Why economic models have gone out of the window

I had a good economics teacher at school, so I went on to study the subject at university - but by the end of the first year I was lost. Gone were conversations about how people as a group might or might not behave in any given situation and the stories of the emotions behind the graphs; in their place were endless lessons on econometrics and statistics backed up with increasingly complicated graphs and equations.

This stuff is pretty important there's no point in being an economist if you can't manipulate statistics and, of course, an excellent grounding in mathematics is vital in every area of life. But it's also only half the story. Economics is as much about behaviour as numbers, and models of all sorts can be destroyed by changes in the way people behave - as we have discovered in the past few years.

Andrew Smithers, of Smithers & Co, is good on this. He makes the point, for instance, that we have so changed the incentives for senior management over the last few decades that their behaviour no longer fits with the traditional modelling of corporate behaviour. They are hopelessly short-term, not very nice to their staff and devoted to the manipulation of the value of their options.

Subscribe to MoneyWeek

Subscribe to MoneyWeek today and get your first six magazine issues absolutely FREE

Get 6 issues free

Sign up to Money Morning

Don't miss the latest investment and personal finances news, market analysis, plus money-saving tips with our free twice-daily newsletter

Don't miss the latest investment and personal finances news, market analysis, plus money-saving tips with our free twice-daily newsletter

Sign up

You might once have expected something like quantitative easing (QE) to encourage companies to borrow cheaply to invest; certainly, that's what central bankers hoped it would achieve. But instead it just encourages them to push up earnings per share by borrowing to buy back shares, something that then holds up the stock market at levels that make no real sense. Again, a change in behaviour that can't be easily input into models suggests it isn't so easy.

Demographics and private debt levels can do this too. There was a time when cutting interest rates automatically made people spend more. But if large parts of the population are past peak spending age and the rest already spend their evenings staring blankly at their credit card statements, it just can't. People change faster than economic models which is why the models don't always work very well.

The same goes for financial models. History is littered with the wreckage of black box hedge fund strategies. This week brought us a few more. Particular mention must go to Cantab Capital Partners; its main fund rose 15% last year, but has reportedly lost 14% this month alone. Its excuse is that "it's unusual for one to see a sell-off in risk assets and a sell-off in bonds at the same time". That's true - it is unusual. But it wasn't entirely unexpected. If prices in the bond markets are artificially high, thanks to the existence of the central banks as the biggest buyers of debt, why should other asset prices move in the same way relative to them as they have in the past?

Look at the point on buybacks above and you can see why rising bond yields (falling bond prices) might hit equity prices: fewer cheap loans mean fewer buybacks pushing markets up. People kill models. If only Mr Kirk (the founder of Cantab) had been more familiar with Mr Smithers.

Investing has always been tricky, but at least in the past there was a straightforward basis to it. You figured out roughly what you thought a company or a sector's earnings were likely to be and then made a stab at guessing whether those earnings were reflected in its price. There were derivatives of this, of course you might also spend some time guessing what other people thought earnings might be so you could buy or sell before they did (the economists among you will recognise this as Keynes's 'beauty parade' theory of markets).

But now to get things right in the short term, we have to second-guess the big central banks. Then we have to second-, third- and fourth-guess other people's guesses on the central banks. I have a tiny bit more sympathy than usual for the losses of professional investors in this world of model failure.

But it still seems to me that the best thing for us ordinary investors to do is just to step back, make a few long-term assumptions and look at long-term valuations. The former probably means behaving as though the bond bubble is over and inflation is the end game, while the latter at the very least means avoiding the US stock market. It currently trades on a cyclically-adjusted price/earnings ratio of 23 times which, as Peter Bennett at Walker Crips points out, is "historically a straight sell".

My own portfolio is currently in sensible investment trusts and cash, so I'm not changing it before the summer. However, I can't go on holiday without mentioning the free-falling gold price. Gold isn't an investment in the way that it was when it was so gloriously underpriced a decade ago, and it is worth remembering that it hates rising interest rates. But it is still something you should hold as insurance. So falls in its price shouldn't bother you much.

Think of your car insurance. Let's say you pay £500 a year for it. Six months into the year, you haven't crashed. Your insurance remains unused. You are effectively down £250. Do you mind? You might be mildly irritated, but you also know that should you back into another car while parking in a frantic effort to arrive at the end-of-term play in time, you'll get value. So it is with gold. If there was no risk at all of another major financial crisis in our investing lifetimes, you'd want to sell it. But now? When the only thing keeping crisis at bay is an artificial boom in asset prices caused by experimental central bank policies? You don't.

This article was first published in the Financial Times.