The Fundamental Case for the 20,000 Dow

Seth J. Masters

While some people deem stocks expensive relative to 10-year trailing earnings, we take a forward-looking approach. It starts with the premise that the stock market is not a casino and stock prices are not pulled out of thin air: they reflect the intrinsic value of companies’ future earnings.

Let’s start with basics. Stocks represent an ownership claim on a share of company earnings. Hence, stock prices reflect (imperfectly, of course) the value of companies’ current earnings and potential earnings growth. In computing the expected returns for stocks, what matters is the starting price, earnings, dividends (the portion of earnings distributed to shareholders), earnings growth and changes in P/E. As you might expect, low starting prices, high earnings and dividends, high growth, and P/E expansion are all good for future stock returns.

The models we use when investing are complex, but a simple argument makes the point. The expected return for a Treasury bond held to maturity is equal to its yield. Similarly, the expected return for a stock equals its earnings per share (EPS) divided by its price—its earnings yield—if the company has no growth prospects and therefore returns all of its earnings to shareholders. If the company does have growth prospects, it would retain some of its earnings to fund growth. In that case, the expected return equals the dividend yield plus dividend growth. If the company pays out a constant share of earnings as dividends, dividend growth equals earnings growth.

Let’s apply this framework to the S&P 500 Index’s price level of about 1,300. Consensus forecasts call for the index to have $104 in earnings per share this year. If the companies in the index didn’t expect any growth, they would pay out all their earnings as dividends, and earnings and dividends wouldn’t grow. The S&P 500’s dividend yield would be 8%, as the first row of the display below shows.

What the S&P 500 at 1,300 Implies

If the P/E remained unchanged, the total return would also be 8%, but both the S&P 500 and the Dow would stay at their current level. While a flat index price might be disappointing, we think most investors today would probably welcome an 8% return on investment.

Of course, the companies in the S&P 500 do retain a portion of their earnings to finance growth, so the index’s dividend yield is slightly above 2%, rather than 8%, as the second row of the display shows. What kind of earnings growth should we assume?

What About Growth?

Historically, earnings and the stock market have grown with the economy over time, although they can diverge for several years at a stretch, particularly if market euphoria drives stock prices to very high multiples of earnings or if gloom drives stock prices to low multiples. Nominal US GDP, which includes inflation, has grown 7% a year on average since 1947—and so have the S&P 500’s earnings and price. (GDP growth is more commonly quoted in real, or inflation-adjusted, terms. We use nominal growth here to match data for earnings growth and the stock market.)

The three key variables that drive both economic growth and earnings growth over the long term are inflation (which increases the nominal value of economic output), population growth (which boosts the number of people consuming and producing goods) and productivity (which increases the output per person or per unit of capital).

Inflation is widely expected to average about 3% over the long term; population growth, to average about 1%; and productivity, to continue to rise about 2% per year. Since 3% + 1% + 2% = 6%, 6% is a plausible long-term economic growth forecast; it is actually below both the postwar average and the International Monetary Fund’s projections for the next five years.*

So let’s assume 6% economic and earnings growth. With a constant dividend payout ratio, this would lead to 6% dividend growth. Eventually, this growth rate would probably make investors less gloomy, and the market would rise from its current low level of 12.5 times earnings.

If the S&P 500’s P/E rose to 15—halfway back to its average of 17.6 since 1970—the index’s expected return would be 9% per year. At that rate, the S&P 500 would reach 2,000 in five years. The Dow, which typically trades at about 10 times the S&P 500, would reach 20,000 in about five years.

But as discussed above, the market should arguably be trading at an above-average multiple, since bond yields are so low. If the S&P 500’s P/E rises to 20 times earnings as sustained growth in a low-interest-rate environment makes investors more confident, the Dow could reprice to 20,000 immediately, as the third row of the display shows.

Since most investors today would probably welcome an 8% or 9% return for the next five to 10 years (let alone an immediate market revaluation), the current limited appetite for stocks suggests that investors don’t believe in these scenarios. Most likely, they don’t believe in the consensus forecast of $104 in earnings per share this year or 6% economic growth. So let’s examine the implications for stock returns of lower earnings and slower economic growth.

What If Earnings Fall or GDP Growth Slows?

Many people expect earnings to decline because margins are far higher than usual. If corporate spending picks up from the unusually low levels of recent years, margins would fall, and that could drive down earnings.

We think it’s reasonable to expect margins to decline somewhat—although not necessarily to their historical average. But for the sake of argument, let’s look at what would happen if margins declined from 9.5% today to their long-term average of about 6.75%.

Even in this scenario, the S&P 500 would reach 2,000 and the Dow would reach 20,000 in about 10 years. Applied to current revenues, 6.75% margins would reduce S&P 500 earnings by about 30%—to $74, as the fourth row of the display shows.

While there would likely be a severe market pullback initially, if normal economic growth resumed and P/E ratios normalized, the S&P 500 would have a 5% total return and reach 2,000 in 10 years.

But the global economy is now weak, and the European sovereign-debt crisis could end up being a drag on economic growth for years. What if Europe and theUSenter a lengthy period of disinflation? That’s possible, particularly if policymakers are unsuccessful at addressing the world’s serious macroeconomic problems.

So let’s perform a stress test and assume inflation of only 1%, population growth of 1% and no productivity growth at all. That would give us nominal GDP growth of just 2%. A recent survey of professional forecasters said there’s less than a 10% chance that economic growth will be that slow over the next three years.**

What would these dismal economic forecasts imply about future earnings growth and stock returns? If we assume the S&P 500 earns $74 per share this year, 2% growth would still get us to a 4% annualized market return if the market P/E ultimately returns to average, as the fifth row of the display shows. At that rate, it would take 20 years for the S&P 500 to reach 2,000 and the Dow to reach 20,000. Such returns are hardly enticing, but they are still likely to exceed bonds.

Of course, stock-market returns could be worse than 8% (or 4%), particularly in the short term. S&P 500 earnings could fall below $74, and anxiety could cause market valuations to drop even further below normal; both happened in early 2009. Other market shocks are also possible. For example, very high inflation with slow growth could cause price-to-earnings multiples to contract.

But market returns could also be better. Our stress test incorporated draconian assumptions—a 30% drop in earnings plus no productivity growth at all, a very rare occurrence over a 10-year period. Human ingenuity has led to remarkably persistent and steady productivity growth in the postwar period. In recent years, new technology and globalization have driven productivity growth. In the future, these trends and others not yet imagined are likely to continue to drive it.

Faced with uncertainty and traumatized by losses in recent years, investors who are avoiding stocks appear to be assuming that the worst outcomes are highly likely to occur. Or, perhaps, they’ve just lost their stomach for market volatility and are prizing near-term stability over potential long-term gains.

In my next post, I will compare the likely range of outcomes for stocks and bonds.

The views expressed herein do not constitute research, investment advice or trade recommendations and do not necessarily represent the views of all AllianceBernstein portfolio-management teams.

Seth J. Masters is Chief Investment Officer for Asset Allocation at AllianceBernstein and Chief Investment Officer of Bernstein Global Wealth Management, a unit of AllianceBernstein.

*World Economic Outlook: Growth Resuming, Dangers Remain, International Monetary Fund, April 2012

 **“Survey of Professional Forecasters,” Federal Reserve Bank ofPhiladelphia, May 11, 2012



  1. Thomas Rekdal

    An exceptionally clear analysis and explanation. Also, I might add, rather reassuring, because I am an exceedingly patient investor.

  2. Absolutely superb reasoning! Anyone who so agrees with my analysis, right down to the expected return of the S&P 500, must be a truly brilliant person!

    You did leave out that powerful force known as “reversion to the mean.” Using long-term, inflation adjusted, real total returns, it would be reasonable to see an above average total return on the equity side of the market over the next ten years or so. At the same time, bonds have some down-side reversion coming to them. Oddly, if one ignores earnings and only uses long-term real returns, and then includes 2.25% inflation over the next decade, the expected numbers come out almost exactly the same.

  3. Rich Robb

    Extremely well reasoned. But so is a recent B Gross report which arrives at a different conclusion. His factors are not doubt a bit different but I don”t think that”s why his conclusion is different. Help us reconcile two well reasoned outcomes re stock values over the middle/long term

    • Seth Masters

      Bill Gross’s scenario is certainly possible, but not very likely, in our opinion—for the reasons I outlined in the blog.

  4. Josh Elliot

    If you believe in reversion to the mean, then how can the rest of your analysis not be affected by that? Additionally, why use statistics dating to 1947? Given that we are in a credit crisis driven market, shouldn”t we incorporate data from 1900 forward? It seems the traditional business cycles we have experienced since 1947 and therefore their fundamental statistics would not necessarily be a great gauge for today”s situation. Additionally, what about the likelihood of greater economic volatility (multiple recessions during a credit crisis) and that affect on P/E and any fundamental argument to get the DOW to 20,000?

    • Seth Masters

      We do believe in reversion to the mean, which implies that profit margins should decline from current high levels and P/E ratios should rise from the current below-average levels. The net effect should be to help move the Dow towards the 20,000 mark over the next 5-10 years (we’ll send you a copy of our research paper to give some more background on our thinking.) We also agree that long-term data are important–we’ve done studies that go back to 1900 and even earlier, which we think support our conclusions.

      Finally, market volatility is definitely a concern, and it should continue to be a factor in investment decisions. But investors shouldn’t fall into the trap of overloading on bonds trying to minimize volatility in the short term, because that’s not a sound long-term investment strategy, particularly with today”s ultra-low bond yields. We recommend balancing the short-term risk of market fluctuations with the long term risk of prematurely running out of money. And that is why maintaining an appropriate allocation to equities is vital.

  5. Thanks a lot for that adjudication, Mr. Masters. I was hard to find any intelligible descriptions about long-term returns but here they are – your text really explained me a lot. My deepest gratitude for that and best regards, Michael

  6. As the S&P nears that magic 2000 number I want you to know how proud one of Bernstein”s satisfied clients is of this projection.

Leave a Reply

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


You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>