When you realise that 2010 to 2016 has been a tremendous bull-run, to see the number of big decliners outpacing big runners makes my point.
Now based on hard-data based on my premise (avoid big moves that are just dead-cat bounces from prior sell-offs, or sell-offs following prior run - no new 3-month price level record being set), I went back ten years ago (June 2007 till date on the SPY), and here are moves meeting my definition. Only two run-ups, and 22 declines
Run-ups
Date Fractional rally
04/05/09 0.03402
27/10/11 0.034835
Decliners
Date Fractional decline
15/09/08 -0.04759
17/09/08 -0.04496
29/09/08 -0.07836
06/10/08 -0.05093
07/10/08 -0.04479
09/10/08 -0.06984
24/10/08 -0.05071
27/10/08 -0.0355
19/11/08 -0.06408
20/11/08 -0.07423
23/02/09 -0.03578
02/03/09 -0.04504
05/03/09 -0.04085
04/02/10 -0.03087
20/05/10 -0.03776
04/06/10 -0.03514
29/06/10 -0.03088
04/08/11 -0.04684
08/08/11 -0.06512
21/08/15 -0.03108
24/08/15 -0.04114
24/06/16 -0.03643