Monday, 30 March 2009

Dunfermline Building Society - irrational incentives?

Robert Peston reports that Dunfermline Building Society is going to be taken over by Nationwide, and asks "why Dunfermline took such risks that ultimately cost the society its independence".

Here's an article with a bit more detail about the story and what happened.

I haven't been able to find out whether management were incentivised with highly leveraged short-term bonuses, but given that this is a Scottish building society, the obvious assumption is that they were not.

So why would they take such huge risks?

The answer surely is that they did not knowingly do so; but that they felt a pressure to chase profits and closed their eyes to the risk that usually goes with high returns. The pursuit of high profits is ingrained in the behaviour of managers even without financial incentives.

This could be seen as a social pressure to keep up with the next guy; a psychological pressure to be the alpha male; or a boundedly rational economic decision based on an inaccurate estimate of risk. All three point in the same direction: irrational corporate behaviour is not driven only by greed, but also by something inherent in human social interactions.

And don't forget that the depositors may have been influenced by high savings rates too - the management were passing on some of those risky high returns to depositors as interest, and the balance accumulated as reserves, owned by the members - who were also the depositors, as this was a mutual institution. So perhaps they were also guilty of not interpreting the risks correctly - or maybe, rationally, relying on the likelihood of being rescued in a crisis.

So this is a warning: don't expect that by regulating executive compensation, or banning bonuses, we will magically get a rational or risk-free financial system. The non-financial incentives will remain just as strong - and may be strengthened if the (potentially) corrective influence of financial rationality is removed.

Update: It has emerged that the society's 2007 accounts are highly suspect as they did not accurately explain the risks being taken. This highlights another important constraint on rational behaviour: availability of information. It's critical that appropriate information is available to regulators, depositors and shareholders to let them make the right decisions and minimise risk.

We don't know for sure whether the board itself had access to clear information to allow it to make correct decisions about risk. There are cognitive constraints on any set of individuals, which can stop a board from evaluating the portfolio of assets held by the company it runs. In the case of a large bank like Citigroup or RBS, it's a non-trivial task to design the right mechanisms to transmit risk information from the level of individual assets, up to managers and through the chain to the board.

In a small society like Dunfermline (with less than 300 people in head office) you would think this would be less of an issue. But it's quite likely that the board did not think they were running high risks, so transparency would have allowed this view to be validated or challenged externally.

3 comments:

Tom P said...

Interesting post - and I agree with your point about pay not being the driver. Had a stab at this issue myself recently -

http://labourandcapital.blogspot.com/2009/03/pay-as-evidence.html

Anonymous said...

As has been pointed out before by others, the packaging of CDO's and other complex credit instruments was really an exercise in passing risk not from those who had it to those that were willing to bear it, but from those that held it to those that didn't understand it. The ratings agencies gave these high yielding instruments a clean bill of health and smaller institutions such as the Dunfermline just accepted that Moodys would know.

The guys just didn't have a clue I'm afraid and didn't suspect just how much they were being shafted, and the instituation fell because of it.

Anonymous said...

If you look at Dunfermline what you will discover is that their net interest margin was very small (typically 40bps). This level of basic profitability meant that they were restricted in their expansion capabilities and were also forced to pursue volume to meet fixed overheads (Head Office, branches etc).

The expansion into commercial (IMHO) was an attempt to expand the asset base with minimal overhead increases. Unfortunately their timing was exceptionally poor and since they were trying to do it on the cheap they probably cut corners in terms of risk assessment and diligence of loans. I dont think you can label the 07 accounts as flawed as the book more or less doubled between FY07 and FY08. Furthermore, the LTV ratios which were probably quite normal in 07 looked crazy by 08.

As for the second poster's comment on CDO's this just isn't relevant in this case. Dunfermline bought mortgage books (not CDOs) from Lehmann and GMAC. These were not 'subprime', but did have self certification - I think this would classify them as ALT-A.