Stretchy Slope. Blame Bunker for this one.

[feucht]

Same here Pipley, and then overnight now a buy on GBPJPY. Bit dull at the moment I know, but hopefully for a good reason!

[feucht]

Update up to the end of last week attached. With the default settings, loss of about 10-15 pips off one new trade (USDCAD), on other settings a pretty bad week, particularly if running with a tightish SL. Let's hope for a bit more action this week!

[feucht]

4 closed trades this week, all winners, 200 pips in the pot

I'm currently looking at separate CloseOnContraction targets for trades in profit and in drawdown, early signs are showing we might be able to squeeze out some more profit and reduce risk...

J

[Pipley]

This is great to "hear"! I only had two since Feucht5 (AUDJPY and AUDNZD), but they were both good trades. I just realized that I had removed Stretchy from several of my charts while transitioning and forgot to turn them back on. That is probably why I only got two trades. Oopsa!

Regardless, this is looking much better than last week now that you figured out the CloseOnContraction thing!! Thanks again!!

[Pipley]

Feucht, I am wondering how a M15 / H4 (1/16 ratio) would work as compared to the Original 1/24 ratio Stretchy. At the risk of looking real stupid, I will ask if it is possible to use the data that you've collected to test that idea.
I am sure that the added noise will decrease the win/loss %, but it might be an option to diversify the risk over more pairs. (A M5 / H2 would be the best ration match, but the noise would be deafening!)
Just a thought from my impatient mind.

[feucht]

I would have to start collecting the M15 slope values going forward to support testing this, which I can easily add in to the EA. Obviously it would be a while before I had enough data to be able to make much of a judgement about it's worth or not. I think the pivot filter has made a big difference to the stability of Stretchy, and I would have to look at whether another pivot relationship (day vs week?) might need to be taken into account.

As discussed above I have been looking at having trades close at different stretch contraction levels depending on whether they are in profit or not. Attached is a table of results, across the columns are the contraction levels for trades in drawdown, and down the rows are those for trades in profit. We seem to be able to get over 200 pips more by closing trades in DD at -0.1, rather than waiting for them to hit -0.4 (which is the sweet spot for trades in profit). At that point we are also looking at over 6 profit pips for every loss pip, which is great for risk management.

I will also look at the crossover point between the 2 contraction level targets, would it be better to apply the "profit" target on trades that were over 10 pips (or 20 or 120!) in profit.

[Pipley]

Wow! Again! If your findings were a few weeks of trending market data they would still be exciting. The fact that these represent 10+ weeks of strange markets is...... well..... really, really exciting. (I'll never be a creative writer!)
I wish I could contibute somehow! Is there a way to use a simulator (or something like it) to collect past data to plug into your calculations? I'm guessing that if there was a way, you would have done it by now, but thought it was worth a mention.

[feucht]

There may be a version of TMASlope that gives you what it was at the time, rather than the standard one that can change a bit (repaint) after the fact. I'm sure that there was some discussion about it in the early 10.2 days. If I could get "real" M15 slope values even at a M1 or M5 resolution (a single value per M1 or M5 bar), that might be enough to get some simulations rolling...

[Pipley]

Okay. I will do some reading this weekend and see what I can find.

[feucht]

I am on it, Pipley, looking at Baluda's indicators, he has an option to "ignoreFuture' when calculating TMA slope, and I have managed to generate M15 values (at a M15 resolution of course, I was talking rubbish before) going back to early this year. I am collecting M15 values tick-by-tick going forward now. The old stuff won't be as accurate, but hopefully good enough to get a feel for whether this is a goer or not.

At this rate I will need a new PC just to house and process the database, it is getting huge!