When Professors Eugene Fama and Robert Shiller were each awarded this year's Nobel Prize in Economics (along with Lars Hansen) for their work on asset pricing models, the financial media immediately began to focus on the different views the two have about bubbles.
What's a bubble? Can we define it? Do we know one when we see one? When asked these type questions Fama's response is that the term drives him nuts. On the other hand, Shiller provides a very long definition with a checklist which includes rapidly rising prices, mass media coverage, and investors showing emotions, like envy and regret, for failing to have participated. Why's there such a difference in views? Hopefully the following will help you think about the issue.
I think a good place to start is with the term "irrational exuberance," which we might define as unsustainable investor enthusiasm that drives asset prices up to levels that aren't supported by fundamentals. The first reference to the term is believed to have been by Alan Greenspan, former chairman of the Federal Reserve, in a December 1996 speech, "The Challenge of Central Banking in a Democratic Society." In that talk, Greenspan stated that low inflation reduces investor uncertainty, lowers risk premiums, and implies higher stock market returns. At the time the S&P 500 was at 744 (SPY). It closed 1996 at 741. With earnings of $40.63, the price-to-earnings (P/E) multiple was 18.2. Were we in a bubble? Remember, if you state something is a bubble that implies that you expect stock prices to come crashing down as the bubble bursts.
A year later, the S&P 500 had closed 1997 at 970. With earnings now at $44.09, the P/E ratio had jumped to 22. Were we now in a bubble? A year later the S&P 500 closed at 1,229 and with earnings at $44.27, the P/E had jumped to almost 28. Was this now the bubble? By year end 1999, the S&P 500 had risen to 1,469 and with earnings now at $51.68, the P/E had risen to 28.4. And by March 24th 2000, the S&P had risen to 1527, putting the P/E at about 30. That was the peak of the market - a level that wasn't reached again until September 18, 2013 (remember though that the S&P 500 is a price-only index, that doesn't include the return from dividends).
The problem, as I hope the above illustrates, is that it's always easy to call something a bubble once it has burst. Throughout history we have had forecasters calling asset prices bubbles. And when they are on occasion correct, the media anoints them as a guru. But, rarely does the media hold such prognosticators accountable when their forecasts of bubbles bursting turns out to be wrong. In other words, the question is really: How do we know when irrational exuberance has unduly escalated asset values, which then become subject to unexpected and prolonged contractions? And can we know the answer? The history of Google's (GOOG) stock price provides a compelling answer to the question: How can we know?
Google went public on August 19, 2004 at 100. Less than five months later on January 3, 2005, the stock crossed 200 for the first time. Was a doubling in price in such a short time a bubble? Well, by June 27 of the same year it had crossed 300 for the first time. Was it now a bubble? By November of the same year (2005) it crossed 400 for the first time. Bubble now? By November 21, 2006 it had crossed 500 for the first time. Bubble yet? By October 8 of 2007 it had crossed 600 for the first time. Fast forward to the fourth quarter of 2013, and on October 18, 2013 it crossed 1,000 for first time. And as I write this it's trading at 1,033. And I think it's safe to say that Google meets all the requirements on Professor Shiller's checklist, including demand fueled by the media and envy and regret among investors who failed to jump on the bandwagon. So, is Google a bubble or not? As always, my crystal ball is cloudy. And that's really the point I think Professor Fama was making - it's hard to know. Thus, while sometimes it's easier than others to say there's a bubble (such as when the NASDAQ (QQQ) was trading at over 100 times earnings at its peak in March 2000), in general I'm more in Fama's camp than Shiller's, but not everyone feels the way I do.
The bottom line to me is that it's often very hard to know if high valuations simply reflect investor's views that risks are low, and thus risk premiums are low, or whether irrational exuberance and a "this time it's different" mentality has taken over. Another problem is that it's a very slippery slope in trying to determine if a valuation is irrational, and thus should be acted upon - is a P/E of 25 irrational, but 24.9 is not? In other words, at what point is there a clear signal to get out of the market. And, importantly, how will you know when to get back in? Remember, when timing the market you have to be right not just once, but twice.
The historical evidence shows that there are few if any investors who have successfully played that game. And that's why Warren Buffett not only advises you to ignore all market forecasts, but also that his favorite holding period is forever.
With all that said, some are questioning whether the Federal Reserve's easy monetary policy has created a bubble in asset prices. All we can really say is that valuations are relatively high, and valuations do matter. In fact, they matter a great deal in terms of predicting future returns. While they cannot tell us what returns will be, the Shiller CAPE 10 (cyclically adjusted price earnings ratio) does explain about 40 percent of future stock returns. It was that insight for which Shiller won the Nobel Prize (while stock returns are random in the short term, they're not random in the long term).
As I write this, the CAPE 10 is 25.25. Inverted (E/P) we get an "earnings yield" of just under 4 percent. However, because the Shiller PE is based on the lagged 10-year earnings, we need to make an adjustment for the historical growth in real earnings, which is about 1.5 percent per year. To make that adjustment we then multiply the 4 percent earnings yield by 1.075 (1 +[.015 x 5]), producing an estimated real return to stocks of about 4.3 percent, or 2.5 percent below the historical return. (We multiply by five because a 10-year average figure lags current earnings by five years.) Further, Cliff Asness of AQR found that when the P/E 10 was above 25.1, the real return over the following 10 years averaged just 0.5 percent - virtually the same as the long-term real return on the risk-free benchmark, one-month Treasury bills. The best 10-year real return was 6.3 percent (so you can get a good outcome), just 0.5 percent below the historical average. But, the worst 10-year real return was -6.1 percent (that's compound annual return over 10 years).
The bottom line is that you should be setting your expectations based on current valuations, and what they imply, not on historical returns. And, importantly, you should also be prepared to deal with the wide dispersion of potential outcomes that Asness showed were possible. In other words, the crystal ball is cloudy.