Even adjusting for extraordinarily low interest rates, global equities are expensive and finding double-digit annual real returns over the next five years is going to be tough.

What’s worse, key markets, notably the US, are so overpriced that there is a high likelihood of an upcoming correction, according to a new study by Joachim Klement and Oliver Dettman of economics and investment consulting firm Wellershoff & Partners. (http://papers.ssrn.com/sol3/papers.cfm?abstracttid=2470935)

The results should give pause to all who think that going along for a central-bank-underwritten ride in the equity markets is always a good idea.

“We show that high valuations like those currently recorded actually lead to lower expected future returns and to increased risks of significant drawdowns, including possibly permanent loss of capital,” the authors write.

“To be clear, today’s high valuations are an alarm bell for the future that investors should take very seriously.”

The study looks at equity market valuations in 38 global markets using two main approaches. The first is by using a cyclically adjusted price-to-earnings ratio, often called CAPE or a Shiller P/E, which is like a straight P/E but which uses average company earnings over ten years adjusted for inflation.

The second is to do an adjusted ‘fair’ CAPE which takes into account economic conditions like interest rates, growth and inflation. That’s crucial because it allows us to see through the effects of tepid growth and extraordinary monetary policy to get a better sense of where stocks are and what history suggests may be next.

For the US, CAPE is now at 24.5 which is 37.6 percent higher than a ‘fair’ economically adjusted CAPE at a still chunky 17.8.

That implies a real risk of a correction over the next five years.

“In the United States, when the CAPE has been at levels comparable to today’s, the average drawdown over the next five years has been an eye-watering 26 percent,” according to the authors.

Of course those corrections range from the 80 percent suffered after the crash of 1929 to 1995 and 2003 when the following five years brought no correction of 15 percent or more (though of course the great financial crisis began in 2008).

That’s the rub for investors now. Overvalued stocks may be, but there is no promise that corrections arrive on any timetable. One difference between now and the pre-1997 period is that central bankers are far more prone to supporting falling markets than capping rising ones. Remember too that a lot of committed equity sceptics lost their jobs as money managers just before the 2000 and 2008 corrections.

Those high US valuations don’t just bring with them a high risk of correction, they also imply quite low returns over the coming five years, with the study predicting a 1.4 percent annual gain in real terms.

Taking a broader look, things look a good deal better. Global developed markets should return 7.5 percent on a real (inflation-adjusted) basis over the next five years, and global emerging markets 6.6 percent.

That’s not fantastic, but could be a lot worse.

Highlights include India, which should return 10.5 percent on a real basis, and Italy and France, which should do just about as well.

While this time always can be different, CAPE is a good guide to future returns because markets historically revert to mean. The authors show that correlations between the CAPE and future five-to-10-year equity market returns are usually higher than 0.7 in almost all covered markets.

If anything, one concern is that the earnings part of the CAPE calculation is now very high, with earnings in recent years at or near all-time peaks as a percentage of GDP. That may well be a permanent change, perhaps due to the effects of globalisation on wages and production costs.

But earnings can revert to mean in the same way that the price investors are willing to pay for them can. A small change could have a big impact.

Writing earlier this year James Montier of fund mangers GMO showed how if you use ten-year trend earnings rather than trailing ten-year earnings the P/E of the US stock market rises to a teeth-aching 34 from about 25.

Neither main conclusion — that we may well have a correction over the next five years and that returns will be moderate (well, lousy in the US) — is hugely surprising.

It does make it hard to be committed to being overweight, or even equal weight an asset like US equities which will barely beat inflation and carries with it a sizable risk of meaningful losses and volatility.

The arguments for international diversification get stronger seemingly every day.

(James Saft is a Reuters columnist. The views expressed are his own.)