TSX Energy; Seasonality vs. Proper Analysis

A few weeks ago I examined a seasonality clip from DVTechtalk January 18, 2010 – promoting Thackray’s 2010 Investor’s Guide and highlighting Seasonality in the Energy Sector. According to Thackray’s 2010 Investor’s Guide, seasonal influence on the U.S. Energy sector is from February 25th to May 9th. Brooke also has completed other studies in the sector and has found that seasonality in the Canadian energy sector, U.S. Oil Services sector and the U.S. Oil Exploration and Production sub-sector are slightly different. Their period of seasonal strength is from January 30th to May 9th.

I back tested the seasonal trades and discovered the seasonal Energy calls simply never worked over the last 10-years. A simple buy-and hold from January 2000 to date returned 315% and the seasonal trades generated 179% over the same period. The worst period was during the great 2000-2006 advance when the seasonal trades had us sell high and buy back even higher.

I also found that when traditional and legitimate technical studies are used, better performance is achieved with much lower trading activity and the associated costs. Now “traditional and legitimate technical studies” can be long term moving averages, primary trend lines or simple weekly or monthly momentum studies. In the study pictured I used simple relative averages (RA) or the relative performance of the TSX Energy vs. the large cap S&P/TSX60 Index – monthly data. I have also laid on a True Range Price Channel to illustrate a supporting study using totally different math. In each case we were out during the bumpy 1998 – 2000 window and were long through the great 2001-2007 advance during which the seasonal strategy generated 12 in and out trades.

At this time our simple RA model has us long on July 31, 2009 following a sell on October 31,2007. Most notably is the low trading activity with only 5-trades over 11-years. The last sell is to be ignored because we only have a partial February bar.