A Quant Test on a Trading Thesis

BellaGeneral CommentsLeave a Comment

Hi Mike,

I was halfway through typing my question when I remembered a blog post similar to my original question: http://www.smbtraining.com/blog/how-do-you-determine-how-far-stock-can-move

I was wondering if you’ve found any statistical information by back-testing the above since then. For me personally, I first find important support/resistance levels on a stock & after that do I then consider the likelihood of those support/resistance levels coming into play (intraday) based on the stock’s ATR, gap %, and relative volume. It has been my experience (which is hardly a statistically sound statement) that if a stock is gaping at least 10% with “good” volume, it tends to see an intraday range of 2.5-3X its ATR prior to the news catalyst/gap. I would just like to hear your thoughts as well as to see about how you could back-test this in a meaningful way.

@mikebellafiore                                                        

I asked one of our quants for help here and this was his response:

Support and resistance levels are a great project. It’s relatively easy to define things like high and low of a day or high and low of first x-period (15min, 30min, hour, etc.) or 20-day highs/lows. It’s more difficult to systematically define whether those are actual important tested support or resistance areas from a given day’s trading.
{
Test
I could look for stocks that gapped 10% and had relative volume > 2 by say 10am and then look at the percentages of those that moved x-atr’s by the close.
So the output would look like this:
during the last 20-days there were 200 occasions where stocks fit your criteria. Of those 70% moved 2atr’s from 10am to the close, 20% moved 4atr’s.

You can be better tomorrow than you are today!

Mike Bellafiore

The PlayBook

One Good Trade

No relevant positions

Leave a Reply