September is the cruellest month

It may only be curious coincidence, but the durability and reliability of the September Curse is striking. Since 1929, US stock prices have declined in September on average by more than in any other month - down by 1.2 per cent, compared with a gain of 0.6 per cent for all months of the year.

And, in case you were wondering, that's not because the average of all those 80 Septembers has been driven lower by one or two horrendously bad months. There's something puzzlingly consistent about September's slide, in fact. In the 30 years between 1970 and 2000, for example, September was the worst-performing month of the year in each of the first two decades and the third-worst month in the last.

Analysts have never really come up with a convincing explanation for Black September Syndrome. So there's always hope. But as this one begins, the omens are not good. Hurricane Gustav is, as I write, still barrelling over the Gulf of Mexico, threatening to halt, for at least a while, as much as 20 per cent of all American oil refinery activity and possibly delivering Katrina-type infrastructure damage along the Gulf Coast.

I don’t know about being an indicator. September is the worst month of the year in the market, with the largest cumulative losses over 1971-2007.

This is true for both the NASDAQ comp and the DJIA.

Triple-witching week in September is often a nightmare, and the week after is often pitiful. There’s a lot of end-of-quarter sales by mutual and hedge funds. This is part of what sets up October for being a good month for either a crash or pretty nice gains.

The last time that September was one of the best months of the year for gains was 1939. September, however, shows up many times as the worst month of the year (by percentage):

DJIA: 1931, 2001, 2002, 1974
SP500: 1931.

By points:

DJIA: 1931, 1929, 1937, 2001, 2002, 2000
SP500: 2002, 2001, 2000.

But the worst pattern has been since 1999, triple witching week and the following week has been abysmal for the DJIA:

1999: TWW -224.80, week after -524.30
2000: TWW -293.65, week after -79.63
2001: TWW -1369.70, week after +611.75
2002: TWW -326.67, week after -284.57
2003: TWW +173.27, week after -331.74
2004: TWW -28.61, week after -237.22
2005: TWW -36.62, week after -222.35
2006: TWW +168.66, week after -52.67

September isn’t a good month, any way you examine the history. Dunno exactly why, it just is. The second half of October is often much better, and November-December-January are the most profitable three months (consistently) in the markets on the long side.

Over what timeframe?

If you’re a buy-n-hold investor, the difference is so small as to not be worth your trouble.

If you’re a swing trader or day trader, well now, there’s some patterns that matter.

Since 2000, Monday and Friday have been battling for the worst day of the week (or the first/last trading day of the week has been the worst day of the week on short weeks like this week). Wednesday has been the best day of the week (for longs) since 2000.

Down Fridays, especially Fridays that are down hard into the close, are often followed by down Mondays.

Here’s the long-term patterns, expressed as a percentage of the times the market closed higher than the previous day, per day of the week, based on SP500 index. “Monday” denotes the first trading day of the week, “Friday” denotes the last trading day of the week.

During 35 years of bull markets:
Mon: 51.5%
Tue: 52.9%
Wed: 58.4%
Thu: 53.2%
Fri: 60.2%

During 20 years of bear markets:
Mon: 40.9%
Tue: 48.5%
Wed: 52.1%
Thu: 50.6%
Fri: 49.8%

This is from June 1952, to April 2007.

As you see, during bear markets, traders don’t like to hold positions over the weekend, or they clean up positions on Mondays.

We’re in a bear market.

All this information and more can be found in the Hirsch’s Stock Trader’s Almanac. They also have a version for commodity traders. They look at seasonal, weekly, daily, hourly, presidential cycles, business cycles, etc over decades of stock market behavior.

One thing that comes through from their research and quantification is that “sell in May and go away” is still very, very true. The months from October to April are the best in the market, and if you found the best stocks to buy (based on MACD indicators) in October, bought and held through April of the following year, then sold, you hold on to more gains than if you held the position through the summer. Summers are just not a good time to hold stocks, overall, if you’re a trader. If you’re an investor, then you can ignore these patterns and stats. Just buy SP500 index funds with a cost-averaging program and you’re going to beat 80% of mutual funds with active management out there.

Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.