The Importance of the Post Labor Day Trade

By Dr. Duru written for One-Twenty

September 5, 2006


{ RSS (XML) - Subscribe to my latest articles here: XML...... AND/OR ! }


Many traders look forward to the action after Labor Day. The Tuesday following Labor Day effectively marks the end of summer and the listless, low volume trading that supposedly occurs every year. Then again, we have all heard how September and October have historically averaged the worst monthly performances (on the S&P 500) of all months. We have learned to dread the return to trading in the early Fall! How do we reconcile the anticipation versus the anxiety?

When I last tried to measure the importance of post Labor Day trading, I compared the day change in price of the S&P 500 at the close the Tuesday after Labor Day to the remaining index performance for September and October. The results were mixed. This year, I decided to tackle the issue from a much broader perspective. I looked for how strongly correlated the post Labor Day trade has been to other price changes over the year. I also looked at whether these various time periods tended to show higher or lower % gains than the post Labor Day trade. My results were quite surprising!

I measured several time periods all based on the closing price of the S&P 500 on the day in question:
  1. Post Labor Day: Friday before Labor Day to the Tuesday after Labor Day.
  2. Remaining Year: The Tuesday after Labor Day to the end of the current year in question.
  3. Year Over Year: The last day of the prior year to the last day of the year containing the Labor Day in question.
  4. Year to Labor Day: The last day of the prior year to the Friday before Labor Day.
  5. Year Over Year, Labor Day: The Friday before Labor Day of the prior year to the Friday before Labor Day of the current year in question.

The time period covered 1962 to 2006. I next measured the correlations of the post Labor Day trade to the other four time periods. I also measured the correlation of the Year to Labor Day vs Year-Over-Year. Since all the correlations turned up positive, I added a table comparing the size of the price change from the post Labor Day trade to the other time periods.

I have posted the results below to allow you to form your own conclusions. (I can also send to you the raw data of this analysis if you need it. Just email me.) This is what I conclude:
  1. It has been much more common for the post Labor Day trade to move in the direction of the trade over any of the time frames of interest in the past 10 years than over larger samples of history.
  2. The Post Labor Day trade has been more important in the past 5 years than in any other time frame measured in this sample of history.
  3. The performance of the S&P 500 for the year up to Labor Day is very highly correlated with the year-over-year performance of the index. This correlation is far more important than the daily trade the day following Labor Day.
  4. However, in the past five years something has happened that is unusual in this 45 year sample. The Post Labor Day trade actually performed better than all years except for 2003 where the S&P500 had a monster gain of 14.57%. For contrast, note that the year-over-year performance was better than the year to Labor Day performance in four of the last five years. 2002 was the exception when the carnage up to Labor Day was three percentage points better than the 23% loss seen that year.
  5. The market tends to perform better over time than the single day trade on the day following Labor Day. Of course, the longer the time frame, the more likely performance will exceed that of any given day.
  6. Finally, if previous history's tendencies hold true, then the post Labor Day trade for 2006 is very likely to show a positive gain, and a strong one at that (1.0 - 1.5% or so?). (See my earlier missive declaring that we are likely to see on Tuesday a sharp move toward current resistance levels). Similarly, we are practically a lock for a positive gain on the year.

(If the table below has formatting problems or is hard to read, you can click here for a properly converted table)

S&P500"LaborDay"Analysis(asof9/4/06)

Last # of Years (including 2005)
Correlations 5 10 25 44
Post Labor Day vs. Remaining Year Performance 0.571 0.588 0.389 0.333
Post Labor Day vs. Year-Over-Year 0.813 0.780 0.552 0.443
Post Labor Day vs. Year to Labor Day 0.815 0.564 0.231 0.216
Post Labor Day vs. Year-Over-Year, Pre-Labor Day 0.681 0.415 0.225 0.260
Year to Labor Day vs. Year-Over-Year 0.992 0.847 0.742 0.839
Last # of Years (including 2005)
% of Years Gain of Second is Higher Than First 5 10 25 44
Post Labor Day vs. Remaining Year Performance 80% 80% 80% 75%
Post Labor Day vs. Year-Over-Year 60% 70% 76% 70%
Post Labor Day vs. Year to Labor Day 20% 50% 72% 68%
Post Labor Day vs. Year-Over-Year, Pre-Labor Day 60% 80% 80% 73%
Year to Labor Day vs. Year-Over-Year 80% 80% 72% 68%
2006 Performance Change
Post Labor Day ?
Remaining Year Performance ?
Year-Over-Year ?
Year to Labor Day 5.02%
Year-Over-Year, Pre-Labor Day 7.63%


Be careful out there!

DrDuru, 2006