A growing number of investors believe that US stocks are overvalued, creating the risk of a significant bear market, according to research by Yale University market scholar Robert Shiller.
The Nobel economics laureate told the Financial Times that his valuation confidence indices, based on investor surveys, showed greater fear that the market was overvalued than at any time since the peak of the dotcom bubble in 2000.
"It looks to me a bit like a bubble again with essentially a tripling of stock prices since 2009 in just six years and at the same time people losing confidence in the valuation of the market," he said.
However, he made clear that it remained impossible to time any fall in the market, and cast doubt on whether stocks would drop should the Federal Reserve raise rates later this week.
"I'm not looking for any big effect," he said. "It's been talked about for so long, everyone knows that it's coming. It's just not much of a big deal."
Prof Shiller added there was no historical evidence for a link between interest rates and share prices. "You would think that when interest rates are higher people would sell stocks, but the financial world just isn't that simple."
He defended his now famous measure of valuation, often referred to as the Cape (for cyclically adjusted price/earnings multiple), which compares share prices to average earnings over the previous 10 years. This adjusts for the cyclicality of earnings.
The indicator showed stocks were seriously overvalued before the market peaks of 2000 and 2007 — but it has also suggested stocks have been overpriced for the past several years, while prices have continued to rise. That prompted several attacks on the Cape, saying it did not take account of changing accounting and tax rules over time, and it was distorted by the sharp fall in earnings that followed the Lehman Brothers bankruptcy in 2008.
Mr Shiller pointed out the fall in earnings in 2008 came as part of a severe recession. "Companies like to take write-offs right away during a recession. Then their earnings can recover from there. If I average over 10 years I don't see that as a problem. The average includes the actual losses that companies have made."
He said changing accounting standards could create difficulties for his model but added: "I think we're better off with changed accounting standards than if we ignored all the changes that happened since 1871."
He said the recent bout of volatility "shows that people are thinking something, worried thoughts. It suggests to me that many people are re-evaluating their exposure to the stock market. I'm not being very helpful about market timing but I can easily see aftershocks coming."