Can We Really Trust What We Think We Know About Markets?

Can we really trust what we think we know about markets? Even the professionals shouldn't be so sure about the basic precepts on which they rely. For example, in a research report recently highlighted on RealClearMarkets.com, a pair of portfolio managers wrote,

"...we do know that long-term (i.e. sustainable) stock market returns are anchored to corporate earnings growth. S&P 500 returns will not exceed earnings growth indefinitely."

That sounds pretty basic, right? It seems so intuitively obvious, most readers probably accepted it without blinking. Yet, logic and history indicate this seemingly fundamental belief is wrong. Not just a little off base, but logically flawed and historically backward.

From a logical perspective, a simple example illustrates that earnings growth is not required for positive stock returns. Imagine a profitable company that has stable earnings (i.e. 0% earnings growth) and pays out all its earnings as dividends each year. If you paid a 10x PE multiple for that stock, and both earnings and the multiple remained stable over time, you could get a 10% annual return from the dividend. Simply put, in this example, 0% earnings growth can support 10% stock returns, indefinitely.

I believe people often forget (or never consider) that a stock's return is ultimately a return on capital invested in the issuer, or rather, the cost of equity capital for the issuer. How much that equity capital costs, and whether a company can afford it, is not necessarily determined or limited by the earnings growth rate. However, for many people, the cost of equity capital is a fuzzy concept because, unlike the cost of debt capital, it is not typically directly quantifiable (except as an ex post observation of stock returns).

Yet the example above shows the cost of capital for equity and debt are not too dissimilar. The hypothetical company's stock would act very much like a bond. A bond's yield is also a return on capital (lent to the issuer). The coupon on a traditional bond-loosely analogous to a stock's earnings per share-has zero growth. Yet traditional bonds almost always offer a positive return if held to maturity. The exact rate of return for a particular investor may depend on many factors, but as long as the purchase was made with a positive "yield-to-maturity" and the issuer didn't default, then the bond could deliver positive returns on a zero-growth coupon stream.

Back to stocks, for a healthy, growing company (or an index of mostly growing companies like the S&P 500), long-term earnings growth could even be viewed more as a floor for long-term returns than as a ceiling. If you assume that P/E multiples are generally range-bound and will remain relatively stable over the long-term (ironically, this is the assumption that may lead some people to think returns can't exceed earnings growth indefinitely), then price should grow in line with earnings. That's simple algebra. For P/E to remain constant, "P" must be growing at the same rate as "E." In other words, the price appreciation would equal earnings growth. But if the stock or stocks in question also pay dividends, then the return for owning that stock would be a combination of the price appreciation and the dividend yield. For an index like the S&P 500, as long as some stocks paid dividends, then stable multiples would imply the index's total return must exceed earnings growth over time.

History supports this. Based on annual earnings and total return data from Global Financial Data, Inc., from 1926 through 2013, the S&P provided a compound annual return of 9.9% while earnings grew at only 5.2% per year.* One may quibble about pragmatic definitions of "sustainable," and "indefinite," but the 87-year history of the S&P is more time than any of us likely get to manage our personal investments. Even over shorter, 25-year periods (which most investors might consider long term) the S&P's returns always outpaced earnings growth. From 1926 through 2013, using calendar year earnings and returns, there isn't a single 25-year period when returns failed to outpace earnings growth. Not one.

That isn't to say that earnings growth cannot impact stock returns, especially in the short-term. But there is no reason to believe that earnings growth rates must prove a limit for stock returns over the long-term. Expected earnings growth rates are just one factor that may play into the cost of capital that returns ultimately represent.

I would guess there is something in human nature that makes us misconstrue seemingly basic market principles. It can happen to financial professionals - portfolio managers in charge of hundreds of millions of dollars - and individual investors alike. None of us is immune, so when it comes to capital markets, we should all re-examine our guiding precepts from time to time.