For fun, I calculated accuracy of my predictions for the past 3 months (December, 2012 through February, 2013)–the post is as of Feb 25, so I will update this in a few days.
EDITED on March 08, 2013 with data for Feb 26-28:
Accuracy overall, regarless of the magnitude of DJI move
31/56=55.36% (period Dec 1, 2012–Feb 28, 2013)
Move over 50 pt (abt 0.36% of DJI at 14,054 on Feb28, 2013)
19/31=61.29%
Move over 141 pt (abt 1.00% of DJI at 14,054 on Feb28, 2013)
4/6=66.67%
See table stats here: Dec12_Feb13_accuracy_stats_web_Feb28
Arrows show my prediction of DJI going up or down, as well as what DJI actually did on that date.
One can see that the accuracy improves for days with stronger market moves–it is not surprising, as stronger astrological aspects are clearer to read & easier to base predictions on. For example, Saturn conjunct Venus transit pretty much always coincides with a down move, whereas Jupiter square or opposition Mars can go either way (although it will surely coincide with increased activity–buying/selling–whichever, but ppl will go overboard with excitement!).
Tolstunka 
For Feb’13 alone, corresponding stats are 11/18 (61%), 6/8 (75%), and 1/2 (50%)–a bit better, but too few datapoints to reach any major conclusions.
I do think taking transits to the US chart into consideration helps to improve accuracy though–for DJI, they seem to often override the global trend.