Hussman Funds


Market Comment Archive

Investment Research & Insight Archive

Long-Term Stock Returns: How Low Will They Go?

The Debate between Rob Arnott and Jeremy Siegel

William Hester, CFA
April 2004
All rights reserved and actively enforced.

One of the most spirited debates at the New Directions for Portfolio Management conference centered on expected equity returns. It was an engaging discussion because there was no middle ground. Differences were not split by nuance, but by wide-open spaces. Stocks are either going to offer ample returns relative to bond investments, or barely any at all.

Each side of the argument had an impressive advocate. Jeremy Siegel, a professor at the University of Pennsylvania and the author of the best-selling Stocks for the Long Run , argued that although equity returns will be lower, their lofty advantage over bonds would continue.

Rob Arnott, a hedge fund manager and the editor of the Financial Analysts Journal, argued that because past equity performance was predominantly fueled by P/E multiple expansion and dividend payments, there's no clear reason why stocks should continue their dominance 1 .

To develop their argument they each covered three topics: stock valuations, the equity risk premium, and the outlook regarding the demand for stocks.


Both earnings and P/E multiples determine the level of stock indexes. In good form, the debaters disagreed on both components.

Earnings are vastly overstated, says Arnott. He prefers to use reported earnings, which include all charges except for those that are truly extraordinary 2 . But even this conservative measure is inflated, he says. At their peak in 2000, reported earnings for the S&P 500 topped out at $37 a share, not the often quoted $54 a share.

Companies are playing games to make their earnings appear higher. If stock options were expensed profits would fall by $6. Earnings restatements by the likes of Enron and Worldcom would lower that figure by another $3. If companies assumed their pension assets (which includes a mix of both stocks and bonds) would grow at 6 percent, and not the highly optimistic 9.5 percent they are currently forecasting, $8 more of earnings would be subtracted. (When a higher rate of return on pension assets is assumed companies can set less money aside, boosting earnings. But if their assumptions turn out wrong, a larger portion of profits will be needed in the future to meet these obligations.)

Even though the year 2000 figure represents peak earnings, Arnott feels it is fair to use $37 a share as a normalized earnings figure for the present as well. So with the S&P recently trading at 1110, and if normalized earnings are $37 a share, the market's P/E multiple is currently 30.

Arnott's Valuation Calculation


S&P 500 reported earnings

in calendar year 2000


If pension return assumptions

were 6%, not 9.5%


If management stock options

were expensed


FY2000 earnings restatements

(so far.....)


Calendar year 2000 reported EPS

net of accounting biases


S&P 500


P/E Multiple


Arnott doesn't venture a guess on where P/E multiple's are headed. But he did say that he can't understand why investors are willing to accept the idea of cyclical bear markets (which last in the neighborhood of two and half years) but not secular ones. "The 1929 and the late 1960's tops were followed by secular bear markets. Why isn't it accepted that a secular bear market could follow the greatest bull market of all time?"

When will it end? "Secular bear markets end when valuations get cheap."

Siegel thinks that a proper earnings figure falls somewhere between reported earnings and operating earnings (generally, a more favorable profit figure with a greater number of expenses ignored).

Instead of looking back as Arnott does, Siegel looks forward. For 2004, strategists forecast that S&P 500 operating earnings will be $60.74. He thinks that it's fair to deduct the cost of options and pension expenses, which amounts to $5.10, according to Standard & Poor's. (This figure is lower than Arnott's because S&P calculates pension expense using the rates of return assumed by the companies in the index. The pension and option expense calculations also cover different time periods).

That would bring his earnings figure to $55.64, and a P/E multiple of 19.95.

Siegel's Valuation Calculation


S&P 500 estimated 2004 top

down operating earnings


Pension and Option



Adjusted estimated 2004

operating EPS


S&P 500


P/E Multiple


Siegel said that a P/E multiple of 20 feels about right. He disagrees that this ratio will fall to previous bear market lows. "People argue that valuations need to fall to their averages of the last 100 years. But those people first have to convince me that the economy and the financial markets haven't changed tremendously in that time.

"50 years ago the dividend yield of blue chip companies was 7 and 8 percent, and those dividends were well-covered, but investors still didn't want them. They ran to 2% Treasury notes and 1% passbook savings. There is no way that we're getting back to that."

Index investing has made it possible to obtain diversification at a low cost, which makes investor portfolios less risky. He also feels that the public now understands that stocks are the best investment for the long term. "I would love to see stocks at a P/E of 8 or 10, I would back up a truck to buy them, it's just not something we're going to see."

The Equity Risk Premium

Siegel expects stocks to outdo bonds by about 3% a year. Arnott thinks equities will be lucky to beat them at all.

Forecasting the equity premium is a competitive activity in the academic and practitioner world. The bulls have history on their side. Whether you look at the last 50, 100, or 200 years, real stock returns have been surprisingly constant, registering about a 7 percent annualized gain after inflation.

The amount by which they beat real bond returns have averaged between 4.5 percent and 5.5 percent over the last 75 years, according to Siegel. The optimists wonder why that should change.

Equity risk premium bears argue that so much of these past stock returns have been driven by increases in earnings and dividend multiples, it would be nearly impossible for a further expansion in these to contribute to future returns.

So that saddles earnings growth with the job of providing additional return. Which leads us to the root of why Arnott and Siegel's equity return premium forecasts differ by so much.

Corporations will retain more of their profits, which will boost future real growth up to 3.5 percent a year, says Siegel. Arnott argues the opposite, that low dividend payout ratios precede periods of low earnings growth. With the dividend yield at near-historic lows, he estimates that earnings will grow at 1.1 percent, after inflation.

Equity Risk Premium Forecasts




Current Yield



Expected Real Growth



P/E Multiple Expansion



Cumulative Real Return



Expected Real Yield



Change in Bond Valuation



Equity Risk Premium



The tricky part of this discussion is that both Arnott and Siegel have data to support their conflicting views. In my opinion, there was not enough time devoted to this contradiction in either of their discussions at the conference.

This is really the core of the issue. If you believe real earnings per share can grow at 3.5 percent, outpacing the long-term average of about 2 percent, then current valuation levels wouldn't look so extreme and the equity risk premium could continue to be favorable.

If you expect real earnings growth of 1 percent then the other side of the debate looks sound. I'll outline each of their positions as best as possible, both from their prepared comments at the conference and from earlier published works.

Siegel argues that today's low dividend yields can be explained by an increase in prices and a reduction in the amount of dividends companies pay as a percentage of their earnings. This reduction in the payout ratio should lead to faster earnings growth.

Siegel compares the change in the long-term average dividend yield with the change in long-term average earnings growth. By his calculations, they move inversely. The dividend yield averaged 5.07 percent from 1871 to 1945 and averaged 3.45 percent from 1946 to 2004. While the average dividend yield dropped 162 basis points between the two periods, average earnings growth increased 222 basis points.

Bringing this up-to-date, Siegel calculates that there's been a change in the dividend yield of 185 basis points (taking the 1946-2004 period average of 3.45 percent and subtracting the current dividend yield of 1.6 percent). Adding that to the recent long-term real earnings growth of 2.88 percent (for the 1946-2004 period), profits could grow at a real rate of 4.73 percent.

Siegel's Calculation of Expected Earnings Growth

Historical Average Dividend (1946-2004)


Current Dividend Yield


Change in Dividend Yield



Earnings Growth (1946-2004)


Potential Real Earnings Growth


Siegel admits that this is an optimistic scenario. But assuming that just a portion of the drop in dividend payments fuels earnings growth, then attractive returns are still achievable. If half of the reduction in the dividend yields boosts earnings growth, that would result in a 5.4 percent real return (leaving roughly a 3 percent equity return premium, his current forecast).

"You can't use historical earnings growth with such a low current dividend yield," says Siegel.

Arnott, along with fellow hedge fund manager Cliff Asness, found different results to the same question. They show that where dividend yields go, earnings growth follows 3 . That is, a high dividend yield has forecasted high earnings growth over subsequent 10-year time frames. Periods of low dividend yields were followed by slower earnings growth.

Surprisingly, in periods of very low dividend yields the subsequent 10-year period real earnings growth was actually negative. "In general, when starting from very low payout ratios, the equity market has delivered dismal real earnings growth over the next decade; growth has actually fallen 0.4 percent a year on average - ranging from a worst case of truly terrible -3.4 percent compounded annual real earnings for the next 10 years to a best case of only 3.2 percent real growth a year over the next decade".

The behavior of corporate decision makers helps explain these results, say Arnott and Asness. Corporate managers dislike cutting dividends. If executives see poor times ahead they may keep the portion of earnings they pay out low. Good times ahead might bring the opposite action.

Managers may also pursue 'empire building' when they are sitting on piles of undistributed earnings, say the authors. With their hoards of cash they purchase buildings and businesses, but in the end create little value (or destroy it). This explanation fits nicely with the activity of the late 1990's.

In a question-and-answer session at the conference, Arnott did say that it was likely that their results differed from Siegel's because he and Asness looked at shorter intervals of time as opposed to changes in long-term averages.

In addition to arguing that low yields precede periods of slower earnings growth, Arnott also pointed out the role dividends have played in the total return of stocks. He summed up this argument with an example: If relatives of yours invested $100 in the stock market in 1800, you'd be sitting on $500 million today. Inflation adjusted, that's still $25 million. If they chose not to reinvest dividends, you'd have $1,500.

With current dividends yields historically low, expectations for slower earnings growth ahead, and lofty multiples, Arnott doubts a continuation of the past equity premium.

Supply and Demand

A message that was repeated throughout the conference is that one of the most critical situations facing the U.S. is its aging population. People are certainly living longer; the average life expectancy of men and women has jumped from 65 in 1940 to about 77 now. In 30 years, that average is projected to rise above 80.

In addition to the aging population, the group of people about to retire is much larger then the labor force that follows it. This will send the retiree/worker ratio soaring from .2 to almost .4 over the next quarter century.

This bulge in the population is still primarily made up of savers. But when these baby boomers want to turn their investments into Winnebago's and Palm Beach condos, who will buy their shares?

Siegel is unconcerned with the demographic situation 4 . He agrees that the size of the U.S. work force will not be sufficient to offset the aging population. But he believes that China (and, eventually other developing nations) will grow quickly enough to provide the goods and services the U.S. needs. "With the dollars they receive from selling us goods, they will be buying not only our government bonds, but corporate bonds and stocks and even real estate."

Arnott looks skeptically at this argument. Considering the current rate of purchases by China, they would need "to buy even more of our equities and bonds than they already do. That's not likely".

Arnott's view is that when the baby boomers begin to cash in their portfolios, stocks will be among the first asset classes to be sold, putting further pressure on their long-term returns.

Why It Matters

This debate is tremendously important. Real returns of 3 percent, compared to the historical real return of 7 percent, will make a tremendous difference in the saving patterns needed to meet future liabilities.

Take saving for retirement. Even if corporations help their workers save through 401(k) plans, low returns could undermine the effort. Assume a thirty-five year old employee, who makes $35,000 a year in salary, and expects real growth in her compensation of 2 percent a year. She invests 6 percent of that in her 401(k) plan and half of that is matched by her company. At retirement she rolls this money into an annuity paying the same investment rate she received during her working years.

If a 7 percent real return were assumed, she would retire with a real sum (i.e. today's dollars) of $394,541. Her annuity would pay out $37,241 a year, which would be 60 percent of her ending real salary ($62,154).

If 3 percent real returns were assumed, she would retire with a real sum of $198,408. Here her annuity would pay out $13,336 a year, covering only 20 percent of her ending real salary.

Arnott wasn't all doom and gloom. Though the investment management industry will likely end up with low returns as a group, there are opportunities for select investors to earn higher returns. He suggested that investors consider a range of asset classes and alternative markets, looking for those that are priced to offer better returns. Investors should seek out managers that are skilled enough to generate consistent risk adjusted returns. Lastly, he advocated investments in vehicles that have the ability to boost returns by using leverage wisely.


(Click here for the first article in this series on the 2004 AIMR conference )

1 As its new editor, Mr. Arnott has been shaking up the usually staid Financial Analysts Journal. His latest "Editors Corner" is titled Is Our Industry Intellectually Lazy? and is available at

2 See The Market's Reality Show - A Guide to Earnings Season

3 Arnott, Rob and Cliff Assness. Surprise! Higher Dividends Mean Higher Earnings Growth. Financial Analysts Journal. January/February 2003.

4 For an opinion piece by Siegel discussing this view (that was published in the January 26,2004 Wall Street Journal), see



The foregoing comments represent the general investment analysis and economic views of the Advisor, and are provided solely for the purpose of information, instruction and discourse.

Prospectuses for the Hussman Strategic Growth Fund, the Hussman Strategic Total Return Fund, the Hussman Strategic International Fund, and the Hussman Strategic Dividend Value Fund, as well as Fund reports and other information, are available by clicking "The Funds" menu button from any page of this website.

Estimates of prospective return and risk for equities, bonds, and other financial markets are forward-looking statements based the analysis and reasonable beliefs of Hussman Strategic Advisors. They are not a guarantee of future performance, and are not indicative of the prospective returns of any of the Hussman Funds. Actual returns may differ substantially from the estimates provided. Estimates of prospective long-term returns for the S&P 500 reflect our standard valuation methodology, focusing on the relationship between current market prices and earnings, dividends and other fundamentals, adjusted for variability over the economic cycle (see for example Investment, Speculation, Valuation, and Tinker Bell, The Likely Range of Market Returns in the Coming Decade and Valuing the S&P 500 Using Forward Operating Earnings ).

For more information about investing in the Hussman Funds, please call us at
1-800-HUSSMAN (1-800-487-7626)
513-326-3551 outside the United States

Site and site contents © copyright Hussman Funds. Brief quotations including attribution and a direct link to this site ( are authorized. All other rights reserved and actively enforced. Extensive or unattributed reproduction of text or research findings are violations of copyright law.

Site design by 1WebsiteDesigners.