Taking Vonasi’s suggestion to try this long-only, I broke it into 2 separate strategies, long and short. With each optimized for just one task the result seems better on every metric. It may have the added advantage that in an uncertain market they hedge each other – would require forward testing.
Image at left is the combined performance in one code. When you map the other two onto each other, every month where the long has taken a loss the short has more than made up for it. That doesn’t happen when combined.
Unfortunately, in this case, it doesn’t get around the fact that it is probably curve-fit to the last 100k but i think it does make a good case for separating strategies into long and short.
Can’t believe I hadn’t tried this before. Is there ever a good reason for having one code go both ways?
Original code is here: https://www.prorealcode.com/topic/mother-of-dragons-trading-strategy/page/2/
Hi nonetheless,
Are you able to share the code for the long / short strategies?
Thanks!
We have a new “vectorial”.
Publicate 10 days ago.
No comments in 10 days.
We see???… Or sleep???
Hi @nonetheless
thank you for the code
I have had a quick look at the results and have some observations that I would welcome your thoughts on:
Based on the attached image it appears that if you are able to get the trailing stop to kick in then the strategy will make money but if it just draws down then a loss much greater than the return of an average trade tends to occur, have you experimented with a closer hard stop/alternative stop?
Given the mix of a long term average and a short term signal it looks like you have occasions on which the trade is stopped out but then opened again promptly as you are looking for a market condition ie x is above y rather than x cross y. Have you looked into altering the moving stop so that they would not get stopped out when the conditions still exist or perhaps having a condition that prevents a quick reopening?
I do like that you have success on the short side on a market that has been trending up for so long
Thanks
Ruark
looks like you have occasions on which the trade is stopped out but then opened again promptly
Hi Ruark, thanks for your suggestions – you’re absolutely right, i had noticed that as well but didn’t know how to fix it. Take a look at Fifi’s version in the other thread (you’ve probably already seen it) – he added updateonclose to both the 2h and 15m TFs which if I understand correctly should help exactly that problem.
I have posted another version, but I think his is the one to go for!
Hi, im referred to this Mother of Dragons.
No to others Mother of Dragons.
This Dragons so good
Separated estrategies, please.
Sleep???
@Fran55 — glad to hear it’s working for you! Try this one, NASDAQ 5m, doing quite well in forward testing over the past few weeks.
Hi.
I think Mother of Dragons in two separated strategies… Up and down.
You have the down versión yo Nasdaq???
Good job, friend… I like the versión to Dow.
?thats really in future?…
No, I never bothered with a short version for NAS but you should be able to adapt the LONG and reset the variables quite easily.
@Fran55 The people don’t understand! 🙄 😆
The people don’t understand
… that Daenerys is actually alive and well! 😂