How High (or Low) Could the Stock Market Go?

Stock chart on trading floor of NYSE
Stock chart on trading floor of NYSE

The U.S. stock market has been on a tear since September, gaining more than 20% in a mere five months. For perspective, annual equity returns average about 5% over the long run.

That means the market has logged four years of average gains in only five months. And it raises the question: What's next for the U.S. markets, not just next month, but in the next year or two?

The bullish case is well established: The economic recovery is solidly advancing, corporate profits are still rising, inflation is low and some evidence shows companies are starting to hire again.

Let's look at a 10-year chart to identify the bullish targets for the Dow: 13,000 and then 14,000.

Chartists have long noted that long-term tops often form what is called a "head and shoulders" pattern in which lower "shoulders" precede and follow the peak or "head." This pattern is clearly visible in the Dow's 2007-2008 top and decline.

The sharp 84% rise from the March 2009 low has brought the Dow back above the 12,000 level and into a band of resistance and potential support around 11,900 to 12,200. The next stop for the bulls is 13,000, the left shoulder on the chart above, last touched in 2008. Beyond that, the next goal is the 14,000 level that marked the 2007 top.

From the view of a 10-year chart, we can see that this advance from 6,500 to 12,170 has been meteoric compared to the more leisurely recovery from 2003-2006, when it took about four years for the market to advance from 7,600 to over 12,000. This suggests that the past two years have been extraordinary rather than typical, and so we might expect more typical returns in the years ahead.

What's Typical?

You'd think figuring out what "typical returns" are would be a relatively straightforward calculation, but -- as with many things financial -- it turns out to be complicated. Some calculate long-term annual returns of around 7%, and others estimate 6.5% as a reasonable expectation for total returns, or dividends plus appreciation/growth. Yet other careful analyses reckon that a return of roughly 4.1% is more realistic.

Sponsored Links

Why is it so difficult to assess mean returns over the long term? Economists Eugene Fama and Kenneth French have shown that the uncertainties of expected returns don't diminish over long time frames, so the uncertainties of 30-year and 50-year returns are higher than shorter-term yields. In effect, the uncertainty over two years is four times the uncertainty over one year.

Calculating long-term returns and mean returns turns out to be an inherently iffy proposition. "In particular, we don't know the true expected returns on portfolios," Fama wrote in an investment forum in 2009. "We typically use historical average returns to estimate expected returns, but the estimates are quite noisy, and they leave lots of uncertainty about true expected returns."

In other words, projections using average annual returns are guesstimates because historical returns aren't reliable guides.

To put this truism into perspective, let's turn to some longer-term charts showing the Dow, from 1977 to the present, and the broad-based S&P 500 index , from 1965 to the present.

We can see that stocks really took off in 1995, and made an unprecedented ascent in five short years to the dot-com top in 2000. In the Dow, this top marks what could be a left shoulder in a multiyear topping pattern, with the peak reached about seven years later in 2007 tracing out the head. If this pattern holds, then the current rally may be the right shoulder.

Alternatively, the Dow might reach for the 14,000 level, and either form a double top there or move on to new heights.

The S&P 500 has already traced out a multiyear double top, with the first peak in 2000 and the second in 2007.

A Return to the Long-Term Average

One tool statisticians use is the "regression (or reversion) to the mean," which refers to the probability that extremes of activity or response tend to revert to the long-term average.

This suggests that any period of extreme outperformance, such as the past 22 months, will be followed by lower and more average returns.

We can estimate the mean return in several ways. On the charts, I took 2% above inflation as a baseline return. At this rate, $1 invested in 1989, six years into the great 1982-2000 Bull Market, would have grown to $1.52 in 2010. A dollar in 1989 now equals $1.76 in 2010 dollars, so I've multiplied our return by 1.76 to adjust for inflation.

By these calculations, the Dow should be around 5,300 and the S&P 500 should be about 800.

If we forget inflation and just plug in an annual mean return of 6.5%, then the S&P 500 should be about 1,125. If we go with a 5% mean return, then the S&P 500 should be around 835.

A Safe Bet, If. . .

Author and analyst Jeremy Siegel found that since the early 1800s, equities had never offered a negative return, after inflation, if held for 17 years or more. That suggests stocks are a safe bet for long-term investors, if they can handle short-term volatility.

Technically speaking, these long-term charts suggest that stocks entered an unusual period of outperformance in 1995 that will eventually end as returns revert to longer-term averages. Nobody knows what those averages will be, but we can look to history from some guidance, and to charts for possible future outcomes.