Skip to main content

"Investors Expect Libor To Be Replaced Within Five Years," Enter Five-Year Libor Swaps Anyway

  • Author:
  • Updated:

It's time to play survey results versus revealed preferences. First:

A key interest rate for more than $500 trillion of securities worldwide will be replaced by a benchmark subject to greater government control, according to a plurality of global investors.

Forty-four percent of those responding to a quarterly Bloomberg Global Poll said the London interbank offered rate, known as Libor, will be supplanted by a more regulated model within five years. Thirty-four percent predicted the rate will continue to be set by banks in the current fashion, while 22 percent said they didn’t know.

That's from a poll of 847 randomly selected Bloomberg users, which is sort of a fascinating data set; like, Dealbreaker is a Bloomberg user (but, sadly, not surveyed). The substance is interesting too, beyond the Libor question.*

But, anyway, the Libor question: the plurality answer is "Five years from now, do you think LIBOR will ... Be replaced by something more like a government-run rate." What is a government-run rate? Meh, whatever, but have a look at revealed preferences:

That is from LCH.Clearnet's data depository of cleared interest-rate swaps and actually tells you nothing about what we're talking about but it looks pretty. My unsupported hypothesis, though, is that interest rate swaps - which have a whole lot of Libor in them, especially at longer tenors - keep getting signed up; see also, for instance, data from the CME, focusing if you'd like on the almost $1bn of >5-year interest-rate swaps that were cleared over CME yesterday.** The OCC's more comprehensive data on swaps only goes up to 1Q12, before the real Libor fun, and is a bit down since 2010 peaks, but total interest rate swaps notionals are still in the $180 trillion area. Again not all Libor but safe to guess: lots of Libor, lots of it longer than 5 years. (Also: not much future-Libor, no? Like, much of this may reference treasury rates or Fed Funds or whatever, but not much of it references "whatever replaces Libor in the next five years.") Also lots of syndicated loans; casual examination of a recent big one (with five-year maturity) shows the telltale traces of, yes, Libor.

The basic point is that if you think Libor is going to vanish as a thing in the next few years, you might or might not share your hopes/fears with a Bloomberg survey, but you'd definitely think twice about entering into a five-year Libor swap or Libor-priced credit facility. (Because: unwinding swaps costs you money! Unwinding swaps because Libor Went Away probably costs you a lot of money, in expectation.)

A while back I pondered the excuse that it might be difficult to reform Libor because "any substantial changes could affect existing contracts that reference Libor." That sounded a little fake to me: the existing contracts tend to refer to a set of outputs (a Reuters page, a BBA determination), not a set of inputs and processes ("take 18 banks, ask them where they think they might be able to borrow, throw out the top and bottom 4 submissions, average them, etc."). So you can change the processes and inputs that feed into those outputs and as long as you can convince, like, Reuters of its goodness and justice, then it's still "Libor" and everyone continues as before. Perhaps a closely-government-supervised Libor submission process works for that; a "government-run rate" like the Fed target rate or whatever probably doesn't but I suppose it depends on how Reuters is feeling that day.

Which makes me suspect that the future of Libor is something like a return to its past. Richard Robb got some attention this week for pointing out that he'd told you so, where "you" is the CFTC, and "so" is that Libor would be manipulated. The important bit:

When the Chicago Mercantile Exchange launched its Eurodollar futures contract in 1981, it devised a nearly foolproof way to calculate the London interbank offered rate.

The CME would select 20 banks at random from a much larger pool and survey those banks for their perception of interest rates. It would discard a subset of the highest and lowest and average the responses that remained. At a randomly selected time within the next 90 minutes, the CME would conduct a second survey with a fresh random set of banks. Finally, it would publish the average of the results of the two surveys without revealing the identities of the banks that participated.

Sooooooooo ... I feel like you can't really go around saying "nearly foolproof" in the financial-regulatory arena because that's just setting yourself up to look like a chump but whatever. The thing to focus on is that this is not just a decent way around some of the Libor gaming, but more importantly that it is plausibly within the spirit of current Libor definitions, insofar as (1) it is a survey of banks and (2) it takes into account term and credit risks, unlike leading Libor alternatives like OIS swaps or Treasuries.*** You could just about imagine shoehorning this process into current contracts and calling it "Libor." Which is probably important, since, given all the swaps and loans that keep referencing it, the thing most likely to replace Libor in the next five years is "Libor."

Investors Expect Libor to be Replaced Within Five Years [Bloomberg]
CFTC Ignored Warnings of Libor-Gaming – 15 Years Ago [AB]
The 'mystery Libor' precedent [FTAV]

* Fun facts, I mean plurality opinions, include that the US will be the best investment opportunity and the EU will be the worst, though the EU comes in second for best too. Stocks will be the best asset class, bonds the worst. The global economy is deteriorating, especially in the Eurozone but also in China. And 56% agreed both that Greece will be out of the euro by the end of 2013 (69% by 2014) and that "Economic troubles in Europe will cause social instability including riots or other unrest"; those are not necessarily the same 56% but it stands to reason.

Also Bloomberg's phrasing is pleasingly informal throughout; the non-answer for "what [region /asset class/what have you] will perform [best /worst] next year" is "Have no idea," which, really, only 2%? They should have sent us the survey. We have no idea and are happy to admit it.

** Like LCH.Clearnet, the CME is cleared swaps and thus misleading: basically no swaps were cleared five years ago, and eventually they all will be, so it's a growth industry. But anyway the CME does show $125bn of USD swaps of 5-40 year tenors, so, yeah, a lot of Libor in their future.

*** So I was not following banking news that closely in 1981 but I gather that the world was different then. If you imagine that today there are like 10-20 banks that are global TBTF whatever banks, and the rest of the banks aren't, then a survey randomly chosen from a pool of 100-200 or whatever banks should in expectation show a higher interest rate than a survey from only the 10-20 big global banks, since those big global banks have TBTF subsidies and their smaller friends don't. So moving the calculation to New Libor would make rates broadly higher and screw borrowers, while arguably helping municipalities and making Bill Gross happy. You could build around that but it'd be klugey. Do you, like, run surveys in parallel for 3 months and then go forward with New Libor minus the average difference between the New and Old Libor surveys during that three-month parallel period? Then, of course, the big banks have a ton of incentive to game Old Libor in those last three months. I dunno, it strikes me as an interesting exercise.