In the frenzied days after the bankruptcy, Lehman’s estate claims, Citi capitalized on the distress to stick Lehman with a $2 billion bill swollen with “phantom transaction costs.” When one Citi trader suggested they were being too rough on Lehman as they sifted through its commitments, his boss responded: “Let them come challenge, let them come sue us.”
Needless to say, Citi denies it did anything wrong. The “phantom fees” Lehman decries were actually appropriate compensation for replacing a hopelessly tangled knot of complex derivatives, Citi lawyers argue. Still, Citi hasn’t been the only bank to face complaints from what remains of Lehman. Of the dozen-odd banks Lehman has challenged over post-bankruptcy derivatives contracts, 10 have already settled, including JPMorgan. All that’s left after Citi is a $1 billion claim with Credit Suisse. Zürich is undoubtedly tuning into this case.
Lehman’s demise meant two big tasks for Citi’s derivatives desk. First, risk management: Rebalance the bank portfolio to hedge for the loss of the $1.2 trillion in notional derivatives that Lehman had with Citi. Second, come up with a number, unrelated to step one above, of what it would hypothetically cost Citi to replace those 30,000 wagers, and leave the bill sitting in the smoldering heap of debts that was once Lehman Brothers. Lehman attorneys have called this second task an “imagination exercise.”
The first part isn’t really an issue here. Following Lehman’s demise, Citi’s derivatives desks found willing customers for about a third of the trades – racking up some fees the process – and traded with other derivatives dealers to cover the rest, Lehman has argued.
The trial hinges on the imagination exercise. As it turns out, there’s no single method for closing thousands of derivatives trades following the largest bankruptcy in U.S. history. These so-called close-outs require making broad assumptions about how much it will cost to replace the canceled derivatives.
Let’s say one of Lehman’s terminated contracts was a five-year interest-rate swap. Citi could be charitable and assume that it could buy a similar swap on a narrow spread and with a minimum of labor. Or Citi could pretend it’s not a global derivatives powerhouse and price its hypothetical replacement trade at the widest possible bid-offer spread then tack on liquidity fees for the trouble.
Lehman’s lawyers are arguing Citi took the latter approach. And they have Citi traders’ unvarnished trading-desk chatter to help make their case.
On a 2008 phone call played in court, for instance, Citi’s head of emerging market credit Marc Pagano told an employee calculating close-outs to “pretend they weren’t a dealer, they were a real money, slow moving deer account.” In other words, while pricing these close-outs, don’t act like Citi, but like one of Citi’s clients: an oblivious ungulate loping directly into the crosshairs.
During cross-examination, Pagano confirmed what he meant by the remark:
Q: Treat it as a real money account, slow moving deer account, means treat them like a customer at which you can get -- with which you can get away with charging full bid/offer, right?
As Lehman’s team argued, Citi rarely actually paid the full bid-offer on its actual derivatives trades. For sophisticated players, derivatives trading necessarily involves haggling, and no one in their right mind pays what’s initially quoted. In normal times that was true of Citi, as the bank’s head of electronic trading Biswarup Chatterjee testified on the stand:
Q: We didn't see in your testimony or in the document shown to you, there are no examples of trades where you were paying rather than receiving a bid/offer spread, right?
CHATTERJEE: That's right.
At another point Lehman’s lawyers asked Citi's credit expert witness Stephen Padovano about a specific instance in which Citi handed Lehman a bill of $4.5 million for closing out a position Citi had actually executed at a profit the day before.
Q: So what I'm saying to you, sir, is that if Citi already knew when it closed out the position on September 16th, if Citi already knew that it was successfully executing hedge trades at no cost, okay, or actually earning money as it was hedging, okay, notwithstanding that fact, right, Citi charged Lehman $4-1/2 million, assuming that it would have to pay bid-offer and liquidity on hedges that it was already successfully doing the day before. Right?
PADOVANO: In this case, yes.
That's not a commercially reasonable thing to do; is it, sir?
PADOVANO: In this case, I would agree with that.
Citi argues these types of examples weren’t the norm. Lehman didn’t leave Citi with a neatly ordered set of exposures rounded to the nearest $5 million and dated to highly liquid maturities. Instead the portfolio was a welter of on- and off-the-run derivatives, many of them far larger and weirder than the market could reliably supply. The more unusual the name, the higher Citi would have to pay to replace it. That holds especially true in a volatile time like September 2008.
According to Citi, about two-thirds were of non-standard sizes, and more than 90 percent non-standard maturities, requiring Citi to go trade-by-trade to determine what it was owed. Indeed, as Citi pointed out, Lehman took a similar trade-by-trade approach with a defaulted counterparty six months prior to its own collapse.
Still, the sausage-making details still aren’t pretty. To close out the terminated derivatives, Citi traders consulted actual broker quotes in the days surrounding Lehman’s bust, picking and choosing spreads to attach to each close-out. Unsurprisingly, says Lehman, Citi’s process chose the highest spreads they could find:
Q: Well from Lehman's perspective, it was the worst number of the bid, mid and offer for every single trade, right?
PAGANO: It would -- well it would mark -- yes, it would be the highest number because it would mark every trade against the bid or the offer.
But sometimes those spreads weren’t enough. So occasionally, Citi’s traders did what any resourceful derivatives desk would do: supplied their own quotes to be used in the Lehman close-outs:
Q: So the purpose for which you're telling Mr. Agrest to put prices on the screen is so that those prices would get picked up by the brokers, you would see the broker screens, print them and use them for purposes of closing out the Lehman claim?
On the stand, Pagano affirmed a few instances of Citi tossing up its own quotes, in each case offering wider spreads than the rest of the market had supplied, then using that to price the Lehman close-out – a process that Lehman compared to painting the tape (a definite no-no). But Citi says this only happened four times, and that one shouldn’t compare Citi’s quotes to others posted on different days. Prices can fluctuate widely over the course of a day, particularly in times of stress.
And to be fair to Pagano, he was doing just what you’d expect from a trader whose job it is to get the best deal for his bank, particularly during a time of worldwide instability. Legally speaking, the question isn’t whether Citi was insufficiently nice to Lehman. It’s whether making the assumptions Citi did was “commercially reasonable” behavior, as the contracts required. As Citi points out, the contracts also allow for pricing the close-outs based on “quotations (either firm or indicative) for replacement,” not prices the bank can negotiate for after the fact.
Lehman has also seized on the exact size of the bill Citi eventually put forward. As it turns out, that number was roughly equal to the $2 billion collateral Lehman already had with Citi. “We say the traders knew, the traders knew about the $2 billion, and the managers, certainly the senior managers knew about the $2 billion,” Lehman’s lead attorney said in his opening statement. “They set about going to try to close-out to meet or exceed that $2 billion threshold.”
Citi scoffs at the suggestions. The $2 billion in collateral covered all of Lehman’s obligations with Citi, not just its derivative bets. And although some traders knew about Lehman’s kitty, Citi points out that Lehman hasn’t gotten anyone to testify that they were actually trying to pad the numbers up to $2 billion.
The trial still has a ways to go and the two parties seem to be arguing past one another. But the case underlines the fundamental imprecision at the heart of derivatives trading. In the aggregate, banks manage derivatives risks with a fair degree of certainty. But topsy-turvy market ruptures like those of 2008 throw all these careful calculations into question. Lehman might be over-exaggerating when its lawyers suggest that Citi was big and powerful enough to make virtually any derivative trade it wanted, at any time, and still eke out a profit. But it also strains credulity to presume that replacing Lehman’s derivatives would require trading at the highest bid-offer available in each and every case.
It’s all the kind of murky and opaque stuff that makes you want to throw your hands up and let the lawyers figure it out. Or in the words of our friend Marc Pagano: “Let them come sue us.”
CORRECTION: In the original version of this story, one of Citi's witnesses was misidentified as Andrew Padovano. We regret the error.