Hi Popov,

it would be nice if we would get a real brute force / genetic optimization possibility in EA Studio. Right now, the optimization just optimizes each parameter on its own (e.g. first ONLY stop loss, then ONLY take profit, then ONLY parameter 1 of indicator 1, etc.), but never in combination like a real optimization would do it. I understand that the current optimizer is fast, but with the current approach of optimizing just one parameter after the other, it will never be possible to exploit the full potential of a strategy. So my request would be: keep the current mode as default, but add another optimizer mode that allows us to either brute force test all combinations (and I do not mind if that takes weeks or months, as my server is anyway always running) or by adding a genetic optimization algo like most other platforms have it. That would be a great addition.

Thank you :-)

Hello Popov,

that is great to hear. Based on the other platforms I´ve been using the last 10 years, brute force (or rather genetic optimization which is a short-cut to brute force) makes a lot of sense. At least it allows the optimizer to not just optimize every variable on it´s own, but in combination - and that can make all the difference in the world for a strategy as you know :-)

I always save the settings along with the strategies. Once I´ve exported my strategies in the format Year-Month-Day Symbol Timeframe Number, I also export the Settings from the Settings page in the same format, so I always know what was used to make them.

Dear Mr. Popov,

First of all, thank you for all your great effort. It is amazing to see how your already fine tools keep constantly improving.

I would kindly ask you if it is possible to add "Minimum months on profit %" as a selection criteria in the "Sort collection by" field. I think that would be a very useful addition.

Thank you in advance and have a Happy New Year!

Best regards,

mw

Settings download should be under Settings like it is now, I find this confusing otherwise, but that´s just me. :-)

Hi Popov,

One feature that I see as a big limitation is not being able to move the OOS data, ie if we are using 10 years of historical data and 30% OOS so 3 years OOS. Be able to shift that OOS 3 years of data anywhere within that 10 year window. That way we are taking into account the most recent data for optimisation.
Also is there any timeline as to when the walk forward optimisation will be able to be used instead of the generic optimisation of only have 1 in sample and 1 OOS to optimise with?

Thanks

I started working on better multinational Portfolio Editor.

For now, you can import the Portfolio in the Validator and to recalculate it on the newest data.

I don´t think it´s a great idea for EA Studio. EA Studio is for generating strategies and there are already existing apps to track results. MT4 Tracker, or if you want it more advanced: https://strategyquant.com/quantanalyzer
EA Studio should concentrate on what it can do best: generating good strategies FAST. Let´s don´t overload it with stuff where there are already other good solutions, as it will slow it down in general and generating strategies will be slower as well then. Not a good idea, but that´s just my opinion.

What also would be good and most likely easy to add is a simple "Break Even" function, in addition to the SL and TP. Many strategies would benefit from this and EA Studio can find completely different strategies with this being available for the Generators usage. It would need one value: "activation level in pips" (e.g. after 70 pips in profit, set Stop Loss to break even) and it would be optional like the TP of course.

Of course, a better trailing stop would also come in handy. The current trailing stop is never used by any strategies by the generator - I have generated thousand´s of strategies on many symbols, all on 32 years of data, but never has the Generator found one that uses the trailing stop. This is because it´s always active, even if the trade is in minus territory. What we would need is a trailing stop with 2 options. "active after X pips" and "distance in pips" once activated. E.g. "trail after position reaches a profit of 70 pips, with a distance of 20 pips from thereon".

I think he just has TOO much on his to-do-list, as he works alone and has no coding-team. Sooner or later we will most likely see all these features and the pace for 2019 has been superb already - we had like 4 updates since new year already, if he can keep this up, then.... :-)

I am very grateful to Popov for everything he does, I can't think of anyone who works harder to give value to their customers.