Diane Coyle: Evidence indicates not every market crash lowers spending

'Today's market is a teddy bear in comparison with the mid-1970s downturn'

Equity markets have injected fresh meaning into the word "volatility" over the past few weeks. A daily rise or fall of 1 per cent is now met with blasé shrugs. However, the downward trend is clear, and it has raised a worrying question: If the economic good times depended on share prices rising, do falling share prices make economic bad times inevitable?

There is a widespread belief that the "wealth effects" of a rise or fall in the value of shares on consumer spending must be greater now than in the past. One estimate, from economists at the OECD, suggests that a global 20 per cent market decline (in real terms) could lower GDP by 0.7 per cent after two years through this route.

The reason for believing the wealth effects of a bear market must be more important now is the argument that consumers have more of their wealth in shares. Shareholding has always been more important in the US than elsewhere, but in Britain a far stronger culture of individual equity investment has developed, thanks to big privatisations under the Conservatives, a number of tax incentives and the euphoria of last year's dot.com bubble. A fifth of households on either side of the Atlantic hold some shares directly. Adding in indirect holdings such as mutual funds or personal pensions takes the proportions to about a half in the US and a third in the UK. It has proved very difficult to pin down firm estimates of the size of the wealth effects, despite many efforts by econ-omists to do so. A comprehen- sive study of the US evidence by the New York Federal Reserve in 1999 noted that the value of shares held by American households had soared 260 per cent, to $9.45 trillion, between 1991 and mid-1998, with most of the increase occurring between 1995 and mid-1997. Economic theory suggests that consumers will spend only a portion of an increase in their wealth, partly to smooth out the benefits of a one-time change over time, partly because they might not be sure whether the gain is permanent or temporary. But in any case, with a huge gain in wealth like that generated by the 1990s bull market, a greater wealth effect than in the past seems likely.

Surprisingly, the evidence seems to point the other way. The researchers found that the effect varied greatly over time, was small and often not statistically significant, and was lower in the 1990s than both the 1950s-60s and late-1970s-80s. The maximum impact was a 10 cent rise in consumer spending generated by a $1 rise in wealth, and more often the figure was 3-4 cents. They concluded that if economists only had the evidence from the 1990s to go on, "We would be hard pressed to conclude that there is a linkage between the stock market and consumer spending."

One reason for this surprising result might be that we tend to over-estimate the increased importance of shareholding. Figures from the Fed show that 19 per cent of households in 1998 (the last time the three-yearly survey of consumer finances was conducted) held some stocks, with the median value of the portfolio $17,500 in 1998 dollars. More held shares indirectly, 45 per cent through retirement accounts and 32 per cent life assurance. The same survey for 1962 showed 18 per cent holding shares directly, with a median value of $4,000 in 1962 dollars, and 58 per cent indirectly. The proportions are not that different, so there does not seem to have been the clear upward trend we imagine. US equity ownership has only spread if you look at the trend since the recent low point in 1990. There has been a genuine upward trend in the UK, but most individual shareholders hold stocks in the demutualised companies, given to them in windfall handouts.

A second reason is that people recognise the ephemeral nature of stock market gains, and probably more so in economies where shareholding is widespread and investors correspondingly more sophisticated. This is the conclusion of a recent circular by John Butler at HSBC. He writes: "If, as we expect, the equity price falls are a temporary, confidence driven shock, then consumption will be lower but only temporarily so."

There are two sensible reactions to temporary windfall gains. One is to ignore them, recognising that it is long-term returns on equities that matter if you are saving for retirement or for the higher education of today's two-year-olds. The other is to sell the shares when you have a big enough gain to buy a car or house or whatever durable item the bull market has brought within your reach, but not to spend capital gains on day-to-day items. So perhaps individual investors are more like the rational agents of economic theory than we give them credit for. According to HSBC, the Bank of England's model implies the fall in share prices is likely to reduce consumer spending by 0.3 per cent over two years, reducing GDP growth by 0.2 per cent. Big enough to make a difference to the outlook for interest rates, but not enough to take the economy into recession territory.

The final point is that one reason it is hard to pin down statistically significant estimates of a wealth effect from stock market booms and busts is that there are not many examples for which enough data on the value of financial assets and consumer spending is available. There have not been all that many big market swings since the early 1960s.

What's more, other economic circumstances have been very different each time. In the mid-1970s bear market, commodity prices and consumer price inflation were roaring ahead, real interest rates were negative and the post-war exchange rate system based on capital controls was breaking down. Today's version is a teddy bear by comparison, taking place against a background of low inflation, new technological opportunities and free capital flows. And other forms of wealth ­ notably housing, which accounts for half of household net worth ­ are still showing healthy gains in both the US and UK.

Like Tolstoy's unhappy families, all stockmarket crashes are different. So it is just possible that the one that has been slowly under way for the past year might turn out to be the one that generates huge negative wealth effects. But in the absence of any evidence, to believe that it definitely will do so is an act of faith.

d.coyle@independent.co.uk