Gap Fade Stats for August
Wow – August was another month with a high percentage of overnight gaps, but let’s see how many of these overnight gaps in the DIA (Dow Jones) were filled intraday.
I define a gap as being an open that is at least $0.20 different than yesterday’s close and use Excel for the statistics.
There were 21 trading days in August, and 17 of these (81%) showed an overnight gap of at least $0.20 in the DIA.
Of these 17 gaps, 8 gaps filled, giving us the first under 50% month this year, with 47.06% of overnight gaps filling intraday.
There were 7 up-gaps, and 4 of these filled (57.14%).
There were 10 down-gaps, and only 4 of these filled (40.00%).
Despite the daily chop, there was not one gap greater than $1.00 (there was a gap of $0.98 on August 5th which did not fill).
The following print-out shows the DIA gap fill statistics for August:
I’m including the “year to date” numbers from January 1st until August 31st 2008 in the DIA (of gaps greater than $0.20):
As you see from the table below, 77% of days have opened with a gap at least $0.20, and 62.50% of these gaps have filled. The table shows the whole picture so far:
Up gaps have been more likely to fill than down gaps this year, and almost 3 out of 4 up gaps have filled, while 1 out of 2 down gaps have filled.
2008 has been a generous year so far for the gap-fade strategy. Will it continue?
Check out these prior gap fade statistics months:
January Gap Fade Statistics
February Gap Fade Statistics
March Gap Fade Statistics
April Gap Fade Statistics
May Gap Fade Statistics
June Gap Fade Statistics
July Gap Fade Statistics
Hey Corey,
This is an interesting study, is there software that you use to get your results? Mainly, is it something that I can find somewhere myself?
These stats are great! I assume you used end-of-day data?
PW,
I download data from Yahoo Finance on the DIA depending on how much data I want to test and then have created my own sheet in Excel that performs these calculations and sums them up for the final results. Once the calculations are in place, you just import the data and verify that the results are correct by comparing the results to a price chart. I don’t know of any software (for free) that does this. It can also be programmed into TradeStation, TraDecision, or other strategy testing software.
Prospectus,
I do – I import Yahoo’s end-of-day data into Excel.
I tried building this into TradeStation with intraday data and have found Excel easier to use because of the numerous calculations you can tease out and work with – I’m also more comfortable with Excel at the moment, though am working on more programming in TradeStation.