Sunday, December 19, 2004

Seasonality of the Market

I had a rather interesting conversation with my wife today about market timing. Specifically, the peculiar way markets tend to surge just a few days/weeks a year. The rest of the time...the market just churns.

We were trying to determine if there would be a way to reduce the number of days to have funds in the market. In a typical year you have approximately 250 trading days. If you could figure out a pattern when most surges occur...take steps to ensure you were 100% invested during that timeframe. And possibly step out of the market during the churn cycle. Similar to the Sell in May, walk away adage.

So, I started testing. Looking for the best months of each year in the DJ-30, S&P 500, and Nasdaq Composite. Here's some results from the study.

Best months of performance were April (13), July (13), November (12), and October (10).
Data included years 1915 - 2004.

S&P 500:
Best months of performance were October (6), March (6), May (5), and January (5).
Data included years 1962 - 2004.

Nasdaq Composite:
Best months of performance were April (2), March (2), May (2), August (2), and December (2).
Data included years 1990 - 2004.

In the above examples, the DJ-30 April month acheived the best gain of the year 13 times during the years of 1915 - 2004. The S&P 500's best month was split between October and March. And the Nasdaq Composite split its best performing months between Spring months (March - May), a late summer month (August), and a winter month (December).

After just a quick review of this looks like the old adage of Sell in May, walk away might not be as simple as it sounds. But, of course, more studies of the data needs to be done.

More to come in tomorrow's post.

Until then...

Daily System Updates
Received a close trigger for the QQQQ Harami long system at Monday's market open. But, this weekend received a long system trigger for the QQQQ from a different system (BBLower). As a result of this...I'll continue to hold the QQQQ position until the new BBLower system issues a closing trigger.

The BBLower system is based on the Bollinger Bands. Specifically, a pricebreak below the lower Bollinger Bands. This is not one of my favorite systems. It has an 80.10% win ratio and an average profit of 1.81%. I typically do not like to use such a low average profit on a system. As a result, this system is still under evaluation. There's a good chance this system could cycle into the "Not Ready for Primetime" folder. Unless I can find some way to improve it's performance.
  • 1 QQQQ (Harami) long with current loss of -.05%.

Weekly System Updates
Long system trigger for CARN at Monday's market open. The system that generated these trades is the same system that generated EGHT, ZICA, VNWK, SPDE, and AATK trades the previous weeks. So, as you know, system caters to penny-type stocks with high volatility. Holding periods are mid-term in length. This system is one of my riskiest systems. I trade very little size with trades from this system because losses can be extreme.

Current open system positions:
  • 1 QQQQ (SimpleUp) long with current profit of +4.03%;
  • 1 QQQQ (TurtleTrader) long with current loss of -1.60%;
  • SPDE (PennyLag) long with current loss of -13.08%;
  • AATK (PennyLag) long with current loss of -18.25%.
  • VNWK (PennyLag) long with current profit of +8.12%;
  • EGHT (PennyLag) long with current loss of -4.10%;
  • ZICA (PennyLag) long with current loss of -1.00%.
Please read the disclaimer on the website. This is not a recommendation to buy, sell, or trade securities. Just a journal of my travels through Wall Street. I can buy, sell, or hold any positions mentioned on this website at anytime. So, be warned.

No comments: