Estimating future stock returns

What are the long-term expected returns for stocks? That’s a fundamental question every investor needs to consider when deciding on an appropriate asset allocation. Unfortunately, it’s not a question anyone can answer with certainty—though there’s no shortage of gurus with opinions.

0

by Dan Bortolotti
February 25th, 2013

Online only.

0

What are the long-term expected returns for stocks? That’s a fundamental question every investor needs to consider when deciding on an appropriate asset allocation. Unfortunately, it’s not a question anyone can answer with certainty—though there’s no shortage of gurus with opinions.

In a paper published last October, researchers at Vanguard examined 15 commonly used methods for forecasting stock returns to see how much predictive power they would have had in the past. These included price-to-earnings (P/E) ratios, dividend yield, earnings growth, economic fundamentals, and recent stock returns. And just for fun, they threw in a red herring: the trailing 10-year average rainfall in the US.

For each variable, the researchers set out to find whether it would have helped predict US stock returns during the 10 years that followed. Suppose, for example, you measured the trailing one-year dividend yield on stocks in 1950. How useful would that variable have been in explaining inflation-adjusted returns from 1951 through 1960? They repeated this for all the factors, in all rolling 10-year periods starting in 1926.

Turns out about half the variables were entirely useless: “Many popular signals have had a lower correlation with the future real return than rainfall,” the researchers wrote. The biggest flops included consensus GDP growth, consensus earnings growth, and corporate profit margins. The variables most effective at predicting future returns turned out to be trailing one-year P/E and cyclically adjusted trailing 10-year P/E, commonly known as the Shiller CAPE ratio. This latter variable explained about 43% of the subsequent decade’s stock returns.

Peeking under the CAPE

The Shiller CAPE ratio is named for Robert Shiller, professor of economics at Yale and author of several books including Irrational Exuberance. Instead of trailing one-year earnings, Shiller’s ratio uses the average annual earnings of companies over the past 10 years, adjusted for inflation. The idea is to smooth out the numbers over an entire business cycle.

The practical difficulty with using Shiller CAPE is the data are not easy to find unless you happen to have a Bloomberg terminal. But for a recent post on his blog, Justin Bender at PWL Capital (who knows a guy with a Bloomberg terminal) tracked down these numbers for Canadian, US, international and emerging markets. As of December, the data predict future real returns (after inflation) of 5.7% for Canada, 4.7% for the US, and 6.6% for international and emerging markets.

Following the advice of Larry Swedroe, Justin then adjusted the figures downward by 1% to account for “slippage” in earnings yield. Assuming an equity portfolio that’s one-third Canadian, one-third US, and one-third international/emerging, they now project a real return of 4.7%. If you assume 2% inflation going forward, that’s a nominal return of 6.7%.

For expected bond returns, it’s reasonable to simply look at the current yield to maturity of the DEX Universe Bond Index, and Justin used an estimate of 2.25% (nominal). Combining these two projections, here are the expected nominal returns for portfolios with various asset mixes:

Stocks 80% 70% 60% 50% 40% 30% 20%
Bonds 20% 30% 40% 50% 60% 70% 80%
Expected return 5.8% 5.4% 4.9% 4.5% 4.0% 3.6% 3.1%

Remember, these figures do not include costs. If you’re using ETFs, you likely need to subtract at least 0.50% for fund MERs, trading commissions, bid-ask spreads, withholding taxes, cash drag and other incidentals.

A framework for your plan

So what’s the takeaway message from this little exercise? Let’s be clear: any forecast of expected returns is only so useful. After all, even if Shiller CAPE was the most reliable variable in the Vanguard research, it still only explained about 43% of subsequent returns. “We feel it is important to stress that even conditioning on initial P/E ratios leaves approximately 60% of the historical variation in long-term real returns unexplained,” said the authors of the paper.

But these figures can be useful for managing expectations. In our DIY Investor Service, it’s not unusual for  clients to say they expect returns of 6% to 7% from a balanced portfolio. Current Shiller CAPE ratios and bond yields suggest that’s unlikely even from a portfolio of 80% stocks. Investors who are comfortable with a traditional 60/40 portfolio will likely need to save more (or spend less) than they bargained for.

The Vanguard researchers go out of their way to stress that long-term expected returns are a moving target, not “point forecasts.” They can be helpful as a framework for planning, but it’s not like you can do this at age 35 and expect your projections to be valid until you retire at 65. You need to revisit the numbers every few years to make any necessary course corrections to ensure you’re taking as much risk as you need, but no more.

Leave a comment

Your email address will not be published. Required fields are marked *