Stock valuations are no help in timing trades

How can we value the stock market, and how can we time our entrances to it and exits from it? And are these questions related?

The questions are topical for two reasons. On valuation, a widely popularised measure (strongly advocated in this column), suggests that US stocks are overvalued. The Shiller cyclically adjusted price/earnings ratio compares prices to an average of earnings over the previous 10 years, and shows stocks are their most expensive since the 2000 Nasdaq bubble.

It has come under heavy fire of late. A popular objection is that accounting standards have shifted, making earnings look lower than they would once have done, and thus making the multiple look higher. But even a straightforward multiple of next year's earnings, at 17, looks high.

Separately, there is great concern that the S&P 500 is ready for a 10 per cent correction (or fall from peak to trough of at least 10 per cent). The S&P 500 is involved in a remarkably consistent rise. Since its last correction, when Standard & Poor's downgraded US sovereign debt four years ago, the upward trend has been inexorable. With analysts writing down earnings forecasts sharply, weakening US economic data after the bad winter (emphasised by deeply disappointing gross domestic product growth in the first quarter), and jitters over the Federal Reserve's intentions for interest rates, this makes a case for a correction - which has so far failed to come true.

Despite this, US stocks have ploughed through earnings season so far while still staying close to their record highs. And this should not be a surprise. Valuation says little or nothing about the short term.

Exhibit A comes from Barclays data featured in the Long View last week. Going back to 1988, it tracked the S&P's subsequent 12-month performance after it had been on each forward earnings multiple from 10 to 25. It has never fallen in the year after being valued at 16, 17 or 18 times forward earnings.

However, the S&P did once fall from a multiple of 11, and it suffered eight annual falls after trading at a multiple of 15. Unsurprisingly, it frequently suffered falls from multiples of 19 or higher. So the data strongly suggest that current valuation tells us nothing about the chance of a correction in the short term.

Looking at average returns in the year after each multiple, there is a clear relationship - the lower the multiple at which you buy, the higher your likely return over the next year - but it is so erratic as to be of no use. Strangely, the three instances of a 25 multiple saw an average subsequent return of almost 4 per cent. This is far better than several cheaper multiples, and demonstrates that once a market is in a bubble, it can stay there for a while.

Shiller p/es are no better. When it comes to predicting the next year, which is what an unhealthy preponderance of professional investors care about, valuation is useless.

But the longer run is different. To quote Rob Arnott of Research Affiliates, Shiller p/es are "breathtakingly powerful" at predicting 10-year returns. Research by AQR split historic Shiller p/es into deciles, and looked at subsequent 10-year returns after p/es in each decile. The result was an almost linear relationship - the cheaper the market on buying, the stronger the subsequent 10-year real return.

So you eventually pay a price for buying an expensive stock, but it may take a decade for this to become clear.

Does that mean we should get out of the S&P altogether? No. Even at its worst, the S&P tends to rise in real terms over 10 years. Exiting the market totally from time to time is dangerous. Also, valuation should take account of the cost of capital. Professor Shiller himself publishes his p/e data on a chart alongside long-term interest rates. The cheaper capital is, the more it makes sense to pay for stocks. Rates are, of course, very low - 10-year yields are at present almost exactly 2 per cent.

So is there nothing we can do with valuation data? Those with a 10-year view, who can invest in multiple markets, can shift from relatively expensive markets (like the US) to places whose Shiller p/es look cheaper (everywhere else, especially Europe and the emerging markets). Thankfully that is what I recommended at the turn of the year.

Four months is far too short a period to judge these things. But as many enjoy looking at short-term market moves, note that the S&P has started dramatically to underperform in the past few weeks. Valuation could not possibly have told you when this was going to happen, but it would have helped asset allocators steadily move their portfolios so as to take some advantage of it.

[email protected]

© The Financial Times Limited 2015. All rights reserved.
FT and Financial Times are trademarks of the Financial Times Ltd.
Not to be redistributed, copied or modified in any way.
Euro2day.gr is solely responsible for providing this translation and the Financial Times Limited does not accept any liability for the accuracy or quality of the translation