How Now, Mighty Dow? , by Kim Zussman
Dividing DOW weekly closes into non-overlapping 25 week segments, I checked for large advances similar to now using:
{max close (recent 25W)} / {min close (prior 25W)}
Currently this quantity - the move to the high weekly close of the 25 weeks ending last week, compared to the low weekly close of the prior 25 weeks - is 0.51 (+51%). I checked back to 1929; there has not been a gain over 50% since 1983, then 1974, and in the 1930's (see attached).
Looking only at periods with gains >40%, here are the returns over the subsequent 25 weeks:
Date 25 max/min nxt 25W
02/19/34 0.960 -0.142
08/28/33 0.957 0.024
02/27/33 0.793 0.925
06/27/83 0.584 0.021
12/05/38 0.529 -0.072
10/27/75 0.512 0.197
01/20/36 0.498 0.116
05/05/75 0.454 -0.017
10/19/87 0.446 0.032
03/29/43 0.441 0.033
07/29/35 0.439 0.173
05/12/86 0.418 0.072
10/21/29 0.403 -0.024
. avg 0.103
Anton Johnson comments:
A quick look at all non-overlapping 25-week periods for weekly Dow 1929-Current yields a mean return of 2.73%; using this as the baseline results in a 7.56% excess return for the R>40% periods (10.29-2.73 = 7.56). However the 92.5% return in 1933 significantly skews the small R>40% sample. Median returns are 3.57% and 3.20% respectively, not much different from each other.