Should I buy or should I sell?

The U.S. stock market dropped last week, but the S&P 500 index is still 13% above its year-ago level and a whopping 181% above its March 2009 trough. If you are an investor, your goal is to buy low and sell high. Looking at the stock market, what would we do today?  Are prices too high? Are they too low? Or, are they just right?

The truth is that economists can’t answer these questions with much confidence. The reason is that the stock market makes fairly efficient use of available information. And even on those occasions when it doesn’t, riskless profit (arbitrage) opportunities generally don’t exist. That’s one reason why many economists neither actively manage their (usually modest) stock portfolios nor delegate that responsibility to active managers. For the most part, trying to beat the market is a game for fools or for criminal insiders, even if there are notable exceptions – like Warren Buffett, who has produced a remarkable risk-return tradeoff over several decades through a mix of leverage and exposure to observed pricing anomalies.

But there are occasions when asset prices get so far out of line with plausible fundamentals that safety seems prudent. In the late 1980s, for example, Japanese stock prices set repeated records, implicitly anticipating an endlessly rapid rise of profits. As they did, land prices in Japan rose to such ridiculous heights that people claimed the value of the Imperial Palace grounds in Tokyo exceeded that of all the real estate in California.

On such extreme occasions, we can say with some confidence (but still with trepidation) that the prospective return on such highly priced assets will prove quite low (or negative) in the coming years. What that statement means is that we expect the relationship of asset prices to fundamentals to revert to historical norms, but it could take a long time. Periods of extreme valuation just don’t last.

That said, there are several big issues that lead us to be cautious. First, one should always be humble in questioning the collective wisdom of millions of investors as reflected in the prices of widely traded assets. Second, it’s difficult to compute fundamental values. Third, even large investors have gone broke waiting for asset prices to return to fundamental values. And, fourth, structural economic changes can warrant changes in the mean valuation even over the long run. Put differently, historical norms evolve.

With those caveats, to see how the U.S. equity is doing today, we went to Robert Shiller’s website and downloaded the data for the following chart. (In 2013, Professor Shiller received the Nobel Prize in Economics in part for the work that is behind this chart.) Each month, he computes the “cyclically-adjusted price earnings ratio” (CAPE) for the Standard & Poor’s composite stock price index. Shiller’s calculation uses the average earnings of the firms in the index over the previous 10 years. Implicit in this comparison of prices and historical earnings is: (1) that the fundamental value of stocks derives from the corporate earnings that eventually will be paid out to stockholders; and (2) the ratio changes over time because future earnings deviate from past patterns, because stock prices move, or both.

The chart shows the level of the CAPE ratio since 1881 (the blue line). There are at least two big peaks that highlight the mean-reverting pattern of the CAPE ratio: the surges of the 1920s and of the 1990s. Neither of these lasted. Stock prices crashed.

Cyclically Adjusted Price-Earnings Ratio (CAPE)

Source: Robert Shiller website and authors’ calculations and update.

Source: Robert Shiller website and authors’ calculations and update.

So, where do we stand today? Well, that’s a matter of perspective. Over the entire period for which we have data, starting in 1881, the CAPE ratio has averaged 16.55. From its recent perch (on July 31) of 25.42, it would have to fall by 35% to go back to that level.

The adjustment could occur through a sustained rise in earnings or through a decline in stock prices. But earnings are already quite high: in fact, the share of corporate profits in gross domestic income (GDI) reached 10% in 2013, its highest since 1968. And the average rate of growth of GDI seems unlikely to top 5% annually over the medium term, reflecting the stability of prices and diminished real growth prospects.

Even so, there are good reasons to be less worried about the stock market than the historical CAPE gap of 35% suggests. To see why, look at the graph again. In addition to Shiller’s raw data, we show the 25-year lagged moving average of the CAPE ratio (red line). Throughout the 20th century, this measure fluctuated between 11.80 and 17.54 (with an average of 14.71). But at the beginning of the 2000s, the slow-moving trend broke out of this range, and today stands at a record 25.14. If the current, higher ratio represents the “new normal,” then the latest CAPE gap is only 1%, not 35%.

So, should we look at the full history since 1881, or just the last 25 years? Do we know which norm is the right one? Not really. But there are several reasons to be cautious in focusing on the higher number. First, the current level of the 25-year lagged average CAPE ratio probably is overstated because stock prices were unsustainably high in the 1990s. Second, another reason stock prices are so high today is almost surely that interest rates are temporarily so low. Third, and most importantly, we should not be cavalier in discarding the valuable information in the long history of data that Professor Shiller has collected.

However, to the extent that U.S. economic growth prospects have slowed – as we believe – then real interest rates will remain lower than they were on average during the second half of the last century. And, with inflation expectations and inflation still tame (see our earlier blog post), long-term nominal interest rates likely will be lower going forward than they were in previous interest rate cycles. To equalize risk-adjusted expected returns on stocks and bonds, the current ratio of stock prices to earnings should be higher than in the past, as the 25-year lagged average CAPE suggests.

So, what should we conclude? Should we buy, or should we sell? The honest answer is that we don’t know. What we can say is that the rate of return on stocks purchased today likely will be lower over the longer term than it has been in the past – especially the recent past. But if you’re looking for asset price bubbles, you’re more likely to find them in bond markets than in today’s stock market.