Sometimes once I have what appears to be a working strategy I like to start removing things to see what true effect they are actually having. In this strategy I have three entry conditions:

c1 = High > highest[p-1](high[1])
c2 = (close - low) / (high - low) < x
c3 = (my secret entry condition that no one has guessed yet!)

So one by one I take them away and run a test – here are the results:

c1 and c2 and c3 – our datum.

[attachment file=”c1 c2 c3.png”]

c1 and c2

[attachment file=”c1 c2.png”]

c1 and c3

[attachment file=”c1 c3.png”]

c2 and c3

[attachment file=”c2 c3.png”]

c1 only

[attachment file=73924]

c2 only

[attachment file=73925]

c3 only

[attachment file=73926]

The results seem to confirm that each of the entry conditions on its own is a pretty good entry criteria. It might be interesting to try other combinations such as (c1 or c2) and c3 etc in the strategy.

So some new ideas to test now.

Also while doing this I had a thought about my counting of runs of bars without a close above the last high and whether that would make a good entry filter as well as a reason to exit. So I ran the strategy with one condition that said if the run of bars was below the average then buy and another if the run of bars was above the average then buy. Here are the equity curves:

Less than average:

[attachment file=”c4 less than ave.png”]

Greater than average:

[attachment file=”c4 greater than.png”]

It seems that a below average run of bars with close above the last high might be a fairly good entry filter but I feel that four entry conditions in one strategy might be one entry condition too far and seriously limit the amount of trades. Still it is another idea to play with.

c1-c2-c3.png c1-c2.png c1-c3.png c2-c3.png c1.png c2.png c3.png c4-less-than-ave.png c4-greater-than.png