Question marks around valuations are providing a real headache for Asian investors wondering whether to buy, keep or sell their US equity holdings.

Using equity valuations as a guide to future returns is notoriously hazardous; country GDP is even less useful. Neither will be much use when thinking about whether to shift from US equities this year.

Per Moldrup, head of manager selection at Danish pension group Sampension, has drawn interesting conclusions about valuations on the back of an emerging markets survey of existing and prospective fund managers he employs.

“There was a real range in the return predictions they gave, depending on the time range they looked at when evaluating whether current valuations were above or below the average,” he tells AsianInvestor.

This variable was so influential that he now doubts whether such measures are any use in predicting stock market returns. “I’m surprised when someone comes out and says the market is cheap or expensive,” he says.

Figures about the US illustrate Moldrup’s reservations perfectly.

Managers defending current US stock valuations have argued that forward price-to-earnings (P/E) ratios – which use projected earnings over the next year – are, at 16, close to their 30-year average. But count in the bull market of the 1980s and the average sinks to 13.

Indeed, a recent survey by Goldman Sachs shows that, using this longer time-frame (stretching back to 1976), the only time that forward P/E ratios have been higher than today was in the four years leading up to the tech crash.

In AsianInvestor’s survey of US equity managers, those who defended US equities – notably John Arege of the US value and core equity team at Goldman Sachs Asset Management, and Pam Woo, head of US and global sector equities at BNP Paribas Investment Partners – point to the dangers of relying exclusively on valuation ratios.

“Higher valuations alone do not represent structural risk to future returns,” says Woo.

Yet the majority of opinion among the managers surveyed was that high US equity valuations reflected expectations of future growth.

Last year’s runaway stock market growth seems to have assumed economic growth that is not yet materialising. But even if economic growth satisfies these managers, does this means that US equities are a good buy?

No, not necessarily. Numerous surveys have found the correlation between growth and stock market returns is weak at best, and negative at worst.

A 2012 review by Bank of New York Mellon concluded that “as counterintuitive as it might seem, data suggest that high economic growth rates do not necessarily correlate with the highest long-term stock market returns”.

A February survey for Credit Suisse, conducted by Elroy Dimson, Paul Marsh and Mike Staunton of the London Business School, makes even starker reading.

If equity managers had predicted economic growth based on data from the previous five years and invested accordingly, they would have gained 14.5% a year in dollar terms since 1972.

But if they had invested in stocks predicted to grow slowest, they would have done far better – gaining 24.6% per year.

The equivalent survey from 2010 looked back to 1900 and out across 83 countries, similarly concluding that investment in “high-growth economies” frequently disappointed.

For his part, Moldrup remains cautious that the US macro environment can deliver what last year’s 30% gains on the S&P500 implied.

In short – and despite the recent recovery from January’s near 6% losses – the market is vulnerable to bad news. “We are more sceptical of companies for which we think only a small disappointment would have a strong effect on prices,” Moldrup says.