Qaddafi had high hopes for Citigroup.

Photographer: Marwan Naamani/AFP/Getty Images.

Goldman's Libya Salesman Was a Little Too Good

Matt Levine is a Bloomberg View columnist. He was an editor of Dealbreaker, an investment banker at Goldman Sachs, a mergers and acquisitions lawyer at Wachtell, Lipton, Rosen & Katz and a clerk for the U.S. Court of Appeals for the Third Circuit.
Read More.
a | A

I used to sell equity derivatives at Goldman Sachs, which puts me in fascinating company.  For instance there is Youssef Kabbaj, a London-based equity derivatives salesman who, in 2008, was described internally as "perhaps Goldman’s top salesman globally." Banks tend to throw around superlatives pretty casually, so I don't know how true that is, but if it is true then we should pause to hold a small parade for him. Look: 2008 was the final year of the greatest boom in financial structuring the world has ever seen, and Goldman was -- I am biased here -- perhaps the bank with the greatest mastery of financial structuring, and Kabbaj was, apparently, its greatest salesman of financial structuring, "possibly the No. 1 rainmaker at the world’s most profitable investment bank." He was at the very pinnacle of a certain sort of finance that we now look back on with awe and nostalgia and not a little bitterness. He was just 31. Good for him. 

Then it all went horribly wrong in July 2008, wrong enough that Kabbaj had to call in Goldman special forces from his hotel in Libya:

From his hotel room, Kabbaj called Michael Sherwood, one of Goldman’s top London executives, who said the bank would do whatever it took to get them out. Goldman’s security team called back, telling Kabbaj it was looking at options for “extraction” and ordering him not to leave the Corinthia. The hotel housed the U.S. embassy and a complement of armed U.S. Marines, not to mention hundreds of foreign witnesses to anything unpleasant that might occur. The next morning, a Goldman partner called to say the bank’s security team was increasingly concerned about their safety. They hustled to the airport and a flight to London.

Did you know Goldman had an extraction team? In my time selling equity derivatives, it never came up. 

That is from this amazing Bloomberg Businessweek story about Goldman's equity derivatives dealings with Libya, which were led by Kabbaj, a Goldman vice president (!) who was on track to get paid $9 million in 2008.  We have talked about those dealings a few times before; the basic story is that Kabbaj got the Qaddafi-era Libyan Investment Authority to enter into $1.2 billion worth of equity derivatives trades on bank stocks just before the 2008 crash in those stocks. That worked out extremely poorly for the LIA, and its post-Qaddafi successors are suing in a London court to get the money back.  

The bulk of the story is about the relationship-building leading up to that moment of betrayal in July 2008. Kabbaj did a lot for his Libyan clients. He set up training sessions about finance. ("To help plan the training sessions, one of Kabbaj’s colleagues e-mailed to ask about the level of the Libyans’ knowledge of derivatives. He responded: 'Baaaaaaaasic.'") He helped them prepare internal memos for their bosses, explaining the trades. He treated them to dinners, musicals and prostitutes, with the "Lord of the Rings" musical being a particular favorite. He got an internship for Zarti's brother Haitem. And he "bought the Libyans a few copies of Liar’s Poker, Michael Lewis’s seminal tale of bond salesmen screwing over clients," because if you are teaching someone finance from scratch the very first thing you do is give them a copy of "Liar's Poker." Kabbaj and Goldman catered to the LIA's every whim, except one: When Zarti visited Goldman's offices in London, he "kept asking if he could smoke and kept getting told he couldn’t."

The idyll was ultimately ruined when the LIA brought in a 26-year-old lawyer named Catherine McDougall, who "was struck by their affection for Kabbaj, whom they considered a friend. One LIA staff member showed McDougall Facebook photos of Kabbaj hanging out with the equity team." But:

As she learned more, McDougall began to suspect the LIA team didn’t realize they hadn’t purchased actual shares. No one understood, she wrote, that if the underlying stocks went the wrong way, “they could lose all their money.” She asked to see the due diligence the LIA had performed before committing to the deals. They responded, she wrote, “Due what?”

A few more pointed questions from McDougall, and Goldman had to call in the extraction team.

It is an amazing story but I don't quite know what to make of it. There is something odd at the heart of the LIA's lawsuit. The bad thing that happened to Libya is that it put $1.2 billion into leveraged bets that some bank stocks would go up. In early 2008. The stocks went down, a lot, in short order. Because, you know, it was 2008. Did Kabbaj sneakily trick the LIA into betting that those stocks would go up? Well, not really. From the article:

The talk at the LIA, Kabbaj learned, was that Qaddafi wanted to emulate the leaders of Qatar, who’d invested in the shares of troubled banks. One target was Citigroup, which Abu Dhabi’s sovereign wealth fund had put $7.5 billion into less than two months before.

Kabbaj fanned the flames -- at one point he "texted the head of the LIA’s equities team to note that Citi shares were down, creating a buying opportunity" -- but the spark seems to have come from Qaddafi himself. And while the LIA's bet on Citi was more risky and leveraged than, say, Abu Dhabi's,  it's not so clear that Kabbaj was responsible for that decision. Libya's lawsuit says that "Goldman (and Mr Kabbaj in particular) began heavily to encourage the LIA Equity Team and Mr Zarti to obtain exposure to stocks on a leveraged basis," but there's no suggestion that the LIA didn't want that, or didn't understand it.

Though there is McDougall's claim that no one understood that, if the stocks went down, "they could lose all their money." But that's a weird thing not to understand. It is true of any stock! If you buy a stock, and the stock goes to zero, you will lose all your money. I mean, if you just bought plain old Citigroup stock in January 2008, you'd have lost 96 percent of your money within 14 months.  It may be the case that no one at LIA understood that, if they invested in the shares of big international banks, they could lose all their money -- but that was a widely shared delusion, even in early 2008. 

Instead, the main complaint seems to be that the LIA didn't understand the details of the derivative structure: They didn't understand the fine points of the documentation, and they didn't understand one big point, which is that "they hadn't purchased actual shares." The LIA's Citi trade was "a cash-settled forward purchase agreement for Citigroup shares with downside protection in the form of a put option at the same price as the forward," which I guess sounds a little complicated. (Let's give it an acronym for convenience, say CSFPAFCSWDPITFOAPOATSPATF.) You can also say it was "a cash-settled call option," which is equivalent, and somewhat less of a mouthful. Or you can say it was "a bet that Citi's stock would go up." That is more intuitive, but it is not equivalent. It is information-destroying: A cash-settled call option is precisely the same thing as a CSFPAFCSWDPITFOAPOATSPATF, but there are lots of ways to bet that the stock will go up, only some of which are identical to the CSFPAFCSWDPITFOAPOATSPATF.

There are times where the specifics matter. If a salesman says "hey you should do a hedge that will compensate you if interest rates go down," and you agree, and then interest rates go down, but the hedge doesn't actually pay out because of some weird structural feature that the salesman forgot to explain to you, then you have every right to feel aggrieved. If a salesman offers you a bet that Citi shares will go up, and you take the bet, and Citi shares go up, and you don't get any money, and you ask him what happened, and he shrugs and says "derivatives are complicated, man" -- you should probably sue him. But if a salesman offers you a bet that Citi shares will go up, and you take the bet, and Citi shares lose 96 percent of their value, it looks a little funny to go to court and say "well we didn't understand the structural nuances of the bet." The nuances are not your problem.  

It is a financial cliché to describe all derivatives transactions as "complex," and in a sense they all are. Their pricing tends to involve math, and they tend to come with long boring contracts full of weird provisions to cover unusual scenarios, and it's generally true that the banker selling them knows more about the math and the provisions than the client buying them. And so when the weird provisions are invoked to the client's detriment, or when the complicated math leads to a bizarre result for the client, the client tends to get angry, and courts tend to be sympathetic. But this isn't that. These trades worked as expected: They were levered bets that financial stocks would go up, and those stocks went down, so the bets were losers. The notion that these were "complex synthetic derivative products" is mostly a distraction: They were, sure, but that's not why they went to zero.  

There's a secondary complaint, which is that the LIA didn't understand how much money Goldman was making on the trades. Again it is hard to know what to make of this. They don't seem to have ... asked? 

I have been thinking a bit recently about salesmanship. It is at a low ebb, reputationally, these days. I wrote last week:

This sort of thing is as old as business, but we live in challenging times for it. Objective performance is easier than ever to measure, making it harder than ever to justify choices based on boozy dinners instead of rigorous cost-benefit analysis. Communications are easier than ever to record and monitor, making it harder than ever to operate by means of nods and winks. Technology has devalued human intermediaries: Now we buy stuff over the internet, not over dinner. Our society has an ideology of meritocracy, and a distrust of corrupt elites who use connections to get ahead. It's just a bad environment for the classic skills of salesmanship and relationship-building.

One way to think about the Libya case is that Kabbaj was really good at the classical arts of salesmanship. He did all the best salesman-y things. He spent a lot of time with the clients, paid attention to them, explained finance to them. He helped them write memos so they'd look good to their bosses, he got their brothers internships so they'd look good to their families, he brought them gifts of aftershave so they'd smell good. They wanted to do business with him because they liked him, and trusted him, and thought of him as one of them. He inserted himself so fully into their lives that they forgot he was a salesman.

In a sense, that is exactly how sales is supposed to work : spending time with clients to build close personal relationships and win their trust. But now, in Libya and the U.K. and the U.S. and everywhere, it is held in some suspicion:

The LIA is employing a legal concept called “undue influence” that’s more commonly used by wives against husbands—it’s novel in financial litigation. The idea is that one party to a transaction can have so much power over another that a contract between them isn’t valid. If Libya wins, investment banks everywhere will face the risk of lawsuits by clients claiming they were snowed.

Goldman may have made hundreds of millions off Libya, but it’s put banking dogma at risk. A bedrock principle of the securities business is that sophisticated investors can look out for themselves and don’t have recourse to the courts if they lose their shirts. If a huge sovereign wealth fund can successfully claim it was duped, there’s no telling who else can.

But "duped" is not exactly the right word. Libya knew it was betting on bank stocks to go up, and those stocks went down. It's not claiming fraud. It's claiming "undue influence." The problem is not that Kabbaj was lying. It's that he was too charming. Libya isn't saying it was duped. It's saying it was seduced

Is that a good legal theory? I have no idea. It does seem weird to say it's illegal, or even suspicious, for a salesman to befriend his customers. And yet it strikes me as plausible that a salesman could be too good, too seductive, that he could get so deep into his clients' hearts and minds that they'd have a good claim to send back anything he sold them. The traditional check on that sort of thing is that relationships are reciprocal: If you become your client's best friend, he probably becomes your best friend, and you're not going to put your best friend in a terrible trade, are you? It takes a particular sort of cold-bloodedness to spend months cultivating a close personal friendship with someone, and then use that relationship to rip him off to enrich yourself. One reason that salesmanship is so distrusted, these days, in the financial industry, is that people are more willing than ever to believe that derivatives salespeople are exactly that cold-blooded.

  1. Like Greg Smith, who explained why he left Goldman Sachs in his New York Times op-ed, "Why I Am Leaving Goldman Sachs," and his book, "Why I Left Goldman Sachs."

    Also I feel like that first sentence could have started with the word "Disclosure." Consider my Goldman ties disclosed.

  2. That too didn't work out: After his extraction, Kabbaj was "sidelined to minor accounts" and ultimately pushed out with a $4.5 million settlement.

    Also, I say "vice president," though technically his title seems to have been the European equivalent, "executive director." (See paragraph 17(2) of Libya's particulars of claim.)

  3. As remedies go, that seems rather tame; the Qaddafi-era LIA deputy chief executive, Mustafa Zarti, took a more direct approach when he discovered the trades were not what he had hoped. "He called Goldman 'a bank of Mafiosi' and said that he could behave like a Mafioso, too," and told Kabbaj "I can make you disappear, and nobody will ever hear back from you." Again this sort of thing never came up in my own career at Goldman, which perhaps explains why I was never quite the star salesman that Kabbaj was. The best salespeople take risks.

  4. See page 19 of Libya's particulars of claim:

    At the same time as pitching the Disputed Trades to the LIA, Goldman was, as aforesaid, involved in writing purportedly internal memoranda and presentations for members of the LIA's staff to present to the LIA's senior management, which memoranda and presentations recommended, purportedly independently, that the LIA enter into the Disputed Trades, and some of which misrepresented the true nature of the Disputed Trades.

  5. Haitem's "most recent work experience was at a 'Video Club' in 2003, where he was 'in charge of customer and reservation services,'" but "incredibly, Haitem Zarti outlasted Kabbaj at Goldman."

  6. Mostly financial stocks, that is, though ɐlectricité de France was also on the list. I don't really know how the LIA picked its stocks.

  7. Abu Dhabi's Citi investment was in the form of a high-yielding convertible preferred stock, meaning that it was in a sense less levered than just buying Citi common stock. (That is: For every $1 change in Citi's stock price, Abu Dhabi could expect to make or lose less than a dollar.) Libya's was in the form of options on Citi's stock, and was more levered than buying the common. (That is: For every $1 change in the stock price, Libya could expect to make or lose more than a dollar.)

    Still, Abu Dhabi lost most of its money too, because Citi's stock lost most of its value, and Abu Dhabi also ended up in court.

  8. Paragraph 26(2) of the particulars of claim complains that the final trades "were structured differently from the investments in the shares underlying the Disputed Trades which Goldman had originally proposed to the LIA in the summer of 2007, and which Goldman had structured as, and explained to the LIA as involving, leveraged acquisitions of underlying shares." That suggests that the original plan -- leveraged acquisitions of underlying shares -- was fine, and was what LIA really wanted.

  9. That's measured from the closing price on Jan. 25, 2008, when the LIA did its Citi trade, to March 5, 2009, when Citi hit its low point. (Split-adjusted prices were $266.40 and $10.20, per Bloomberg.) It got better, though if you bought Citi in January 2008, you're still down 83 percent even today.

  10. Though this scene from July 2008, involving Zarti, Kabbaj, and Kabbaj's associate Nick Pentreath, is pretty evocative:

    Kabbaj started drawing on a whiteboard, running through basic concepts like how options could be “in the money” or “out of the money,” and Pentreath began a technical explanation of the derivatives.

    Zarti again interrupted. “Youssef,” he said, “I’m asking you.”

    At that point Citi was down 20 percent from where the LIA put the trade on. If a client put me on the spot and asked me to explain a levered long trade on a stock that had lost 20 percent of its value in six months, sure, I might start stammering out technical explanations, or turn things over to someone else to do that. It's better than just saying "yeah you picked the wrong stock, is the thing."

  11. The first part of that sentence seems to be wrong, and the second part extra-right, in the case of Goldman and the LIA. The LIA claims that "precise details of the relevant Disputed Trades appear only to have been supplied by Goldman to the LIA after the relevant transactions were executed," and that "trade confirmations of the relevant Disputed Trades were only supplied by Goldman to the LIA weeks (and in some cases months) after execution."

  12. The LIA has some real arguments that the structure worked to its detriment, specifically that "each of the Disputed Trades was a complex synthetic derivative product, which unlike direct investments: (a) did not allow the LIA to enjoy dividends or voting rights during the term of the investments; (b) could not readily be liquidated or traded; and (c) might expire entirely worthless (whereas a direct investment was not subject to any term, and could have been retained by the LIA to await a subsequent recovery in the market price)."

    I'm not sure that dividends or voting mattered much here. But a synthetic trade is harder to liquidate than real shares, since LIA would have to negotiate the unwind with Goldman, and it really can expire worthless even if the stock might one day rebound. Citi never rebounded all that much, and LIA never seems to have asked for an unwind, so these complaints aren't exactly central to LIA's concerns, but they are real enough.

  13. Though the internship thing is touchy.

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.

To contact the author of this story:
Matt Levine at

To contact the editor responsible for this story:
James Greiff at