Skip to main content

Libor Epistemology

  • Author:
  • Updated:

My favorite thing about the Libor scandal is that it's an obvious outgrowth of what Libor is. Like:
(1) Banks signed a kajillion contracts with people to the effect of "we will [lend | swap | whatever] you some money at Libor + xxx%."*
(2) Libor is a number that a private data provider calculates based on numbers that a bunch of banks make up and tell to a trade organization.
(3) Everyone knows that.

The latest news is, basically, that some people think that's kind of a bad idea and are going to band together and stop it, which strikes me as vaguely sad in a passing-of-a-more-genteel-era kind of way, but whatever. This is what jumped out at me:

The Libor rate-setting process is not considered a regulated activity under the UK Financial Services and Markets Act, but US and European banks and interdealer brokers have suspended or fired more than a dozen traders in recent months following allegations of abuse.

It probably goes without saying that, when the nice man from the BBA calls you up and asks “At what rate could you borrow funds, were you to do so by asking for and then accepting inter-bank offers in a reasonable market size just prior to 11 am?,” and you just outright lie to him, then that's ... well ... bad behavior anyway, of some sort. But of course you don't actually know where you could have borrowed funds just prior to 11am unless you did so, and if you didn't, you're making it up a bit. So it's hard to know that you "lied."**

The more interesting thing, though, is that "not considered a regulated activity" bit. You could if you wanted to go write a list of the people to whom it is or isn't okay to fudge a little when they call you up and ask you exceptionally personal questions like "at what rate could you have borrowed money this morning?" There's surely an element of bluffing to at least some counterparty relationships: you probably don't want to go around lying about what you're selling them, but representing yourself as stronger or weaker than you actually are to improve pricing is not uncommon. Similarly putting on a brave face to the media is pretty accepted, with mixed results.

On the other hand, you could imagine credit agreements and rates derivatives where the rate set was not "Libor, as defined on some page," but "the rate at which our bank can actually borrow for that term that day." While that agreement would suck for pretty obvious reasons - it would pass on the bank's idiosyncratic credit-unworthiness to the clients - it would probably make it tougher for banks to fudge the relevant rate. Not just because the client would want to scrutinize it, but because fudging would so obviously and directly be a fraud: you'd be lying directly to a client in a way that directly, dollar-for-dollar, takes money from them and gives it to you.

But fudging your borrowing costs with the nice Libor man just seems so ... indirect. Of course it has the same effect - only, y'know, magnified to the tune of hundreds of trillions of dollars. But by separating out the nice person doing the data-collecting from the client who is losing money, you make the fraud less obvious - not to regulators (it's probably more obvious to them - it's all in one place, and also ooh nasty derivatives etc.), but to customers, and also probably to yourself. After all the nice Libor man is just a trade industry guy asking for an estimate; you hardly owe him the same care and scrutiny you'd put in a trade report or filing for a "regulated activity." And the customer's contract doesn't entitle him to a "right" Libor, only to the Libor on that Reuters screen.

Regulators consider Libor overhaul [FT]

* It's maybe worth noting that they're defined that way, i.e. by a Reuters page or whatever, like this:

“LIBO Rate” shall mean, with respect to any LIBOR Borrowing for any Interest Period, an interest rate (rounded upwards, if necessary, to the next 1/100 of 1%) equal to the offered rate for deposits in either Dollars or Euros for a period equal to the Interest Period for such LIBOR Borrowing that appears on the Reuters LIBOR01 Page (or any page that can reasonably be considered a replacement page) at approximately 11:00 a.m., London time, two Business Days prior to the commencement of such Interest Period. If such rate is not available on the Reuters LIBOR01 Page, the “LIBO Rate” shall be the rate (rounded upwards, if necessary, to the next 1/100 of 1%) equal to the arithmetic average of the respective rates per annum at which Dollar or Euro, as applicable, deposits approximately equal in principal amount to such LIBOR Borrowing and for a maturity comparable to such Interest Period are offered in immediately available funds to the London branches of the Reference Banks in the London interbank market at approximately 11:00 a.m., London time, two Business Days prior to the commencement of such Interest Period. The Administrative Agent shall determine the LIBO Rate and such determination shall be conclusive absent manifest error.

In other words contracts don't (except as a backup) define "Libor" as a fact about the world, like "the rate banks can borrow at" - they define "Libor" as a fact about what a Reuters screen says"

** Also, until 1998, the question was “At what rate do you think interbank term deposits will be offered by one prime bank to another prime bank for a reasonable market size today at 11am?," which, like, it would be hard to get caught "lying" in response to that question.