Nov

18

TxgivingOne of the memes developing this year is that Black Friday is being de-emphasized at the margin. WalMart et. al. have started their promotional activities 1-3 weeks early, trying to capture the estimated reduced spend this season. Separately the Street usually gets their hands on very good customer count data from independent (industry specialist) research houses which triangulate physical counts, aggregate credit card data and check-cashing activity among other metrics. Hit up your friendly retailing analyst from one of the big shops for a better vantage point.

Jeff Sasmor adds:

The day after Thanksgiving is always interesting - there are sometimes spectacular pump/dumps in the stock market as the oodles of folks with their Ameritrade and E*TRADE accounts have nothing better to do that day than to lose money till 1 PM E.S.T.

There are also the sales, both in physical and virtual space. I don't go near any mall that day. You usually can't even get a parking space. How full will the parking lots be this year, with all the media talk of belt-tightening and consumer reticence to buy? With modern-day inventory systems, we will know the results over the weekend.

Kim Zussman presents his Thanksgiving analysis:

In recent years, the 30 days before Thanksgiving have been quite bullish (SPY 93-06):

One-Sample T: pr 30D

Test of mu = 0 vs not = 0

Variable N Mean StDev SE Mean 95% CI T P

pr 30D 14 0.04775 0.05824 0.01556 (0.01412, 0.08138) 3.07 0.009

However barring an explosive rally in the next 3 days, this year is not (assuming Weds closes about current levels); the pre-Thanksgiving 30D is about -6%. Which could make the pre-holiday period of 2007 the worst of of the prior 13 years:

Date pr 30D
11/24/2006  0.043
11/25/2005  0.081
11/26/2004  0.065
11/28/2003  0.013
11/29/2002  0.089
11/23/2001  0.053
11/24/2000 -0.032
11/26/1999  0.107
11/27/1998  0.179
11/28/1997 -0.013
11/29/1996  0.072
11/24/1995  0.036
11/25/1994 -0.031
11/26/1993  0.005


Comments

Name

Email

Website

Speak your mind

1 Comment so far

  1. gabe on November 20, 2007 12:59 am

    good exercise worth following the outcome. there are still 48 hrs to go but since the explosive rally looks very improbable, this would be the 2nd highly significant test (at 1%) you ran, after the black mondays analysis (at 2%) that didn’t come true. what are the odds of that? it looks to me the (non)independent observations condition should somehow be counted for.

Archives

Resources & Links

Search