The London Interbank Offered Rate (LIBOR) is an index (actually, there are 150 indexes covering 10 currencies and 15 maturities from 1 day to 1 year) that is reported daily by the British Bankers Association (BBA).  LIBOR settings are established each day as the result of a survey of a panel of large international banks.  LIBOR is meant to represent the interest rate on unsecured loans between large commercial banks.  More specifically, each day each bank supplies its estimates of the rates it would be charged as of 11:00 London time on interbank loans for each of the 150 currency/maturity combinations.  The BBA then takes the submitted quotes and computes a “trimmed mean” (for example, assuming 18 submitting banks, the four highest and lowest quotes are tossed out and the simple average of the remaining ten quotes is the official rate).

One of the contributing banks, Barclay’s in the UK, has admitted that it regularly manipulated the quotes that it supplied in order to achieve various objectives including hiding increases in funding costs during the financial crisis and increasing profits on positions held by Barclay’s traders.

There is evidence that Barclay’s manipulations of the first type (designed to hide increases in funding costs during the crisis) were supported and encouraged by regulators and central bankers, ostensibly in an attempt to project financial stability.  Barclay’s manipulations of the second type (designed to maximize Barclay’s profits) were made without the support and knowledge of regulators and central bankers.  They were designed simply to increase profits on the bank’s positions.  

This is a big deal with many ramifications, none of them good.

LIBOR is used as an index on hundreds of trillions of dollars of securities and derivatives.  For example, interest rate swaps are contracts between two counterparties to exchange interest rate payments, generally where one side pays a fixed rate and receives a floating rate and the other side pays floating and receives fixed.  LIBOR is typically the floating rate index on interest rate swaps.  In the event that LIBOR is artificially lowered relative to what it should have been, there is a benefit to the counterparty paying the floating rate and a detriment to the counterparty receiving the floating rate.  What is the incentive to manipulate the LIBOR setting?  Let’s take an example.  Suppose a bank (Barclay’s?) had ten percent market share of derivative contracts.  This would amount to $40 trillion or so.  Let’s further assume that the floating rate side of these contracts (the side tied to LIBOR) adjusts every quarter.  This means that, on average, every business day there would be $640 billion ($40 trillion/250 * 4) of LIBOR resets on the bank’s positions.  Across the entire book of business, there would probably be about as many contracts where the bank paid LIBOR (meaning they would benefit from a lower setting) as there are contracts where the bank received LIBOR (meaning they would benefit from a higher setting).  But on any given day the net position one way or the other could be substantial.  So, if by submitting a slightly higher or lower LIBOR the bank could move the setting in one direction or the other, there might be a non-insignificant benefit.

Of course, given the trimmed mean calculation, if any one bank submits a value which is way too high or low, it will simply be eliminated and will not affect the result.  In order for the attempted manipulation to be successful in influencing the official rate, it would seem that at least five of the participating banks have to be pushing in the same direction (again, because each individual bank would benefit on some days from higher levels of LIBOR and other days from lower levels, it would generally be the case that the individual banks would have disparate objectives).

During the depths of the financial crisis in 2008, the methodology for setting LIBOR fell apart.  This is because during the crisis large banks were not comfortable lending to one another at all.  Effectively, there was no market and hence no LIBOR.  The number that got reported as LIBOR was probably a lowball estimate; that is, it is likely that LIBOR was badly estimated during that time, and the bias was to the downside.

There is a strong likelihood of substantial litigation.  Investors earning LIBOR during the financial crisis probably earned yields well below market while borrowers paying LIBOR paid well above market.  This seems clear cut, except that most derivatives positions are collateralized so that credit risk is not a significant factor.  It was assumed, incorrectly as it turns out, that LIBOR would not strongly deviate from collateralized rates like Repo rates (Repo is short for repurchase agreement in which a securities holder borrows money by selling a security today with an agreement to repurchase it at a set price at a later date).  When LIBOR did in fact deviate from the collateralized rates, it arguably violated the spirit of the design of the contract.  So, replacement by a lower rate more closely reflecting collateralized rates at that time may have been a supportable strategy.  In any event, after substantial time has passed and at great cost, the courts will eventually sort out any liability. 

Beyond litigation, another obvious consequence of this event is identification of a need for an improved process for setting the benchmark rate for floating rate contracts.   Ideally, market transactions would be the basis for setting the benchmark rate.  These transactions could be in a futures market like Eurodollars, or in a cash market like Repo.

A further consequence is continued decline in confidence in the integrity of our financial system and large financial institutions.  The inevitable response will be a push for more extensive financial regulation.