Understanding LIBOR

In the last week there has been a rather big flap about LIBOR rates.  It is a very serious matter.  The Wall Street Journal pointed out a week ago that these rates might be misleading.  There have been various stories following up on this and noting the defects in the method of calculation and the implications for US markets.

We believe that the stories capture neither the significance of the implications, nor all of the possible reasons for the problem.

Four months ago we worked on this issue.  While we urge readers to revisit the entire article, here were some of the key points:

Readers need to know the following:

• Much of the popular discussion of LIBOR moves relates to other currencies, not dollars.
• The relevant discussion of LIBOR rates in US dollars pertains to so-called "eurodollars."  These are dollar deposits held outside the US (not necessarily in Europe).  These deposits are 20%+ of total dollar reserves.
• The rate is determined by a panel of banks trading in eurodollars.  They are big players in this market, but not necessarily US based banks.
• There is only a loose arbitrage between Eurodollar trading and rates in the US.  Many of the banks involved cannot move between the two markets, for example.
• The LIBOR rate most important to the US housing market is the six-month maturity, linked to some ARM's.  This is not the rate that you read about most frequently.
• US investors can trade Eurodollar futures at the CME. Many people do not understand that this is an interest rate instrument with very deep liquidity and a history of over twenty-five years.

The implications for the CME Eurodollar market are just as important as that for various business and mortgage loans.  No pundits seem to have noted this, but you can be sure the Merc traders have.

A Possible Reason

At the time of the original article, we were exploring the idea that US banks had adopted FAS 157, mostly doing so in advance of the deadline of November 15th, 2007.  Meanwhile, international accounting standards differ.

What if banks are concerned about disclosure, more confident in those who have adopted the US accounting  standard?

We tried to generate some exploration of this idea last December, but without much luck.  It is not a topic that hits the "sweet spot" for the leading economists on the web, and emails to specialists did not engage their interest.

Now it may be different.

Felix Salmon raises the question, as follows:

Are European banks significantly riskier than American banks? Looking at RBS's decision to raise $24 billion in new capital, it certainly seems that way: the move will take RBS's tier-one capital from a normal-for-Europe 4.5% up to a normal-for-the-US 6%.

And so it's maybe not surprising that US interbank borrowing rates are lower than European interbank borrowing rates. The spread at the moment is 4bp, which is significant enough, but Carrick Mollenkamp reports that it could widen further, to as much as 10bp, as Libor continues to widen out to reflect reality rather than wishful thinking.

He wonders whether there is a need for an interbank rate based solely upon US banks.

This has very far-reaching implications.  Most importantly, why should various loans in the U.S. which do not involve non-US banks use this rate?  Existing contracts, of course, cannot be changed.

We expect to see much more on this topic.

Comments

[oldprof]

Thanks, Marco.

[Marco]

Also important to understand the LIBOR OIS spread:

http://www.learningmarkets.com/index.php/20080930456/Stocks/Intermarket-Analysis/understanding-the-libor-ois-spread.html

[Turley Muller]

It's interesting to see the divergence in LIBOR again, we saw that last year and then it pulled in for a bit, and since widened out.

Generally, the differential between 1yr LIBOR and 1yr CMT has been 50bps. Most hybrid ARMS use either of these indices. Mortgage desk I worked on, I offered both, up to the borrower which index he/she wants, the only difference comes in the margin. 225 for LIBOR and 275 for CMT.

1yr LIBOR now is 3.23 and CMT is 1.90, difference of 1.33, and 83bps above historical trend. Kinda sucks for borrowers that are facing a reset and they choose LIBOR instead on CMT, but who has that kind of foresight?

I don't know- but late 2006, Citi started paying up for LIBOR ARMS - 3/1 and 5/1. Not changing the margin, but offering a higher price, (lower initial rate) I couldn't really figure it out, my boss was curious too, as to why. I harvested all kinds of historlcal Data from my Bloomberg, ran regressions, couldn't find much variance. Typical spread 40-60bps. It appears that investors were betting on an impending dislocation down the road with LIBOR ARMS generating higher yield.

For years, I didn't think much about LIBOR vs CMT , but Jeff, when you raise the question about tying US securities to Euro Banks- for what purpose?

I think the lesson is always choose a CMT loan. CMT is the average yield on Treasuries with one year of maturity. The FED can affect this rate, it is "default free". LIBOR seems to be an animal of its own, hard to influence with policy and heavily based on credit risk of financial institutions, opposed to a central bank.

Check out my blog for a couple articles I written about this, if you haven't already. Thanks for the article, great info as always.

[oldprof]

Maestro - The Fed Funds rate is usually for overnight lending and is the specific target of Fed policy. It is not a good indicator of interbank assessment of risks.

LIBOR has a lot of different maturities and is based on various currencies. There is really nothing like it in the US.

Thanks for the question, which is probably also on the minds of others.

Jeff

[Maestro]

"He wonders whether there is a need for an interbank rate based solely upon US banks."

Don't we have that already? Or is the Fed Funds rate different in a significant way that I don't realize?