Depending on the measure used, equities can be cheap, expensive, or anywhere in between.
FROM BARRON'S | JULY 23, 2014
Barely a day goes by of late without someone in the financial media announcing that equities are overvalued and primed for a fall. The most popular article on one of the most popular financial websites recently blared "U.S. stocks will be very disappointing for 10 years." The argument? That on multiple gauges, the current valuation of the market is higher than it was during the vast majority of market peaks in the past.
That is hardly a rare argument. But as ubiquitous as such analyses are, it still begs the question: is it true?
The short answer is: stocks might be pricy but past patterns are not nearly clear enough to warrant certainty about whether market valuations have reached a peak.
It is, of course, impossible to argue against the possibility or indeed the probability of some sort of equity correction. But a sell-off of 10% on one or more indices is not the same as a decade of underperformance, or a looming crash. Current prognostications of market overvaluation frequently ignore how other asset classes are behaving, which is a crucial flaw. Equities are trading today in an environment of radically low yields on fixed income instruments, and that matters greatly in assessing where stocks go from here.
Make no mistake: no one knows with any statistical certainty how stocks or any aspect of the financial system will perform in the years ahead. These are not mechanical systems, and patterns are ever in flux. But the belief that there is a mean valuation of stocks that must inevitably be reverted to overlooks the fact that valuations swing wildly over the years and are always in the context of how other available assets and investments are performing. In short, past valuation analysis is hardly as conclusive as many contend, and many today are also cherry-picking their numbers to make their cases for or against equities.
Market Valuations in Perspective
There are many ways to assess market valuation: the price of the stock to sales, relative to revenue, relative to dividend yield, relative to free cash flow, and of course, relative to earnings. Over the past decades, the price-to-earnings (P/E) valuation has become the preferred reference point. However, many professionals and strategists have tried to distinguish their arguments by coming up with more obscure, arcane, or simply new valuation methods. That is fine and well, as long as one recognizes that all of these metrics are simply different ways to assess whether investors are paying reasonably, excessively, or cheaply for stocks.
It's often repeated that the long-term average of the S&P 500's price-to-earnings ratio is about 15.
According to data from Bespoke Investment Group, the average since 1929 is 15.25; the average since 1989 is 18.90; and the 10-year average since 2004 is 16.95. These are trailing averages, that is, the P/E ratio based on stock prices relative to the prior 12-months of earnings. The current trailing ratio is about 17 times earnings.
This picture says that markets aren't cheap or notably expensive. But averages are just that, and they blend together years when the P/E ratio was very high (such as 1998 and 1999 when it was nearly 30 times earnings), and years when it was much lower (in 2006, it was just below 16 times earnings; in 1980, it was below 10). An average hides such variations: when Bill Gates walks into a bar, everyone in the room on average becomes a millionaire. That is statistically true, but also relatively meaningless.
So on average, today's equities (or at least the S&P 500) are somewhere in the middle between cheap and expensive based on P/E ratios. The picture changes somewhat if we use forward estimates rather than trailing estimates. The argument for trailing estimates is that they are based on real numbers, whereas forward ratios are based on expectations. However, we buy stocks based on what we expect them to do, and not based on what we think companies will deliver in terms of financial performance. You don't buy IBM because they once dominated the typewriter market with the Selectric; you don't buy General Electric because of their dominant franchise in incandescent light bulbs. In short, you don't buy a stock based on what the company has done; you buy a stock based on what you believe it will do. Or at least that is how we ought to invest.
In that sense, forward P/E ratios are a better measure of valuation, because most investors buy stocks on expected earnings. On that metric, stocks are about where they were in the early to mid-1990s.
There is also the widely used index developed by Robert Shiller of Yale that takes a 10-year average.
On that score, stocks today are well above their long-term average but well below the last major stock bubble in the late 1990s.
Even a cursory reading of these historical patterns would suggest that there is nothing conclusive about current valuations if you believe that past patterns are a reliable guide. Certainly, there is nothing clear enough in these patterns to lead to such levels of certainty that so many express when they declare, flatly, that markets are overvalued.
Above all, don't forget the context
No financial asset exists in a vacuum. The current value of all assets is based on some assessment of future growth and cash flow. Every portfolio offers some assortment of assets, from stocks to bonds to cash to less liquid investments such as real estate. The price of each is determined not just in reference to itself but in reference to other possible investments.
The primary alternative to equities has traditionally been bonds. Because equities are seen as riskier and more volatile, they have always commanded a "risk premium" and hence cost more than bonds on a valuation basis.
Today, what is unusual isn't the price of equities but instead the price and yield of bonds, of all flavors. The yield on the 10-year U.S. Treasury has been lower in the past few years than at any point since the 1950s. German and other European bonds recently hit lows not seen since the early 1800s and the time of Napoleon! With yields that low, stocks are the only place to go for investors looking for liquid investments that have any hope of generating long-term returns above inflation.
For many years, it was widely accepted that there was a valuation relationship between stocks and bonds. The so-called "Fed model" held that if the earnings yield of the S&P 500 was higher than the yield of the 10-year U.S. Treasury, then equities were the rational choice for investors for better returns. This model, even using a variety of P/E methodologies, shows that stocks have been cheap relative to bonds since the early 2000s, and extremely so since 2009. Including today.
It's vital in all of this to recognize that the equity crash of 2008–2009 that ended in March 2009 was not predicated on a stock market bubble. It was purely a byproduct of a housing bubble and, more crucially, a derivative bubble that created a financial crisis that forced the sale of liquid assets such as stocks to meet obligations triggered by that crisis. Stocks crashed in 2000 because stock trading had become a bubble. Stocks crashed in 2009 because housing and derivatives had become a bubble.
If you believe that all financial assets are at risk because of the global system of easy money generated by central banks and by too much sovereign debt piling up, then yes, stocks will suffer dramatically if the global financial system is built on untenable foundations. The presence of risk, however, is not proof of a coming decade of stagnation and crash. There are few places for investors and advisors requiring liquid investments to find the potential for yield and appreciation. Relative to bonds and relative to past patterns, stocks are not evidently cheap or expensive. They are, however, unequivocally able to generate growth based on the ability of underlying companies to grow. That may not ensure long-term gains, but if long-term gains are to be had, stocks are where such gains most likely will be found.