Over relatively short periods of time, stock-price fluctuations are basically random. Although we can assign all sorts of explanations to the market’s gyrations after the fact--one day, bad economic news is bad because it means growth will slow; the next day, bad news is really good because it means the Fed’s easy money policies will continue--the truth is that most of the time no one really knows why stock prices move from one day to the next.
Luckily, investing isn’t about the short term. We invest to meet long-term goals: securing our retirement, sending a kid to college, maximizing a philanthropic contribution. And in the long run, stock-price movements are not random. In fact, they very rationally follow the fundamentals. To repeat one of my favorite Ben Graham quotes: In the short run, the market is a voting machine, but in the long run, it is a weighing machine.
Dividends and Earnings Growth Drive Returns
At the most basic level, there are three sources of returns from common stocks: dividends, earnings growth, and changes in price/earnings ratios. This works for individual stocks as well as for the market as a whole. Total return is given by:
Total Return = Dividend Return + Earnings Growth + % Change in P/E + (Earnings Growth)*(% Change in P/E).
The table below shows total returns for the S&P 500 and predecessor indexes over various historically meaningful time periods, along with a breakdown of the sources of return (I lumped the last "interaction term" in with the P/E change). These figures are all in real terms--that is, they are adjusted for inflation.
Over the very long run, the market has delivered a real total return of around 6.5% per year (6.6% in the postwar period). However, returns have varied greatly by time period. In both 1930-42 (which includes the Great Depression) and 1970-79 (a period of rapid inflation), the real return on stocks was negative 2.2% per year. The most recent decade-plus hasn’t been kind to investors, either, with the bursting of the technology and housing bubbles and the global financial crisis leaving investors with a negative 0.2% real annual return in 2000-12.
When deconstructing returns in this way, it becomes apparent that the P/E multiple has been the key swing factor between good decades and bad. Great bull markets occur when the market starts out very undervalued and ends up very overvalued. For example, the 1980-99 bull market started with an average P/E multiple of 6.9 in 1979 and ended with a P/E of 27.5 in 1999. The opposite occurs during bear markets. Both the 1930-42 and 1970-79 periods began with a P/E in the midteens and ended with a P/E in the single digits. The poor returns in 2000-12 were almost entirely driven by a contracting P/E multiple, while the impressive returns so far in 2013 were caused by an expanding P/E multiple.
Realistic Long-Run Return Expectations
The market’s P/E multiple expansion since 1871 was not a significant driver of returns. Instead, returns over the very long run closely track dividend yields and real earnings growth. Historically, dividend returns accounted for the majority of total returns. However, as payout ratios have fallen--and firms have put more cash to work through internal reinvestment, acquisitions, and share repurchases--this mix has shifted. Since 1946, real earnings have grown 3% annually, and dividends have contributed 3.6% to total return. In the 2000s, average dividend yields fell to unprecedented lows around 2%, where they stand currently.
Given the importance of dividends to historical returns, one might reasonably wonder what this means going forward. Finance theory would tell us that a firm’s choice of a payout ratio shouldn’t necessarily matter to long-run returns. Lower dividends mean more retained earnings, which will translate to faster future earnings growth. However, the historical data seems to contradict this theory. While lower payout ratios in the past 40 years have resulted in accelerated growth, it hasn’t been enough to make up for lower yields. We can estimate sustainable total return as earnings growth plus the dividend return (that is, ignoring P/E changes). This measure, shown in the second column of the table, dropped from 7.9% per year in 1943-69, to 6.6% in the 1970s, to 5.5% in the 1980-99 bull market, and to just 4% in 2000-12.
This indicates that incremental retained earnings have not been used wisely, on average--management teams may have destroyed shareholder value by overpaying for acquisitions, making poor internal investments, or repurchasing shares at inopportune times. S&P 500 firms might be able to raise their payout ratios (from a 2012 level of 36%) without sacrificing much in the way of earnings growth. The recent trend in sustainable returns implies that real returns are likely to be lower going forward than they had been in the past. Dividend yields and real earnings growth haven’t been enough to support a 6.5% real return on stocks for more than 30 years. Based on recent trends, real returns of 4% to 5.5% seem more realistic. Some combination of higher payout ratios and better returns on retained earnings would likely be necessary to achieve the high end of this range
A version of this article appeared in the August 2013 edition of Morningstar StockInvestor.
Matthew Coffina is the editor of the Morningstar StockInvestor newsletter.
