Monday August 1st Weekend FX gaps, page-9

Hi pinozi

I had a look at the AUDUSD data set, which totals 181 out of the 420 data points (31 gaps this year & 150 gaps prior to this year).

In this data set, 171 gaps (94%) close within 4 days, another 1 in a 5th day, and another 2 in a 6th day. Of the remaining 7 gaps closed, 2 more closed within 20 days, 4 of the last 5 within 6 months and the last 1 took 3 years.

Based on the above I rationalised it is better to exit still open trades after day 4 than day 5 - trading off 1 more profitable trade for smaller stoplosses on the losing trades. One observation looking back at the charts for those gaps, when the gap wasn't closing after 4 days it was typically still opening on the 5th day and then of course there is the gap to the following monday (day 6) which could open it even further. Hence a 4 day close-out seems the least risky to me at this point.

To model the outcome, I assumed the following:

1. 1 position taken at gap open (only if > 10 pips)
2. 1 additional position taken targeting at least the average of the gap open price and the max open price during the 4 days
3. Calculate the average entry based on 1 & 2
4. Calculate the profit on all gaps that close within 4 days based on the above average entry price
5. Close any gaps still open at the end of day 4, and calculate the stoploss on each of those trades

From the data set the results were (all trades same sizing):

1. Of the 181 data points, 104 opened with gaps > 10 pips, so 104 tradeable gaps.
2. Of the 104 tradeable gaps, 94 closed within 4 days and 10 did not close within 4 days
3. The profit in pips from the 94 gaps that closed within 4 days was 3319
4. The stoplosses from the 10 trades closed at the end of the 4th day of each trade was 1185 pips
5. The profit/loss ratio is 3319/1185 = 2.8:1
5. The net profit from the data set was 3319 - 1185 = 2131 pips

I could re-calculate based on a 5 day exit or 6 day exit, but thats a bit of work & I think the swing either way will be only a few %. Rather than do that my next step is to validate the results from the above data set by calculating the results of the same rules using the 240 data points from the other currency pairs from this year, and see how repeatable the result is.

Finally, the above doesn't of course preclude monitoring those gaps that remain open to identify possible turning points for another entry attempt to close it & recapture some of the stoploss (since they all do eventually close, according to the data set).

Cheers, Sharks.