Oct

17

Black Mondays? from Kim Zussman

October 17, 2007 |

Are they worried about black Mondays in October?  If so, one might expect Friday returns (Thr close - Fri close) for Octobers to be lower than other months, as there would be selling to avoid down opens on Mondays.

Tested this using SPY 93-07:

Two-sample T for F cc vs oct Fri

          N    Mean   StDev  SE Mean
All Fri   742  0.0003  0.0107  0.00039 t=-2.3
Oct Fri   64   0.0039  0.0120  0.0015

No, they are not afraid to own Fridays in October. In fact October Fridays are significantly higher than all Fridays in the period.

Are October Mondays blacker than the rest (Fri cls - Mon cls)?

Two-sample T for F-M all vs oct F-m

          N     Mean   StDev  SE Mean
All F-M   672  0.0010  0.0108  0.00042 t= 0.8
Oct F-M   65  -0.0004  0.0124  0.0015

October Mondays are slightly lower than all Mondays, but not significantly.


Comments

Name

Email

Website

Speak your mind

3 Comments so far

  1. gabe on October 17, 2007 7:19 am

    assuming you found Friday’s mean to be significantly, say 2% higher. How would one go about to trade it? The bid-ask spread on options is alot higher than this, at least from my broker.I think a study of Highs-Lows would be in order? but that complicates the matters more.

  2. James Smith on October 18, 2007 5:45 pm

    Kim, that is a nice regression and reasonable area for inquiry. Can this also be extrapolated to indicate that there is little superstition in the market regarding any various days in the calendar?

    One way the numbers might be compared more meaningfully is to take just 64 samples from pre 1987 trading and compare that with the 64 samples since. That way the “law of large numbers” (which tends to reduce variability toward a more normalized mean). As the number of observations increases, the trend toward a more stable independent variable is observed. Comparing 64 observations with 742 is a bit skewed in favor of 742 std. deviation and mean, they are not equal statistically, are they?

    It seems that Black Monday (equity market), and the tech wreck of 2000 (tech sector of equities), and the Yen (fx and Japanese realestate) losing value in the mid 90’s were all rather predictable using an intrinsic value appraoch to see these markets were overvalued and due for a correction.

    Does taking a “sell the winners” and “buy the losers” approach based on market fundamentals, make sense? One major firm (not an investment bank or alternate strategy group) of asset managers believes so and has apparrently, consistently, followed a disciplined trading analysis and methodology to allocate assets out of the areas where the most profit is to be taken (selling in the winning sectors now overvalued) and buying into the sectors where the most profit is to be made (i.e. out of favor or “losing” sectors).

    Jim Smith

  3. gabe on October 19, 2007 9:08 pm

    today’s observation sure puts things in perspective. could you re-run the statistics with today’s data? I’m curious if one data point can change the significance

Archives

Resources & Links

Search