Libor manipulation and the invisible whistle | FT Alphaville

Libor manipulation and the invisible whistle

There was a piece in the Telegraph on Sunday that may well sum up the thoughts on the Libor scandal of many who worked, or are still working, in banks. It’s called “Libor scandal: How I manipulated the bank borrowing rate“. It gives a sense of how it is that a hell of a lot of people didn’t question the manipulation of the rate.

Frankly, anyone with a Bloomberg terminal in 2008 would have been in on this, as one could see that the rates various banks had submitted did not reflect where they could fund. Deals were getting torn apart all over the place because no one could ramp up funding at a decent rate, despite what those screens said.

People knew how Libor was constructed, that it was a survey with barely a disciplinary mechanism in sight (other than being kicked off a panel, which seemed a ridiculous threat when so many were lowballing rates at the same time).

But it was just accepted. A great number of people didn’t even clock that there was a whistle to blow. Why might that be? We’d posit it’s because they didn’t even perceive what was going on to be wrong, and that it probably has a lot to do with social conformity (within the sphere of banking).

A classic study, the Soloman Asch comformity experiment (1958), showed that if a test subject witnessed others giving the wrong answer to a simple question, they too will often give the wrong answer. More on this and how we think it relates to Libor shortly.

To set the scene, a quote from The Telegraph piece we mentioned above, concerning lowballing the (anonymous) bank’s Libor submissions:

According to the trader, “everyone knew” and “everyone was doing it”. There was no implication of illegality. After all, there were 20 to 30 people in the room – from management to economists, structuring teams to salespeople – and more on the teleconference dial-in from across the country.

The discussion was so open the behaviour seemed above board. In no sense was this a clandestine gathering.

This particular person even went around to clients telling them about how Libor wasn’t the rate his bank could fund at, as part of communicating why the rates for loans were going up:

…we had to explain the “dislocation of Libor from itself”. As the trader put it, everyone knew that we couldn’t borrow at Libor, you only needed to look at the price of our credit default swaps – effectively survival insurance for the bank – to see that.

What that meant was that even though Libor may have been, for example 2pc, the real Libor rate the bank was paying was more like 5pc or 6pc. So in fact, we needed to be lending money at Libor plus 3pc or 4pc just to break even. That is what we were telling clients.

That’s not to say that everyone is somehow innocent in this and all should be forgiven. Of course not. As story in The Telegraph continues:

Looking back, I now feel ashamed by my naivety. Had I realised what was going on, I would have blown the whistle. But the openness alone suggested no collusion or secrecy. Management had been in the meeting, and so many areas of the Treasury division of the bank represented, that this was clearly no surprise or secret.

Now over to a modern version of the Asch conformity experiment that we mentioned earlier. It’s similar in many ways to the original, but with brain-scanning fMRIs.

Here’s an example of the type of question the subject would be asked in the old-style experiment (the new one uses shape rotation, but similar difficulty level):

Question: Give the letter that corresponds to a line that is the same length of the line in the box on the left.

Answering on their own, the subjects would respond C, almost without fail. If however, they were in a group of fellow subjects (the rest of the group were in fact actors co-opted by the experimenters) and those fellows all gave the answer A (speaking out loud so that the real subject could hear), then a good number of the real subjects also started saying A.

In the original Asch experiment, when asked 12 questions in this way (where the real subject answered last, after a group of 5-7 co-opted subjects had all given the wrong answer) out of a total of 18 questions, at least one wrong answer was given by 75 per cent of the participants. Despite how obvious the right answer is!

But, is this because one wishes to conform, hence knowingly gives an incorrect answer, or is it because one’s perception of the world, hence the lines, has actually altered?

This is where the fMRIs come into play, in an experiment done by Berns et al (2005).

The idea is to examine which areas of the brain are active when a subject gives the wrong answer. Activity in the area of the brain that deals with conscious decision-making and conflict resolution would perhaps indicate that the subject was going about deciding whether to go along with the group, despite what they thought was correct.

Alternatively, activity in a part of the brain that deals with vision and spatial awareness would perhaps indicate that the individual’s perception had changed. That is, they were giving the wrong answer not consciously, but because the consensus view had actually altered what they saw in the world — in effect, changing the length of the lines.

Want to guess the results?

The researchers found increased activity in the area of the brain that deals with spatial awareness, indicating that the subjects’ perception had indeed changed. As for those decision-making areas, they didn’t show heightened activity.

Even worse, when the subject did go against the group, there was heightened activity in the amygdala and the right caudate nucleus, which are associated with emotional salience — to an extreme, note that the former is associated with the fear of rejection too. Hence Professor Bern refers to this result, when the subject had gone against the group, as the “pain of independence”.

The NYTimes reported on the study when it came out and noted this (emphasis ours):

In many areas of society — elections, for example, or jury trials — the accepted way to resolve conflicts between an individual and a group is to invoke the ”rule of the majority.” There is a sound reason for this: A majority represents the collective wisdom of many people, rather than the judgment of a single person.

But the superiority of the group can disappear when the group exerts pressure on individuals, Dr. Berns said.

The unpleasantness of standing alone can make a majority opinion seem more appealing than sticking to one’s own beliefs.

Or, being a whistleblower is tough, and sometimes society is in dire need of those who are unwilling to conform.

Of course, Libor-setting is a bit more complicated than which line is longer than the other. However, the point remains that:

If other people’s views can actually affect how someone perceives the external world, then truth itself is called into question.

Hence a call for a change in the culture of banking makes sense. As would further measures to protect and encourage whistleblowers. As well as perhaps, further education in our own shortcomings:

There is no way out of this problem, Dr. Ariely said.

But if people are made aware of their vulnerability, they may be able to avoid conforming to social pressure when it is not in their self-interest.

Our message to you if you work in banking — the next time you see something that smells even a bit iffy, try sharing it with a few friends of yours who don’t work in finance. Better get their opinion… before the way you look at the world changes, as it may well be for the worse.

Related links:
What Other People Say May Change What You See – NYTimes
Neurobiological Correlates of Social Conformity and Independence During Mental Rotation – Biological Psychiatry 2005
Libor scandal: How I manipulated the bank borrowing rate – The Telegraph