Libor Manipulating RBS Traders Were Apparently Cool With Libor Manipulation
It's hard to see what is news about the latest Libor news but it exists so let's paste it here:
Royal Bank of Scotland Group Plc managers condoned and participated in the manipulation of global interest rates, indicating that wrongdoing extended beyond the four traders the bank has fired.
In an instant-message conversation in late 2007, Jezri Mohideen, then the bank’s head of yen products in Singapore, instructed colleagues in the U.K. to lower RBS’s submission to the London interbank offered rate that day, according to two people with knowledge of the discussion. No reason was given in the message as to why he wanted a lower bid. The rate-setter agreed, submitting the number Mohideen sought, the people said.
One way to conceptualize Libor is that it's the interest rate at which banks lend to each other on an unsecured basis. This is fine as far as it goes but late 2007 was farther than it went; by that point banks were skittish about unsecured lending and Mervyn King was already conceptualizing Libor as the interest rate at which banks don't lend to each other. But of course there are lots of rates at which banks don't lend to each other; 714.03% per annum is, for example, a perfectly good interest rate at which I will assert banks don't lend to each other. So King's formulation insufficiently specifies.
But that means you need a new concept! It just does; you can't avoid it by saying "well just try harder to say what rate you borrow at when you don't borrow." How do you get the new concept? Beats me; the CFTC has listed factors that were kosher to consider and they include prior Libor submissions, actual and expected central bank decisions, and "research documents," which are all, like, things you can look at, but which are none of them information about rates you can borrow at.1
Faced with that, a tempting way to conceptualize Libor is that it is the variable that interest rate derivatives solve for. There are eighty gazillion dollars worth of Libor-referencing derivatives in the world, and half of them are long and the other half are short, and if you submit a really high Libor all the short people2 will yell at you, and if you submit a really low Libor all the long people will yell at you, and if you get it just right you will have peace in your time until the next morning. And so you do. This is a stupid and wrong way to conceptualize Libor and we can stipulate that it constitutes massive fraud but what else are you gonna do? Here's a guy:
“This kind of activity was widespread in the industry,” said David Greene, a senior partner at law firm Edwin Coe LLP in London. “A lot of the traders didn’t consider this behavior to be wrong. They took it as the practice of the trade. This is how things operated, and it seemed harmless.”
The more I think about this the less I understand it, but here is a critical lacuna in my understanding: I can't for the life of me figure out whether the manipulating banks were mostly long or mostly short the rates they manipulated. I once speculated on pure a priori no-arbitrage grounds speculated that banks mostly profited by higher rates but I have no confidence in that; it seems at least in certain times and places and banks to not have been true.
So here's a question: did the Libor submitters know? Perhaps - actually it seems likely - this Jezri Mohideen character knew all of RBS's yen-Libor DV01 and so the submitter could be pretty sure that by lowering his yen Libor submission he was making RBS money. (Or making his clients money?3) It's less obvious that this would be knowable in, for instance, USD, given that RBS had a fair amount of USD-denominated, Libor-referencing loans that were not managed by the derivatives desk that was hassling and sitting next to the Libor submitters.
In any case, the RBS Libor submitter surely didn't know whether the other banks were net long or short; he could well have imagined that they were all managing offsetting books. In other words, that there was supply and demand for higher and lower Libors, and that they were doing their self-dealing part in an overall socially beneficial system of balancing that supply and demand, just as they would be if they were trading an actual thing instead of their imaginary guesstimate of an imaginary interest rate. If they conceived of their submissions as trades, rather than as scout's-honor promises about where they were borrowing, it's perhaps understandable that they weren't too bothered about leaning the way that made them money.
Or not, I dunno. But, again, what else were they gonna do?
RBS Managers Said to Condone Manipulation of Libor Rates [Bloomberg]
1.Really! If a guy on RBS's economics research desk called the Libor submitter and said "I think the Bank of Japan will raise rates, so you should submit a higher Libor," that would apparently be okay. And yes that's less obviously fraudy than saying "I will make more money if you submit a higher Libor, so submit a higher Libor," but it's no less unrelated to your actual costs of unsecured borrowing today, which are supposedly measured by Libor. If your shoeshine guy said "I think the Jets' secondary will really step it up with Darrelle Revis out, so you should submit a higher Libor," would it be okay to do it? Are "internal ... research documents" closer to that, or to fraudy self-interest, or to your actual borrowing costs? Blargh.
2.Who have your phone number. Or don't need it; here's Bloomberg again:
The bank’s seating arrangements helped facilitate these interactions. Money-market traders who made the firm’s daily Libor submissions sat on the same desk as derivatives traders whose profits rose and fell depending on where Libor was set, three people said.
3.Right? He too had to balance a market. Presumably if he's short rates and submits, like, "negative 20%," RBS clients who are long swaps against him will call him up and yell at him, which when you're in the business of selling swaps to those people is a bad thing. And if he's flat rates and has a big client who is short, he'll submit a low rate, no?