We didn't have a picture of the Tormar logo, sadly.

Photo by John Parra/Getty Images for AYS.

Ex-Goldman Guys Think Citi Was a Little Abrupt With FX Trades

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

Last month, we talked about Citigroup's lawsuit against Tormar Associates, "the joint family office of former Goldman Sachs partners Ron Marks and John Tormondsen," which happened to be short a whole lot of Swiss francs the day that the Swiss franc went up a whole lot. It's a sad old story: Tormar suddenly owed Citi, its prime broker, millions of dollars on trades that had gone horribly wrong. Citi demanded collateral, and Tormar didn't post it. So Citi blew Tormar out of its trades, seized the $10 million of collateral it had already posted and sued for another $25 million in losses beyond what was covered by the collateral. This all basically happened in a few hours. I mean, not the suing. But Citi liquidated Tormar's positions within hours of demanding collateral.

This left Tormar feeling a bit aggrieved, because those really were the wrong hours: The dollar quickly recovered a lot of its losses against the franc, and Tormar's trades mostly ended up looking fine. It's just that they weren't Tormar's trades anymore; Citi had blown them out at exactly the worst moment. Tormar said: "Had Citibank taken an appropriate approach, as required by our agreements, and worked with us, neither Citibank nor Tormar would have suffered any losses, as the positions quickly and inevitably rebounded in value."

Tormar has now filed its answer and counterclaim to Citi's lawsuit, expanding on that statement. Much of the dispute between Citi and Tormar comes down to whether Citi had the right to liquidate Tormar's positions as quickly as it did. Citi claims that, under the terms of the International Swaps and Derivatives Association agreements governing Tormar's derivatives trades, Tormar was in default as soon as it slipped below its required margin, and that Citi had the right to terminate the contract and blow out Tormar as soon as that default happened. Tormar argues instead that there were some delays built into those agreements, and that Citi didn't have the right to move as quickly as it did.  

Who's right? My general attitude toward ISDAs is that no one can know what they say, and that disputes over what they mean are either resolved amicably by business people or else litigated viciously for years. You can read the agreements if you want; here are the prime brokerage agreementISDA master agreement and credit support annex. It is safe to say that banks generally think that ISDAs give them the right to do what Citi did, blowing out clients if they have any worries that the clients won't be able to pay. It's reasonable that the clients might think otherwise and take the view that they have a few days to get the money together before being liquidated. The language of the CSA does not seem super clear to me on this point, which I suppose gives both sides comfort most of the time but runs the risk of causing awkward and expensive disputes like this one.

But there are two fun points in Tormar's counterclaim. First:

Throughout its history, Tormar has carefully managed its collateral requirements to avoid having to make any sudden margin calls, typically maintaining collateral approaching ten times that which Citibank required. Thus, although Citibank had required Tormar to post approximately $1 million in collateral in connection with its foreign exchange portfolio, by January of this year, Tormar had posted over $10 million in collateral to secure its trades.

Last time we discussed Tormar, I pointed out that it had hundreds of millions of dollars of trades (by notional amount) with just $10 million of collateral. But that was because it was being conservative! In fact Tormar's position was more than $300 million in notional amount, and Citi only required it to post about $1 million of collateral. Citi was offering Tormar leverage of more than 300 to 1, which makes 50:1 retail foreign-exchange leverage look pretty tame. And which looks pretty bad when the franc moves 35 percent in one day. 

Second:

In addition, the ISDA Agreements did not provide Citibank with the right to dictate that the close out of Tormar’s position had to be effectuated through Citibank’s own options trading desk. Yet that is precisely what Citibank required of Tormar, despite the fact that this requirement exacerbated Tormar’s losses.

Indeed, during the approximately five-hour period during which Citibank’s options desk effectuated the forced liquidation of Tormar’s positions, Tormar repeatedly identified better prices from other market participants than Citibank’s options desk was providing. Thus, Mr. Tormondsen continued to object to the manner in which Citibank was liquidating Tormar’s portfolio and repeatedly informed Citibank that its prices were markedly higher than others offered in the market.

Upon information and belief, Citibank’s forced liquidation was conducted through its own options trading desk to enrich itself at Tormar’s expense, as Citibank itself was suffering significant trading losses in connection with its own proprietary trading exposure to volatility in the franc.

One reason that banks like the ability to quickly blow out their clients during periods of panic is simple risk management: If Citi had to wait for days while lobbing notices back and forth with Tormar, its positions could well have deteriorated, exposing Citi to even more losses. But another reason that banks like that ability is that periods of panic give them an opportunity to profit. When it liquidated Tormar Citi was, essentially, trading with itself: Tormar had sold some options in its account, and so Citi shut down those trades by buying those options back (for Tormar's account) from Citi's own trading desk. The more the trading desk charged, the better off the trading desk was. And the more dislocated the markets were, the more opportunity Citi had to trade profitably with its captive counterparty.

Of course since Citi was advancing Tormar the money to buy back the options, those profits may turn out to be a bit of a wash for Citi: Every marginal dollar that it made on its trading desk, it lost in prime brokerage. But if it wins its lawsuit, this trading looks pretty good (for Citi).

This is just Tormar's claim, so I don't know that Citi actually traded like that.  But the claim is reminiscent of the disputes between the estate of Lehman Brothers and some of its derivatives counterparties, particularly JPMorgan and, hmm, Citigroup. When Lehman went bankrupt, its counterparties were in much the same place as Citi was with Tormar: They had the right, under their ISDA agreements, to close out Lehman's positions by trading with themselves, and they seem to have made quite a bit of money doing it.  Which left Lehman to sue, years later, and try to establish that those profits were excessive.

When we last talked about Tormar, I noted that Citi's hair-trigger ability to close out its positions had a pro-cyclical element to it: Any nervousness by Citi could lead to panic selling, increasing nervousness in the broader market. There is not much you can do about that, though; prime brokers really do need to be able to risk-manage their positions, especially if those positions are levered 300 to 1. But it would be a bit worrying if Citi's trading desk did take advantage of the liquidation "to enrich itself at Tormar’s expense." It's probably for the best if banks can cut loose their defaulting clients quickly, but it's not great if they enjoy doing it.

  1. See paragraph 29 of Tormar's counterclaim:

    Citibank required Tormar to agree to the liquidation of its foreign exchange positions just a few hours after it had informed Tormar that additional collateral was required. Accordingly, between approximately 12 p.m. and 5 p.m. on January 15th, Citibank closed out not only the 29 options in Tormar’s portfolio with direct exposure to the Swiss franc against the U.S. dollar, but also Tormar’s remaining 143 options, which had no direct exposure to the franc and were not under any duress. 

  2. See paragraphs 27-28 of its counterclaim:

    However, the express terms of the ISDA Agreements provided that Tormar had until the close of business on January 15th to post any additional collateral Citibank had requested earlier that day and that, even if Tormar did not meet that deadline, an Event of Default under the ISDA Agreements would exist only if Tormar failed to transfer the additional collateral within one business day after Citibank provided Tormar with written notice of its failure to post the required collateral. (Credit Support Annex, at §§13(c)(iv), 13(l)(ii).)

    Moreover, neither of the ISDA Agreements permitted Citibank to force Tormar to immediately liquidate its foreign exchange positions or to do so without any notice. Instead, pursuant to the Master Agreement, Citibank was not permitted to require termination of Tormar’s outstanding positions until after the occurrence of an Event of Default. (Master Agreement, at § 6.) As set forth above, an Event of Default with respect to Tormar’s purported failure to post required collateral could not have occurred under the express terms of the Credit Support Annex until at least one business day after Citibank provided notice to Tormar that it had failed to post the collateral by the deadline set forth in the Credit Support Annex.  

  3. Tormar points to provisions like paragraphs 13(l)(ii) of the CSA, which declares an Event of Default if it fails to post collateral "and that failure continues for one Local Business Day after notice of that failure is given to that party." But Citi points to provisions like 13(l)(iv)(B), which says it's an Event of Default "if at any time Net Equity shall be less than or equal to 60% of Required Margin," with no apparent notice or cure period. I don't know which one controls.

  4. Neither does Tormar; its claim is "upon information and belief."

  5. One delightful JPMorgan trader allegedly closed out a Lehman trade on September 15, 2008, only to re-open and re-close it on the next day, and again the day after, to get a more favorable price (for him). If you're trading with yourself, you don't have to be too careful about exactly when you traded.

This column does not necessarily reflect the opinion of Bloomberg View's editorial board or Bloomberg LP, its owners and investors.

To contact the author on this story:
Matt Levine at mlevine51@bloomberg.net

To contact the editor on this story:
Zara Kessler at zkessler@bloomberg.net