Based on a strategy of this forum, i created a variant that seems to work too fine to be good.
You can see the results of the last 10 years attached (tick by tick mode enabled of course).
You may notice that I have overshadowed everything so as not to openly share strategy.
It seems to me that similar results are not available anywhere, so I prefer to treat it with lead feet because I would not want to lose the competitive advantage.
Also, I don’t share it because it has a huge problem as you can see: some positions remain open for months.
For this, I need the help of some good programmers to work in private (Paul, Grahal, Robo Future Trader etc … you are welcome), you can leave me your contacts (mail, telegram, skype) in order that i can share with you my idea.
On the other hand, it could also be an illusion, but trying does not harm 🙂
If this is not possible or you find ethically incorrect not to share an idea, you can close the topic and I apologize in advance.
No contacts and personal info can be posted.
Any request to contact someone shall be addressed to the admin using the contact form.
Thank you 🙂
I didn’t know. Sorry Roberto.
If someone is interested, you know how to do now. Thanks 🙂
Hello,
You are sure that it is not a boque.
Have you done a backtest several times?
hi Francesco,
Indeed the gain curve seems too good to be true … As an impression that the tick by tick mode is not activated in the backstest.
In my case every time I had this kind of curve, it was a bad estimate of the gains and losses on my part. Suddenly my “utopian” strategy on pivot points in real result was catastrophic.
If the stop loss and the take profit are on the same candle, PRT retains by default only the take profit: therefore the performance of the tests is largely overvalued.
Unfortunately difficult to help you more ….
A +
Florian
PaulParticipant
Master
@Francesco congrats on this nice find!
You’re not given us many leads here! One thing I notice that a long position is kept some time even the index goes down. Somehow it gives me the impression that perhaps you modified HammerNegated OIL 10min strategy.
tested that default one quick on the 4 hour timeframe saf with wrong spread in this case.
Anyway I fully understand you don’t like public sharing if you have something great!
@fifi Sorry but what do you mean with “boque”?
@florian I can assure you this is not without tick by tick. SL and TP are not part of the code.
@Paul It is not based on that, but it does looks good also. But only 65 position in 10 years is not the best.
I hope you can understand my concern for keeping it secret.
But as I said, it could be an illusion, and after a fix that doesn’t leave some positions open for months, it could become a very bad code 😀
By doing the backtest several times,
The result does not correspond to reality.
Very little loss in the ratio in the green and the percentage of position in the green too.
I’m wondering ???????
Tried several times, also in different sessions of PRT. Results remain always the same.
Reasons for ‘too good to be true:
- Because you have your stop and take profit order levels set too close to price.
- Tick by tick is not on when it needs to be.
- Your fixed spread in the back test does not truly reflect real life spread.
- The strategy allows massive draw down by using averaging down/up and just happens to be lucky enough to not open a first trade at the start of a plummeting market or massive new up trend.
- It is heavily curve fitted.
- It is actually a really good strategy!
As we don’t have the code or even the test result numbers we can only guess which of them is the reason for your ‘too good to be true’ feeling.
It seems huge performances are achieved by using a very small capital,
A drawdown of 25,000 (+ required margin) currency units would require 100,000 units as a starting capital not 6,000.
By adjusting capital, performance drops to less than 1/15th.
@robertogozzi Can you please give me your email? Would you like to take a look at the code? Thanks.
Sorry, I thought you were an admin. I will try to reach you through the contact form 🙂
Use contact form at https://www.prorealcode.com/contact/
to ask the admin (Nicolas).
Seems to be an averaging down strategy, at first look of the equity curve? Or orders are not close until they reach a new equity curve high? In this case, that kind of results seems very ordinary, especially for long only strategy on index. I’m curious about the drawdown of the order opened for almost 10 months between mid 2016 and first quarter of 2017? 😉
