Via Stone & McCarthy,
Libor is not a market determined interest rate, rather it is a trimmed mean from a survey of banks participating in a survey conducted on behalf of the British Bankers Association (BBA). There are a number of problems inherent in the survey-based Libor calculation.
First, there is the stigma associated with a higher than average Libor posting. This stigma results in an under-reporting of Libor. Banks that submitted Libor postings above the average submission, risked being punished by speculators, depositors, and other creditors including other banks. This dynamic rendered a "gaming" of submissions that rendered an understandable under-representation of Libor.
The Fed as well as most market participants were aware of this problem back in 2008. See: Everyone Knew, Not Just FRBNY, Banks Underreported Libor (Stone, July 13, 2012)
Second, there have been incentives for banks to attempt to manipulate Libor by submitting Libor postings that would alter the trimmed mean. The ethics of such manipulation is materially different than the aforementioned stigma associated under-representative of Libor. Here the manipulation was an attempt to foster Libor rates that enriched trading operations of the submitting bank.
Chairman Bernanke was asked in testimony several times yesterday whether Libor should be dropped as a benchmark interest rate. His answer was Libor should be repaired or some market determined interest rate should be embraced as an alternative.
He offered up 2 market-determined replacement possibilities for Libor:
(1) Repo Rates
(2) OIS rates
Both are market determined interest rates, but neither in our minds captures the essence of what Libor is supposed to measure. For example, dollar Libor is suppose to measure the cost of unsecured interbank borrowing of dollars by a subset of banks in London. Effectively, this is the eurodollar rate for these banks.
Our preference for a Libor alternative would simply be the eurodollar rate. The Fed publishes this rate daily. The Fed's source for the eurodollar rate is the ICAP eurodollar screen on the Bloomberg at 9:30am eastern time each morning. Certainly, if timing is an issue, a 6:30am snapshot could be taken to be more closely aligned with the timing of current Libor snapshot.
Why Not Repo Rates?
Repo rates represent the cost of secured funding as oppose to unsecured. Repo rates are subjected to issues surrounding dealer positions, the availability of repo collateral, and do not necessarily reflect the cost of unsecured funds to banks.
Why Not the OIS Funds Rate?
The OIS funds rate is simply the implicit expected average funds rate based on the fixed side of a swap agreement. Importantly, the funds rate is not the same as the cost of dollar deposits to banks outside the US. It was for this reason that the Libor-OIS spread and the eurodollar-OIS spread enlarged dramatically during the dark days of the financial crisis.
While the funds rate does represent the cost of unsecured inter-bank funding amongst US banks, it is below that paid be banks outside the US if for no other reason than US banks have direct access to the liquidity safety net provided by the Fed's discount window. Not all banks operating in London have even indirect access to the Fed's discount window.
The Fed effectively targets the overnight funds rate, and the term OIS funds rate is in concept a construct that should mirror something close to the average expected funds target over the specified term. In concept the term eurodollar rate is the expected average overnight eurodollar rate over the term. But the difference between the funds rate and the eurodollar rate reflects both perceived credit risk, and bank funding needs.
When banks outside the U.S. find themselves with increased term-dollar funding requirements, or stains due to balancing sheet dressing by US Money Market Funds associated with quarter or year-ends--the eurodollar rate will go up, not necessarily so for the overnight funds rate or the 1-mo, 3-mo, etc OIS funds rate.