Created by economist George Taylor just three years ahead of the 1929 Wall Street Crash, his theory of the on the so-called ‘Hemline Index’ holds that when share prices go up, so do hemlines.
By the same token, during tougher economic periods, skirts and dresses get longer.
The original thinking seems to have run that, when women had more money and could thus afford to buy stockings, they wanted to be able to show them off and so skirts became shorter.
Whatever its precise origins among the short dresses of the Roaring Twenties, though, the fashion for miniskirts in the prosperous 1960s and rah-rah skirts in the Reagan boom years of the 1980s seemed to add weight to Taylor’s theory.
Correlation, not causation?
As it turns out, however, two Dutch economists crunched the data between 1921 and 2010 and found there was indeed a link – the only problem being that hemline lengths actually lag the market by some three years.
Quite aside from our usual arguments about spurious correlations where no causation exists, therefore, clearly any kind of economic indicator that lags what it is supposed to be indicating is of no practical use.
In fact, the same can usually be said of an indicator that precedes events – after all, if any kind of clear of relationship is shown to exist, then the market will already have discounted it.
In other words, the knowledge will already be built into stock prices and, as we always argue, the price you pay for a stock, not the growth your receive, is the biggest driver of future returns.
And speaking of the future leads on to our final point – the future is uncertain and therefore impossible to predict.
Economics matters but, whatever happens to be the fashion, there are no cheats to help forecast economic data in any reliable way.
- Ben Arnold is an author on The Value Perspective, a blog about value investing. It is a long-term investing approach which focuses on exploiting swings in stock market sentiment, targeting companies which are valued at less than their true worth and waiting for a correction.