Enter the Bull: The DJIA Could Top 15K In 2 Years

Dow Jones Reprints: This copy is for your personal, non-commercial use only. To order presentation-ready copies for distribution to your colleagues, clients or customers, use the Order Reprints tool on any article or visit www.djreprints.com

Based on cyclical patterns of market history, the odds are better than two chances in three that the Dow Jones Industrial Average will reach 15,000 or higher over the next two years. Based on the same cyclical patterns, there's about a 50-50 chance that the Dow could hit 17,000 or more.

Also, the broad fundamentals that could drive the Dow to new highs are fairly clear. The stock market enjoyed double-digit earnings growth in 2011, yet barely rose in response, mainly because of fears over the health of the domestic economy and contagion from Europe. Now that those fears have begun to subside, the market's upside potential can be unleashed.

The market history draws on 141 years of equity performance, from which a fairly straightforward cyclical pattern can be discerned: a strong tendency for periods of worse-than-average returns to be followed by periods of better-than-average, and vice versa. Since the past five years have been squarely in the worse-than-average category, better-than-average returns in the two-year period just begun are now likely.

Enlarge Image

Jeremy Siegel's knowledge of bear markets is enhanced by intimate involvement with two centuries of market history.

The cycles, then, are based on simple arithmetic: two-year intervals following intervals of five years. Also, five-year intervals that are worse-than-than average are objectively defined as belonging to the lowest quartile of all five-year periods in the 141 years tracked. Two-thirds of the time, after a five-year period like the one we've just seen, the market rises fast enough to lift the Dow to 15,000 or higher from present levels over the following two years. The same pattern applies to Dow 17,000 or higher, except that happens just half the time.

These findings are based on the research of Wharton School finance professor Jeremy Siegel, author of the aptly titled best seller Stocks for the Long Run, now in its fourth edition. The Wharton finance professor has amassed numbers on stock-market performance dating back to 1871, the earliest year for which unimpeachable data are available.

His book also presents data that start in 1801.

But cycles are not destiny. And even if they were, these cycles still imply one chance in three that the Dow won't reach 15,000 over the two years.

Professor Siegel offers a bullish scenario based on broad fundamentals that support the two-thirds chance. His forecast keys off the wrong call he himself made early last year. He had expected the strong rebound in the Dow from its March '09 lows to continue through 2011. Instead, the Dow's gains in 2011 ran half the rate of the year before (5.5% versus 11.0%).

Even that disappointing rise was helped by the fact that the blue-chip average ran well ahead of the broader indexes, due to the attraction of its higher dividend-yielding stocks. While the Standard & Poor's 500 index had done even better than the Dow in 2010 (+12.8%), it had zero gains in 2011.

Why the lack of follow-through, despite the growth of earnings? Siegel blames it on two main shocks: the dramatic slowdown in U.S. economic growth that provoked renewed fears of recession, and perceived risks that the economic crisis in Europe would spill over to the U.S.

As both fears have begun to subside, the market has responded on the upside so far this year. Based on Thursday's close, the Dow has risen 5.5% from its close last year, or as much as it gained through all of last year. The S&P has done even better, having risen 7.5% over the same period.

From this point, the market will be sensitive to an easing in both concerns, which Siegel expects will be forthcoming. Economic data so far released have lent credence to a pickup in GDP growth, and future data should lend further support. As for euro land, while the region as a whole is probably in recession already, the market should gain support from continuing signs that a major meltdown is unlikely.

"Many stock bulls are calling for a 10% to 15% gain this year," observes Siegel. "But I would not be surprised to see the market up 20% or more, even if earnings growth slows."

WHILE DOW 15,000 AND 17,000 may sound like dramatic targets, from at least two perspectives -- earnings and inflation -- they are actually rather modest objectives.

In 2011, earnings per share on the Dow grew 12%. Assume a slowdown to half that rate over the next two years, or 6% per year, which is lower than consensus estimates of 9% per year. Since Dow 15,000 from the Thursday's close requires an annual increase of just 8%, the price-earnings ratio on the index would only need to edge up, from the current 13.1 to a still-modest 13.6. Also, if earnings were to grow at consensus expectations, Dow 15,000 could be reached with a P/E of 12.8.

On the same 6% earnings-growth assumptions, Dow 17,000 in two years would boost the P/E on the blue-chip index to 15.4, still average by most standards.

Similarly, in inflation-adjusted terms, Dow 15,000 and 17,000 are actually modest targets (see chart). Assuming price inflation of 2.5% annually over the next two years (last year, it ran 3%), Dow 15,000 in 2007 dollars would still be below its 2007 high. Dow 17,000 would be a new high in 2007 dollars, but would exceed the 2007 highs by only about 7%.

Enlarge Image

IT'S NOT SURPRISING THAT Jeremy Siegel's research has helped turn him into a long-term bull. Over the 141-year period that his data cover, stock market returns, including reinvested dividends, have averaged 8.7%, or 6.4% after inflation. Such is the power of compounding that, at a compound annual return of 6.4%, $1,000 invested 141 years ago would be worth $5.9 million.

It is this research, compiled and updated with the help of one Siegel's former students, Jeremy Schwartz, that forms the basis for projecting the likely upward trajectory of the Dow. (Schwartz is research director of the New York-based WisdomTree Asset Management, a firm with which Siegel is associated.)

Using yearly numbers, Schwartz has compiled returns using rolling five-year periods: 1871-1875, 1872-1876, and so on, culminating in 2007-2011; that's 136 five-year stretches in all. Over the same 141 years, similar data are presented on rolling 10-year (131 periods), 20-year (121), and 30-year intervals (111).

Especially for the five-year intervals, dropping and adding years can often give noticeably different results. For example, the five-year periods ended 2007 and 2008 showed, respectively, average annual returns of 15.0% and -1.4%, even though these periods overlapped by four years. Similarly, the five years ended 2000 and 2002 showed, respectively, average annual returns of 16.6% and -0.9%, even though the periods overlapped by three years.

Starting at the close of each calendar year, the returns assume all publicly traded stocks are purchased on a capitalization-weighted basis, with all dividends reinvested. Yearly percentage returns are equal to the average annual compound rate.

Total returns over recent years have not benefited as much from reinvested dividends. While payouts have picked up since the cut in the dividend tax in 2003, the yield on the S&P 500 is still well below its average in the post-World War II years. As recently as the period from 1990-95, the yield was rapidly declining, but still averaged 3%, compared with a little over 2% today. In the 1980s, the dividend yield averaged 4.2%; in the '50s, 4.9%.

But there has been an offsetting factor. As Jeremy Schwartz points out, "Firms have been substituting share buybacks for dividends at a greater rate over the last 20 years than through much of history." Stock buybacks contribute to total returns by putting upward pressure on stock prices. With cash that might otherwise go to dividends spent instead on buying back stock, the long-term effect on returns from cash infusions might be even over time.

THE LENGTH AND BREADTH of Siegel's historical data have inevitably spawned criticism. In particular, critics have doubted the quality of the 1802-to-1870 data. But as the Wharton professor points out, whether or not that criticism holds up (he doesn't think it does), no one has challenged the data from 1871 on, which forms the basis for the record assembled here.

For comparability with the 21st century, 1871 is a good starting point. That year marks the beginning of the mature phase of the industrialization of America in the post-Civil War period, with a stock market that featured a fair range of different industries, roughly similar to more recent eras. Siegel has also taken care to track all failed stocks into bankruptcy, so there is no "survivor's bias," a common flaw in historical analysis that could artificially inflate performance.

The track record has also been criticized because it assumes purchase of all publicly traded stocks, and there might have been times when certain small-caps were too illiquid to get the assumed executions. But even if that had been a problem, small-caps are too small to affect returns by very much.

Other criticisms that would have applied in earlier periods are less valid now. Dividends and realized capital gains, for example, are taxable, while in this case, no taxes are assumed. But in the era of 401(k)s and IRAs, it isn't unrealistic to assume tax-deferred returns over long holding periods. Somewhat less realistically, management fees aren't subtracted from the returns, either. But given the advent of index funds and exchange-traded funds, some ETFs charging annual fees of just 0.07%.

Enlarge Image

The nearby tables summarize the annual returns for the four holding periods, grouped from lowest to median to highest. In no 20- or 30-year interval have the annual returns ever been negative. Even after inflation, these long-run returns would still be consistently positive, although the very worst periods were hardly enough to build a respectable nest egg.

As the tables show, the returns ran just 2.77% per year in the worst-performing 20-year period, a dubious honor that belongs to the two decades ended 1948, a period dominated by the Great Depression and World War II. But true to the rubber-band cycle effect, the worst 20 years can lead to the best. In the two decades immediately following, through 1968, returns ran 14.83% per year, one of the highest on record.

Through year-end 2011, 20- and 30-year returns, at 8.19% and 10.78%, respectively, were about in line with the median for each. Since these returns were neither very high nor very low, there is no special reason to assume anything much better -- or worse -- over the next 20 and 30 years.

But the story is very different when it comes to the five- and 10-year returns through 2011. They both fall not just in the bottom half of all returns, but in the lowest quartile. Even if we include the strong rebound in January of this year -- turning them into five-years-and-one-month and 10-years-and-one-month intervals -- they still fall into the lowest quarter. It's fair to expect mean reversion from here, which is to say improved performance over the periods following.

Barron's asked Jeremy Schwartz to line up the worst-performing quartile among five-year periods and see what happens over the following two years. To keep his data unbiased by gains during periods of inflation -- gains that are essentially illusory -- the database he used for five-year periods was inflation-adjusted. The recent five- and 10-year periods were also in the lowest quartile after adjusting for inflation.

Schwartz found 33 five-year periods in the lowest quartile for which two-year follow-ups were possible. The first finding was rather stunning: the median annual returns on these 33 periods ran 20%, a strong confirmation of the rubber-band effect. Median returns on all other two-year periods were much lower, at 6.8%. Since those two-year periods all followed higher-performing five-year intervals, the 6.8% tended to confirm the rubber-band effect going in the other direction.

Applying that 20% to the Dow, we first subtract 2.5 percentage points for dividends, leaving us 17.5% a year. Grow the Dow at 17.5% for two years by using the Jan. 31 close of 12,633 -- the final number in the interval of five years and one month -- and you get 17,441. (Of course, at the Dow's Thursday close of 12,891, it's already a bit closer to that target.)

That is why, based on these median returns, we said Dow 17,000 has a 50-50 chance of occurring -- a reasonable assumption given these findings.

Schwartz also found that, of 33 two-year intervals, in 23 cases, or 70% of the time, the returns ran 11.7% or higher. Subtract the same 2.5 percentage points for dividends, and you get growth of 9.2%. Grow the Dow by 9.2% a year and you get 15,064.

Schwartz also found that stocks grew in 30 of the 33 10-year periods, so make that nine chances out of 10 that the Dow will be either flat or higher over the next two years.

How high? The strongest annual rebound post-World War II was 28.7%, in the two years ended 1980. Grow the Dow 26.2% -- again, subtracting 2.5 percentage points for dividends -- and you get 20,120.

Read Full Article »
Comment
Show commentsHide Comments

Related Articles

Market Overview
Search Stock Quotes
Partner Videos