Mother of Dragons trading strategy…
- This topic has 522 replies, 50 voices, and was last updated 3 years ago by
.
-
@StingRe, @josef1604, as I’ve explained elsewhere in this v long thread, rather than the typical ‘buy low, sell high’, MoD works on a ‘buy high, sell higher’ basis – ie it will only enter when the run is well established. Backtest shows that 9 times out of 10 there is enough continuation to turn a profit. This does mean however that it can also buy at the very top of a run (or sell short at the bottom) – and when you see that happening it looks colossally dumb, like only a moron would enter there. Obviously there is still room for improvement in this regard, but I have yet to solve it. Personally I’m at the point where I just accept that sometimes it will do dumb things, but in the long term it tends to get it right. Hope that helps.
I am running DJ v5.1 and NAS v4.1 L & S
5 users thanked author for this post.
Thanks for all the work you put into this strategy Nonetheless!
I had a different idea for a robustness check and your strategy is perfect to test it on.
When in a position, you don’t see normally the same signals. To have them reflected in the equitycurve I used cumulative position orders = true.
However the main difference is normally a strategy uses a set ( stoploss )command, tradeprice(1) or positionprice for exits.
Now it’s all based on the first position, so i.e. stoploss exits are the same but coded differently and all other references to above too but that was changed easily.
The purpose is to let the equiycurve intact or improved when acting on all or selective signals but keeping the total countofposition = same signals relatively in low in backtesting & optimising and having more insight on the number of good & bad signals and improve on them.
Additonally I’ve added the reenter when same signals come as default (not using cum.orders), every-time, when losing or when winning.
Here’s a comparision on the code v5.1, I added r.. to reflect below
// reenter [1] (default) no action on same signals until closed
// reenter [2] action on same signals when in position
// reenter [3] action on same signals when position is in loss
// reenter [4] action on same signals when position is in profitSince the code doesn’t use a hard stoploss command it’s not recommended to use live. Risks are higher i.e. when adding to a winning position all positions have the same exits from the first position, so if it fails completely, all next signals (=new positions) have a higher losses.
6 users thanked author for this post.
I have just discovered this very long topic, and only one word (or 2) : THANK YOU to all of you.
I am learning so much with this topic
thank you again
1 user thanked author for this post.
This does mean however that it can also buy at the very top of a run
As we have this exact situation right now, here’s one way of dealing with it:
Yesterday DJ v5.1 opened a long position at 29988, just 100 points away from the all-time high. As this trade seemed likely to fail, when it started going south I manually opened a hedge for the same amount. These are presently both still open. The short has a breakeven stop and a limit equal to the stop on the long position. So if it does turn around and make good, the short will close for no loss. But if, as I expect, it falls all the way to the stop then the short will cover the cost. It’s a way of letting the algo run it’s course while having an insurance policy against likely failures.
I’m sure that most people already know this trick, but for anyone who doesn’t …
This does mean however that it can also buy at the very top of a run
As we have this exact situation right now, here’s one way of dealing with it:
Yesterday DJ v5.1 opened a long position at 29988, just 100 points away from the all-time high. As this trade seemed likely to fail, when it started going south I manually opened a hedge for the same amount. These are presently both still open. The short has a breakeven stop and a limit equal to the stop on the long position. So if it does turn around and make good, the short will close for no loss. But if, as I expect, it falls all the way to the stop then the short will cover the cost. It’s a way of letting the algo run it’s course while having an insurance policy against likely failures.
I’m sure that most people already know this trick, but for anyone who doesn’t …
Too late for me…. i am waiting for the conclusion of this trade….
Same for me, it’s now closed… I hope the next trades will be better.
Thanks again to all the contributors on this amazing topic.
Anyone running the NAS 5m MoD v4.1 S here ? I launched it without MM (MM=0 & positionsize=1), and it took a position on Nov 3rd, at 06:00:00 UTC, which hit stoploss at 15:15:27 UTC. However, this doesn’t appear in the backtest.
I reached PRT out, and I am now waiting for their answer.
I read that @ had a similar issue a few days ago with the DJ version.
Regarding this issue I mentioned a few days ago, PRT investigated and told me it was due to a problem related to MTF synchronization. They will try to fix it within the next few weeks.
2 users thanked author for this post.
Find exclusive trading pro-tools on 
