"Everyone I talk to says we'll have a fall rally [this year] due to the seasonal patterns," a Wall Street Strategist told Barron's a few weeks ago. The strategists polled by the newspaper are bullish, too. They expect stocks to rally between 5 percent and 10 percent by the end of the year. The median forecast of the 10 strategists was for a gain of 6.5 percent. The two hurricanes that hit the Gulf Coast this month have not swayed the group much. Five of the six strategists polled by Bloomberg this week still expect a year-end rally.

In that event, the market's pattern this year might resemble last year's. In 2004, after spending most of the year in a trading range with a flat return, the market leapt in October as the presidential election passed. The market ended the year up 10.5 percent.

The question of whether the market can repeat this performance is purely speculative. What is more certain is that a rally, if one were to ensue, would need to have its leaders. So it's interesting to examine the industry composition of recent rallies, and look at which industries could lead a market advance this year. Would it likely be the same groups that led last year's run-up? Do those look primed to rally this year? Might a different set of stocks lead?

In some ways, last year's rally was straight from a textbook. As investors gained confidence in a continuing economic expansion, they pushed up the prices of cyclical stocks. The Morgan Stanley Cyclical index climbed 19 percent from the end of October through the beginning of March. The S&P 500 rose 12 percent during the same period.

On closer inspection, though, it was not just any cyclical stocks that led the way. Companies that are traditionally sensitive to the latter part of an expansion registered the best performance. You can see this by comparing the performances of two indexes: The Merrill Lynch Early Cycle Index (XE) and the Late Cycle Index (XT).

The Early Cycle Index includes stocks in industries that are traditionally sensitive to the first part of an expansion, including retail, home furnishings, and auto manufactures. Companies in this index include Maytag, Circuit City , and Nordstrom.

Late-cycle stocks are often those that respond to an increase in manufacturing and the expansion of capacity at plants and factories. This would include companies involved in the supply of materials, commodities, and chemicals, including names like Phelps Dodge, US Steel, and Alcoa.

During the 2004 winter rally, investors showed a clear preference for these late expansion companies. The Early Cycle index rose 6.8 percent from the end of October through the beginning of March, about half of the S&P 500's gain. The Late Cycle Index jumped 20.5 percent.

Considering the age of the current economic expansion, last winter's move in late-cycle stocks made sense. The economy was finishing up its third year of an expansion. That's an economy in the second half of its life, at least by the averages. Since 1945 economic expansions have lasted about 4.5 years.

If investors were to decide that the current expansion has legs, they might turn to late-cycle stocks again. But, so far, the performance of this group has been poor. Since March of this year, the late cycle index is down 16.5 percent, trailing the overall market by a wide margin. Compare this performance with the period prior to last year's winter rally. Both early cyclical and late cyclical stocks held up well enough, even outperforming the S&P 500 through the range-trading period.

Broadening the analysis to the 24 industry groups in the S&P 500 shows a similar pattern. Since early March of this year, the two worst performing industry groups are Materials and Autos. Other cyclical groups trailing the market average include Consumer Services, Capital Goods, and Consumer Durables. Again, these same groups held up heading into last October, and then led last year's winter rally.

Despite this lackluster performance, investors could change their outlook for cyclical stocks. In fact, they often do at this time of year. Since 1990, during the months from November to March, Morgan Stanley's index of cyclical stock has outpaced its consumer stock index in 13 out of 15 years (and the S&P 500 Index 10 out of 15 years). Since 1980, cyclical stocks have outperformed consumer stocks 70 percent of the time during the winter months.

Investor's preference for cyclical stocks late in the calendar year might be explained by their optimism for a new year. Also, the stock market tends to do better through these months, on average. So the higher beta cyclical stocks benefit from the overall markets seasonality. (This pattern is not limited to US markets. In a 2001 study of 37 developed and emerging markets, researchers Ben Jacobsen and Sven Bouman found that 36 of the countries had stronger returns through the fall and winter.)

A similar pattern can be seen in the performance of value and growth indexes. Value indexes beat growth indexes twice as often during the fourth quarter of each year. This performance, and the performance of cyclical stocks during these months, fits with much of the research which shows that during periods of optimism and high expectations cheaper stocks perform better than growth companies.

What other groups might lead?

Again, the question of a year-end rally is speculative, but allowing for the possibility, if cyclical stocks don't spark a rally, which other groups might? Financial stocks are an obvious pick, since they account for roughly 20 percent of the S&P 500. But the group's performance of late has been mediocre. Financials have lagged the market's return in the two most recent periods: during last year's winter rally and during this year's range-trading period. The group continues to face a series of headwinds, including higher short-term rates, a flatter yield curve, and a housing market that looks to be cooling.

The technology group, which tends to be influenced more by its own product cycles than the overall economy, is another candidate. In fact, these stocks have helped to keep the market in positive territory this year. Technology companies are scattered among the best performing industries this year, including Healthcare Equipment, Semiconductors, and Technology Hardware. These tech industries trail only the hot-money three: Energy, Utilities, and Real Estate.

Tech stocks also have the season on their side. The chart below shows the top-performing industry groups in the S&P 500 during the months of November to March since 1989. Of the 24 industry groups in the S&P 500, the three technology benchmarks hold the top three spots. Semiconductors averaged 15 percent returns, Tech Hardware rose 10.2 percent, and Software & Services gained 9.3 percent through the winter months. Yes, the late 90's tech rally influenced the overall returns. But during the full period, the returns for the Technology industry during the winter months were twice those returns the group delivered from March to November.

Even considering the potential for technology stocks to break out of their relative slumber and take a leadership role, it's difficult to imagine any sustained rally without the contribution of some type of economically sensitive stocks (other than just the Energy group). Since the market's peak in 2000, there have been four rallies over various time frames. They began in September 2001, October 2002, March of 2003, and in October 2004. In each of these rallies, cyclical stocks ran briskly ahead of the S&P 500. In the 2001 rally, the gain in cyclical stocks was double that of the market. In the year-long rally beginning in March of 2003 cyclical stocks were up 63 percent, while the market rose 35 percent (after two and half years of relentless losses).

Studying the industry make-up of market moves is always informative. With the aftermath of the Gulf area storms increasing the potential for slower economic growth, and with such wide-spread hope for a winter rally, thee pattern of leadership from individual industries may take on even greater importance.