Mar

7

The DJIA has now gone down for four consecutive weeks after an up (UDDDD), and the total decline is about -20%. Looking back 1929-present, this decline of 20% is the 7th worst out of 104 instances of UDDDD.

What usually happens next ? X for UDDDDX was insignificantly negative: 

Variable    N     Mean     StDev   SE Mean     95% CI             T      P

nxt wk    103  -0.0028  0.0328  0.0032  (-0.0091, 0.0036)  -0.85  0.398

And regressing next week's return vs return of prior 4 weeks suggests bigger declines do not make for a better return (see scatter diagram ):

Regression Analysis: nxt versus dn 4

The regression equation is

nxt = 0.00799 + 0.142 dn 4 

Predictor     Coef   SE Coef     T      P

Constant   0.0079  0.0053  1.50  0.136

dn 4         0.1415   0.0563  2.51  0.014

S = 0.0320660   R-Sq = 5.9%   R-Sq(adj) = 4.9%

Comments

Name

Email

Website

Speak your mind

4 Comments so far

  1. curmudgeon 4361 on March 7, 2009 1:15 pm

    What the scatter diagram also shows is the large volatility that follows a steep 4 week decline like this one. Anything from -12% to +10% for next week is in the cards if we define a “steep” decline as -10% or worse. “Shallow” UDDDD declines are more sedate in their after effects.

  2. Anton Johnson on March 8, 2009 1:14 pm

    For a recent small sample perspective, using the last one-year data there are 6 (this one inclusive) occurrences of weekly UDDDD, with the 5 prior resulting in no 5th week down, and a 3.94% median return.

  3. Andrew McCauley on March 8, 2009 7:52 pm

    The recent sample with regard to the S&P 500 is not as positive, posting a median next week return of 0.41%, N = 5, win rate of 3 in 5, max = 5% & min = -2%.

  4. curmudgeon 4361 on March 13, 2009 3:46 pm

    Well, how did the “X” week turn out:

    16:12 *S&P 500 RISES 11% IN BIGGEST WEEKLY ADVANCE SINCE NOVEMBER

    is the headline that just flashed on the tape.

    It was a big move allright.

Archives

Resources & Links

Search