Since I come from a buyout background, it is hard for me to imagine anything more unsettling than the revelation that LIBOR might be... well... fixed.
The London InterBank Offered Rate (see what I did with the formatting there?) is a disproportionately serious figure used as the basis for Greenspan-only-knows how many debt structures. That its utility as a baseline risk measure would be called into question (even in the absence of any real manipulation or inaccuracy) is, in itself, very alarming.
LIBOR isn't a market priced rate. There is, at least so far as I am aware, no LIBOR index (LIX?) to trade on. Instead, a rather old and more than Byzantine mechanism is used to report the figure to Reuters. Specifically:
Every morning by 11:10 London time, "panels" of banks send data to Reuters Group PLC, a London-based business-data and news company, on what it would cost them to borrow a "reasonable amount" in a designated currency. The dollar Libor panel, for example, consists of 16 banks, including U.S. banks Bank of America Corp. and J.P. Morgan Chase & Co. and U.K. banks HBOS PLC and HSBC Holdings PLC. Reuters uses the reported borrowing rates to calculate Libor "fixings." To reduce the possibility that any bank could manipulate an average by reporting a false number, Reuters throws out the highest and lowest groups of quotes before calculating averages.
Of course, what the banks report to Reuters is a "best guess," of sorts, and, accordingly, in times of limited liquidity, reliability of the figure is less than stellar. But add to this the prospect that banks are understating the figure to disguise their distress and you have a rather volatile brew.
If you can't rely on banks, who can you rely on?
Ok, so that was a joke.
Bankers Cast Doubt On Key Rate Amid Crisis [WSJ]
Stressed Banks Underreporting LIBOR Rates [Naked Capitalism]