• C++ Programming for Financial Engineering
    Highly recommended by thousands of MFE students. Covers essential C++ topics with applications to financial engineering. Learn more Join!
    Python for Finance with Intro to Data Science
    Gain practical understanding of Python to read, understand, and write professional Python code for your first day on the job. Learn more Join!
    An Intuition-Based Options Primer for FE
    Ideal for entry level positions interviews and graduate studies, specializing in options trading arbitrage and options valuation models. Learn more Join!

Libor's Accuracy Becomes Issue Again

Joined
5/4/07
Messages
176
Points
28
Libor's Accuracy Becomes Issue Again

Questions on Reliability of Interest Rate
Rise Amid Central Banks' Liquidity Push


By CARRICK MOLLENKAMP



link to the article:
Business - WSJ.com


The accuracy of a widely used interest rate, seen as critical to judging the health of the financial markets at a precarious time, is coming under question for the second time this year.
Doubts about the London interbank offered rate, or Libor, center on whether banks are understating what it costs them to borrow dollars in stressed financial markets. Libor's reliability became an issue again this week when banks paid higher interest rates to borrow using collateral than they did for unsecured loans.

Those questions come as central banks inject liquidity into the market to restore the confidence of banks that have been reluctant to lend to one another. Other lending markets, including commercial paper, which are short-term IOUs issued by companies, have also struggled, potentially causing a credit crunch to spread throughout the economy.

Libor is supposed to reflect average bank-borrowing costs. Overseen by the British Bankers' Association in London, the rate serves as a benchmark for the borrowing costs of homeowners and companies. During the credit crisis, it has provided a gauge for whether banks trust one another enough to lend money. Last week, Libor rates surged in a sign that banks were having trouble borrowing money amid the problems at American International Group Inc. and Lehman Brothers Holdings Inc.

Concerns about Libor's accuracy emerged out of the rates being paid in another market used by banks to get cash. The Federal Reserve's term auction facility, one of numerous efforts the Fed has been using to fight the credit crunch, allows banks to borrow, but they must put up collateral.

Because of that, banks should be able to pay a lower interest rate than they do when they borrow from each other because those loans are unsecured. It is the same reason why rates for a mortgage, which is secured by a house, are lower than those for credit cards, where the borrower doesn't put up any collateral. In other words, the rate for the Fed auction should be lower than Libor.

But on Monday, the rate for the 28-day Fed facility was 3.75%, which was much higher than Libor. On Monday, the one-month dollar Libor rate was 3.19% while Tuesday's rate was 3.21%.
The Fed facility should be lower, said Scott Peng, a Citigroup Inc. U.S. rate strategist. The "market needs some accurate transaction-based measure of interbank lending."

Earlier this year, Libor appeared to be sending false signals. Banks complained to the BBA that rival banks might not be reporting their true borrowing costs because they didn't want to admit that others were treating them as if they had troubles. That led to a BBA review and the pledge that the rates banks contribute would be better policed. Every morning, 16 banks submit borrowing rates in a process that produces Libor rates at lunchtime in London.

Lesley McLeod, a BBA spokeswoman, said the BBA stands by the Libor. "Libor is accurate," she said. "It is constantly monitored and currently reflects the extreme market volatility present in these unprecedented circumstances."


Write to Carrick Mollenkamp at carrick.mollenkamp@wsj.com
 
Swap Spreads Widen to Record Amid Bailout Concern (Update1)

Related...

***

By Liz Capo McCormick

Sept. 24 (Bloomberg) -- The spread between the rate on a two-year interest-rate swap and Treasury yields surged to a record on concern that U.S. lawmakers may delay a Treasury Department proposal to bailout the banking system.

The rate charged to exchange fixed for floating interest rate payments for two years above Treasury yields, dubbed the swap spread and viewed as a gauge of credit concerns, climbed to 166.38 basis points from 139.25 yesterday. It hit a high of 157.25 last week before the $700 billion proposal was unveiled on Sept. 19. Today's spread was the widest since 1988, or as far back as Bloomberg compiles data...

...The swap spread is the premium charged over Treasury yields to exchange floating for fixed-rate payments. Swap rates are higher than Treasury yields in part because the floating payments are based on interest rates that contain credit risk, such as the London Interbank Offered Rate, or Libor...


http://www.bloomberg.com/apps/news?pid=newsarchive&sid=abZXOgYUy7GY
 
Are LIBOR rates calculated solely based on some mathematical equations or are based on demand and supply ?

What I am trying to understand is, Whether there are faulty models which model these rates or there are people who dont let the real numbers out (due to any reasons)
 
LIBOR rates is the actual lending rate among major lenders. It is self-reported; I suppose it is an average of those rates reported. The problem is not technical, it is human.

London Interbank Offered Rate - Wikipedia, the free encyclopedia

t is a filtered average of inter-bank deposit rates offered by designated contributor banks, for maturities ranging from overnight to one year.

The "again" refers to this, also from wikipedia:

On Thursday, May 29, 2008 the Wall Street Journal released a study suggesting that banks may have understated borrowing costs they reported for LIBOR during the 2008 credit crunch.[2] Such underreporting could have created an impression that banks could borrow from other banks more cheaply than they could in reality. It could also have made the banking system appear healthier than it was during the 2008 credit crunch.

For example, the study found that rates at which one major bank "said it could borrow dollars for three months were about 0.87 percentage point lower than the rate calculated using default-insurance data."
 
So the next question that comes to my mind is

Is the error accidental or intentional ? (If at all there is an error)
 
The problem is alleged, and is alleged to be intentional for purposes of camouflaging credit risk.

If the LIBOR spread is unrealistically low, viz a vis some other benchmark, whether it is "camouflaging" or not, that in itself is an indicator.

The current circumstance, however, isn't one of the LIBOR being quoted too low, necessarily. This spread can be charted historically -- it's at its highest since at least 1990 (if memory serves -- I'm not at my desk right now).
 
We've actually had some passionate arguments at work over this recently. We take yield curves built at the short end from observable and liquid market instruments, which in normal times are in pretty good agreement with LIBOR. Now, not so much; the volatility of the composite 3M market rate, particularly, is far higher than LIBOR would suggest. This makes operating a model of forward interest rates to do risk on an option that fixes based on LIBOR an interesting exercise.

LIBOR is an Olympic diving score (with all British judges). They poll the participants, discard the "outliers," and take an average. Here's the best description I can find of these poll answers :

LIBOR contributors...provide the rate at which they believe they could borrow should they propose so to do. The issue has been raised as to whether this has the potential to stigmatise contributions and therefore the BBA proposes to explore options for avoiding any stigma whilst maintaining transparency.
(From the BBA website.)

Some brilliant language there. What it tells us is that (1) the poll answer has nothing to do with an actual trade, and (2) banks have a reason to be less than forthcoming in some of their poll answers. Stigma is mentioned here; left out is the incidental fact that big swings in LIBOR can wreak havoc in swaps and IR options books with notionals in the trillions of dollars.
 
(From the BBA website.)

Some brilliant language there. What it tells us is that (1) the poll answer has nothing to do with an actual trade, and (2) banks have a reason to be less than forthcoming in some of their poll answers. Stigma is mentioned here; left out is the incidental fact that big swings in LIBOR can wreak havoc in swaps and IR options books with notionals in the trillions of dollars.

This obviously isn't recent revelation. Why was an opinion poll on what people might do a trade at chosen as THE reference for structured securities?
 
Are LIBOR rates calculated solely based on some mathematical equations or are based on demand and supply ?

What I am trying to understand is, Whether there are faulty models which model these rates or there are people who dont let the real numbers out (due to any reasons)

Here's how LIBOR is determined:

Source: http://www.liborated.com/libor_basics.asp#5

Basically, they survey a panel of international banks and take an average (in a way) of market rates those banks are charging. The banks apply a relatively simple formula to calculate the rate. The formula is in the link above.

But despite the mathematics involved it's not solely technical because the banks are basing the rates on what they think they can charge (market demand, as you mentioned).

Hope this answers your question..?
 
Due to the problem of LIBOR mentioned above, ICAP launched New York Funding Rate last year.

http://uk.reuters.com/article/hotStocksNews/idUKNYG00110720080610

The main difference is that NYFR is collected anonymously while LIBOR is not.

But, NYFR has only two maturities yet, 1M and 3M.

Also, I am not sure whether NYFR provides borrowing cost of other currencies.

NYFR's trend is quite similar to LIBOR, but it is usually higher than LIBOR.

Hope this helpful.
 
Back
Top