Friday, 3 July 2009

Bad mathematics and dodgy economics

Robert Peston today attempts to demonstrate that bank executives have ripped off bank shareholders over the last twenty years or so since deregulation of the City.

However he does so with some classic statistical tricks. See which ones you can spot in the article if you wish, before I carry on...

(note that these tricks are actually courtesy of Andrew Haldane, a director of the Bank of England - Robert just communicates them to us)

from 1900 to 1985, the financial sector produced an average annual return of around 2% a year, relative to other stocks and shares...But in the subsequent 20 years, from 1986 to 2006, returns went through the roof: the average annual return soared to more than 16%
OK so far (though that "relative to other stocks and shares" is a bit tricky, because it's aready quite a meaningful excess return - banks haven't always been as safe as Robert and Andrew suggest).

But now what's happened?

The collapse of banks' share prices in the past two years has wiped out most of those gains: to March this year, when the low point was touched, the fall in UK bank share prices was more than 80%, an all-time record plunge.

What this means is that in the full period from 1900 to the end of 2008, the annual average return on financial shares was less than 3%, almost identical to the market as a whole.
Isn't that terrible...poor bank shareholders only made the same money as everyone else. What a con.

Except...for three little things that are really quite manipulative.

First, he picks the low point of the shares instead of a more objective place - like, say, today's share prices, which are about twice as high as the nadir. Anyone can find a low point with hindsight, but hardly any investor successfully calls the low or the high of the market in reality. So a more fair benchmark would be to look at the returns up to today, where shares have fallen 60-70% from peak and not 80%.

Second, that 3% is not the same as everyone else in the market: it is 3% more than the rest of the market. Which means that anyone investing fully in bank shares in 1900, as Andrew proposes, would have ended up twenty-two times richer than someone who spread their money across the whole stockmarket.

Third and most egregiously, he has completely redefined the time period for the convenience of his argument. All those 16% returns from 1986 to 2006 (minus the recent fall) have suddenly been spread over 108 years from 1900 to 2008! No wonder it looks a lot less.

In fact, if you take the current share price of banks instead of their low point, and look at the returns from 1986 to 2008, you get an annual return of 9.7%. Not an absolute return of 9.7% but 9.7% over and above the rest of the stockmarket. Not too shabby, eh?

Unlike the conclusions that Robert comes to. Admittedly these conclusions are informed by Andrew Haldane's confusing mathematics - I'd recommend any investment bank to hire him to help sell their derivatives - but Robert should know better.

The figures do not show that bank executives have been living high at the expense of bank shareholders. They show that executives and shareholders have both been in cahoots at the expense of bank creditors, depositors and taxpayers.

Robert gets the taxpayers bit right but he doesn't spot this: banks have been able to make such high leveraged returns because lots of people lent them money, at low interest rates, without realising that they were exposed to huge risks. The shareholders took little risk because limited liability means their downside was limited to the value of their shares; while the upside was infinite. Governments have now rescued many of these creditors - meaning that the risk wasn't perhaps as huge as all that. But many have still lost out, without any of the compensating profits that shareholders made over the last twenty years.

Robert proposes that shareholders should insist on bank executives being paid less and running the banks less aggressively. Actually shareholders should - if they can get away with it - get the banks to go back to doing business exactly as before.

The disciplines need to come from creditors and taxpayers, because they are the ones really on the hook. One discipline may be reduced leverage; although a fixed leverage ratio is probably too simplistic and too easy to game; another may be higher capital requirements, though that has the same potential problem.

The best single thing the regulator can do is increase transparency in the banking system so that creditors can make more informed decisions about risk, and impose appropriate discipline on the banks they lend money to. After all, a bank would insist on full disclosure of every aspect of my business, and regular up-to-date management accounts, if it lent money to me. Why shouldn't pension fund bondholders or Swiss buyers of mortgage-backed securities require the same?

A more detailed proposal here.

No comments: