Waiting for December's Bloom

Updated from Dec. 11

The Dow made headlines Thursday by crossing over 10,000 and holding it through the close. Could it be the beginning of a sustained "December effect" in stocks?

With the Dow and S&P 500 up over 15% for the year, and the Nasdaq up more than 40%, investors might be hoping that a reluctance to part with winners before the start of the new tax year will drive prices even higher -- the so-called December effect.

If the major indices were able to build on the momentum and put in a strong December, however, it would arguably be the first time this year stocks behaved in a way that is consistent with conventional wisdom.

Consider the one about January being a harbinger of how the markets will perform for the year. By the end of January, tea-leaf readers were sure 2003 would be dismal. The S&P 500 was down almost 6% for the month, a move the Stock Traders Almanac says portends a down year 92% of the time. Barring a breathtaking turnaround, that's not going to hold for 2003.

Another prognostication holds that January's moves are usually corrected in February. But after posting 6% declines in January, the Dow and S&P followed up by dropping again in the year's second month.

March mythology has it that stocks rally the day before St. Patrick's Day and then go flat on March 17. This year, St. Patrick's Day traders clubbed sellers with a shillelagh, driving the Dow and S&P up nearly 4% and the Nasdaq up almost 5%.

But the market saved its worst thrashing for those who adhered to the "sell in May and go away" theory.

When researchers at Charles Schwab looked at the numbers going back to 1926, they found it hard to understand how this legend ever gained currency. Which month produces the highest average monthly return for the S&P 500? July. The weakest month, both in terms of average monthly returns and frequency of negative returns, is September, the time everyone is supposed to be leaving the beach behind and getting back to business.

Source: Charles Schwab

Schwab calculated that $10,000 invested in the S&P 500 index starting in May 1926 would have grown to more than $23 million by April 2002. Had the same investment been withdrawn each May through October, it would have grown to only $5.7 million by April 2002.

In 2003, a $10,000 investment returned less for those following the notion.

Sell in May and Pay $10,000 with and without the 2003 summer
	Full Year	Out June-August	Missed Gain
Dow Jones	$11,644	$10,988	5.97%
Nasdaq	14,036	12,325	13.88
S&P 500	11,655	11,308	3.07

Schwab researchers did find that the S&P 500's average monthly return has been 0.4% higher for the November-to-April period vs. the May-to-October period. But the slightly better return during the winter months is not a strong argument for trying to time the market. Even if this had been a weak summer, alternatives were limited, with bond yields historically low and cash producing a negative real return.

The autumn proved to be no better, as yet another market-timing theory fell apart. Since 1950, November has been the strongest month for the S&P 500, but this year the index was flat, along with the Dow and the Nasdaq.

Looking forward to 2004, investors may be looking for a "January effect," when people buy back their December losers and start selling 2003 winners so the capital gains taxes land in the 2004 tax year. This action has not moved markets in years, as these maneuvers are expected and priced in well in advance.

Of course, 2004 is also an election year. Every time the incumbent has lost the White House, the stock market has declined. While many may feel inclined to bet on President Bush's re-election, it's also true that no one who lost the popular vote has ever gone on to win a second term.

All was not lost in 2003. One predictor to hold up was the Super Bowl theory. When the NFC team wins, the stock market has been up for the year 85% of the time. This year, the NFC's Tampa Bay Buccaneers trounced the Oakland Raiders 48-21.