Tim Worstall points to a fascinating article in the Grauniad about the LIBOR manipulations a decade or so ago. The article is an extract from a book and, although it rather conflates the two separate LIBOR frauds that were identified, it is very well worth reading, if a little long. For those that don’t want to bother below is a potted summary.

LIBOR was (is) set by a bunch of representatives at various London banks reporting what rates they must pay to borrow unsecured cash for various time periods from other banks. Since this was a manual process with little or no oversight or validation it was susceptible to fraud.

One fraud was that traders in the same bank (or who knew exactly who was asked the question each day) would collude with the reporting representatives to shade the rate by a few basis points one way or another each day. Since these traders tended to have enormous open positions in derivatives based on these rates a single basis point difference meant millions of pounds (euros, dollars etc.) of profit. This had been going on for a while and is interesting only as a further example of how markets and organizations are made of individuals who won’t always have the interests of their employers at heart.

The second fraud only occurred as the global financial crises hit in 2007/2008. Essentially the longer term unsecured loan market died because banks weren’t lending squat to each other and banks were frightened that if they reported rates that were higher than their competitors they would be considered a bad risk by the market and then they’d suffer a bank run.

This latter fraud is the one I’m interested in because it shows an interesting twist on the famous prisoner’s dilemma. As the Grauniad article explains:

Soon banks began to submit rates they thought would place them in the middle of the pack rather than what they truly believed they could borrow unsecured cash for. The motivation for low-balling was not tied to profit – many banks actually stood to lose out from lower Libors. This was about survival.

The point being that essentially all the banks were vulnerable to a catastrophic bank run if they were perceived as not being trusted by their peers. So they preferred to lose actual money in loans they were originating (many of which were priced as rates of LIBOR + X) rather than appear as if their peers didn’t trust them. This is remarkably similar to the prisoner’s dilemma where prisoners have an incentive to not rat on each other and thus accept a limited punishment each.

What I find fascinating is that the banks self organised to this strategy.

Years of investigation by financial regulators and fraud offices on both sides of the Atlantic have failed to find any Mr Big or Commission of Secret Illuminati who laid down the rules. The banks all, individually, hit on the strategy of under-reporting their rates to the BBA and none of them saw any benefit whatsoever in ratting on their competitors. They were only found out when some people who were totally unconnected with the LIBOR reporters looked at the rates being reported and smelled a rat.

What is also interesting is that, as Tim says, most people still don’t realize that for a significant period of time the actual rates for interbank unsecured borrowing for almost any period longer than overnight was essentially infinite – as in the banks simply would not extend unsecured credit to each other, let alone anyone else. This is of course exactly what happened in the great depression of the 1930s, but with a critical difference. This time the entire system did not fail and we did not see mass bankruptcies as everyone tried to call in their loans. I think a good argument can be made that this lack of mass failure was due to the banks lying to the BBA and thus pretending that there was a market when there wasn’t one.

This is important because now that the world, and particularly the regulators, know that the LIBOR could be tweaked they have of course set up various checks to make sure that it isn’t. That’s stopped (presumably) all the type 1 frauds noted above, but I wonder if it has left the system more vulnerable in the event that we get to another moment when banks won’t lend unsecured cash to each other.