An underestimate: Three different approaches to estimating outlook on equities

We believe that the prevailing long-term outlook for equities is too pessimistic.

Given the current low-interest-rate environment, it is not unusual to hear investors comment: “In this environment, I would be happy with 4% to 5% return from equities.”

At Janus, we take three different approaches to estimating long-term equity returns, and they all point to meaningfully higher estimates than the 4% to 5% range assumed by many in the marketplace.

Even Bill Gross in his widely read Investment Outlook commented in August 2012:

“Together then, a presumed 2% return for bonds and an historically low percentage nominal return for stocks — call it 4% — when combined in a diversified portfolio produce a nominal return of 3% and an expected inflation-adjusted return near zero. The Siegel constant of 6.6% real appreciation, therefore, is a historical freak, a mutation likely never to be seen again as far as we mortals are concerned.”

The “Siegel constant” represents the annualized long-term inflation-adjusted return for U.S. equities since 1912 as estimated by Jeremy Siegel of the University of Pennsylvania. For the past five years, the S&P 500 index has returned 1.5% per year and for the past 10 years, 7.1%. Given the recent history, the sub-2% real GDP growth for the U.S., and the low-interest-rate environment, many investors believe that the historical 6.6% real return for equities is likely beyond the realm of possibility; in fact, they expect real return for equities to be in the range of 1.6% to 2.6%, nominal return of 4% to 5% minus the market implied inflation rate of 2.4% for the next 10 years.

From a top-down and supply-side perspective, this line of reasoning regarding equity returns appears sensible. However, from bottom-up, historical, and investor demand perspectives, 4% to 5% projected nominal equity return for the foreseeable future seems much too low based on three distinct estimation approaches:

1. Corporate revenues and earnings from a bottom-up perspective


2. History of equity returns


3. Equity returns from an investor demand perspective

Taking a building-block approach, we can estimate the real return from equities in the following fashion:

Real return from equities = Dividend yield + real EPS growth rate +/- P/E expansion or contraction

If we unrealistically assume zero return from dividend yield and no price-to-earnings ratio expansion or contraction, then the real return from equities reverts to the long-term real earnings-per-share growth rate. Further, if we assume that the latter converges with the real GDP growth rate over the long term, then the real return from equities has to equal the long-term real GDP growth rate.

This approach begs the question: What is the appropriate real GDP growth rate to use as a proxy for the long-term real EPS growth rate? In our opinion, and from a bottom-up perspective, the U.S. real GDP growth rate represents a rather poor proxy for the real earnings growth of U.S. companies.

Returning to the building-block approach, consider a more realistic scenario in which we substitute the U.S. real GDP growth rate with global real GDP rate as a proxy for real corporate earnings for the next 10 years:

1. Beginning dividend yield = 2%


2. Real corporate earnings growth for the next 10 years = 3.5%


3. Return from P/E expansion or contraction = zero

Based on these assumptions, the supplied real return on equities for the next 10 years approximates 5.5%. When we add 2.4% — the market implied breakeven inflation rate on the 10-year Treasury bonds — we arrive at an expected nominal return on equities of 7.9%, meaningfully higher than the 4% to 5% nominal returns on equities embraced by many in the marketplace.

It is instructional to compare estimated future equity returns with the history of equity returns. As shown below, the Siegel constant of 6.6% for the past 100 years is almost spot-on with the realized returns for the S&P 500 index from 1926 to 2011.

S&P 500 index real rate of return (1926-2011)

Siegel constant: 6.6%

S&P 500 index (1926-2011):

Historical compound rate of return: 9.8%

Historical change in CPI: 3.0%

Historical real return: 6.8%

Data presented reflect past performance, which is no guarantee of future results.

If we assume that the long-term realized real returns for the S&P 500 index are the best unbiased estimate of returns for the next 10 years, then we should expect 9.2% in nominal returns from the S&P 500 index.

In the first section, we approached the estimation of equity returns from the supply perspective. Now, we approach the estimation of equity return from the demand perspective; that is, what kind of risk premium should investors demand from equities? To be more precise, what should equities' risk premium be over long-term corporate bonds?

Equity risk premium over corporate bonds (1926-2011)

Historical compound rate of return

S&P 500 index: 9.8%

Long-term corporate bonds: 6.0%

Historical equity risk premium over long-term corporate bonds: 3.8%

Between 1926 and 2011, the realized (ex-post) equity risk premium over long-term corporate bonds was about 3.8%. As of October 2012, the yield-to-worst on Barclays Aggregate Long Corporate and High Yield indexes stood at 4.3% and 6.5%, respectively. Therefore, when we combine the long-term historical equity risk premium with the yield-to-worst on Barclays Aggregate Long Corporate index we arrive at an expected nominal equity return of 8.1%.

Although possible on an episodic and temporary basis, we believe that over the long term, the expected nominal return on equities cannot be lower than 6.5%, the yield-to-worst on the Barclays Aggregate High Yield index, because even high-yield bonds rank senior to equities in the corporate capital structure. Moreover, since equities are deemed riskier than high-yield bonds, the former must have higher expected returns than the latter; otherwise, there would be no incentive for investors to hold equities over high-yield bonds.

We took several approaches to estimating future equity returns for the next 10 years, and all three approaches resulted in estimated nominal equity returns that are meaningfully higher than the nominal 4% to 5% return forecasted by many in the marketplace.

The market might not achieve Siegel's optimistic constant of 6.6% from equities, however, the 1.6% to 2.6% real return on equities forecasted by many in the marketplace seems equally unlikely. The true outcome probably lies somewhere in the middle.

Jonathan Coleman, CFA is the co-chief investment officer, and Soonyong Park, CFA, CPA, is an institutional client strategist for Janus Capital Group.