This is a reasonable request. I'll add controls that will set how many slots to be generated by the Generator. (This work well in FSB Pro, so I'm going to make it in the same way.)

I have no Idea to that, sorry, but can you say me, what you make when strategies no more good, that you make this old strategies good? Maybee Optimization? Or what make you then?

Hi Popov,

I'm in favour of having more opening and closing rules.  Can you kindly increase the number of rules available for the EA Studio to generate the rules (with the option to choose how many max/min slots a user wants).

The reason I need more slots because I believe to build/search for robust EA, we need more slots for 2 types/components of rules.

Currently there are only max 4 opening rules and 2 closing rules available for EA Studio.

Over the years of building EA, I've come to realised that a good EA need 2 MAIN components for it to be robust.

A robust EA's rules need to consist of these 2 Main Components.

Component 1: Rules to Trigger a positions (Rules for making money)

These are the rules that identify the trends/conditions to enter.  These rules formed the broader framework in which it allows the EA to find as many opportunities to trade.  These are the signals that often play the part in helping the EA make profit.

Component 2: Rules to eliminate false signals (Rules for not losing money)

These are the rules to restrict/limit the 1st component rules to trade, in another words, it aims to eliminate trades during consolation/ranging/ambiguous trading conditions.  In another words, eliminate false signals, prevent your EA from losing money.

In a nutshell, a good EA is one that not only make money but also that doesn't lose (too much) money.  Often the main reason our EA fail is not because it didn't make money, but rather we having difficulties finding ways to stop it from losing money.

Of course lousy EA are those that not only losing money but can't even have a nice decent gross profit.

If we were to examine the EA's PF value, having a value of 2.0 and below simply means that our EA losing (Gross loss) majority of what it has earned (GP) and hence we ended up with a small Net Profit that resulted to having a PF value of eg. 1.3

Hence, having extra rules to stop the EA from losing what it earned is far more effectively ways to improve our trading rules than to constantly search for new EA and given a limited opening/closing slots, the type or quality of robust EA is hence limited.

Hi Popov,

Can I please request either a snippet of code or better yet the ability to have individual trading times for each day/the ability to not trade certain days. I know this feature is there for Sunday and Friday but doing both back testing and live testing I’ve found plenty of my strategy’s perform better on certain days than others.

Could I please request more advanced position sizing. Ability to set limit orders and laddering in and out of a position which is fully optimisable.

Also the ability to optimise a strategy through multi market ie we build a strategy on one symbol then can choose to optimise over say 5 symbols to give a strategy which is profitable over a number of symbols to help with curve fitting. I understand the performance will be compromised but it would give a more robust strategy.

Also are there any updates in the bot to distant future with new features?

I strongly support that suggestion. For example if we could adopt ATR based position sizing techniques such as the fixed fractional method, then we would find it far easier to hone in on those solutions that may be suitable for multi-markets. These volatility based methods would also need to be applied to profit targets and stop placement for the generator and reactor.

Unfortunately the current methods employed by EAStudio significantly limit the capability of developing robust EA's capable of multi-market use.

I remember in previous posts within this thread that this request was being considered and was in the development pipeline but I have not heard of any recent updates on this request.

Popov, have you an update at all on this wish list request?

The fact you use a backtesting engine, and filtering and sorting criteria, means that you introduce curve fitting.

Whatever goal you set to the app, it will try to find a strategy that best suits it, by fitting trading rules on the provided data.

The opposite may be right - the fewer criteria, the better strategy. And Monte Carlo for robustness testing.

It would be nice to have the option to turn on/off certain trading days or hours of the trading day. Some strategies perform poorly on Mondays or Fridays or during certain hours of the day.

Request: Walk Forward Module, option to run a 'pre-live' optimization stage to get parameters for live trading

@Community and Popov
My two cents on what is wrong with current walk forward approach

The way it is now the user is selecting parameters that works only for the last segment (IN SAMPLE 5) not for the actual trading period ( last in-sample 5 + last oos5 ).

If you select a walk forward with 5 segments, it should do a additional "pre-live" optimization using the ( last in-sample + last oos) to inform the user the parameters for live trading.

Hope this make some sense, if you agree reply and help to make this walk forward module something priceless!

Keep the excellent work Popov, cheers

Hello everyone,

Given that the data are so important to create a good strategy it would be great to have more information about the quality of the uploaded data,(not only the maximum days off), for example:

-How many gaps with more than X pips have the data and how many days off (or hours)  have these gaps

-Where are this gaps, date and time

-A reliable source of data with many years of data (Metaquotes had a lot of huge gaps), integrated in Expert Advisor Studio

Thank you Mr. Popov

Hello

My continuing stumbling block is getting reliable historical data.

If we are to use our brokers data (rather than Tickstory or equivalent), I would like to be able to feel reassured that it is good enough for creating strategies as quickly as possible.

I have seen that there have been some updates to Data Statistics tab under 'Data' in the header menu, in particular, the Maximum Gap and Maximum Days Off statistic readings, which is great, but I wonder if this could be improved further by identifying exactly what "from/until" dates of data are missing?

To go further than that, could EA Studio offer to cut off any data that is before any gaps, giving a nice chunk of accurate and consistent data to work with?

Thanks.

Hi Popov,

I would love to see the ability to move the window of OOS data from the most recent date, ie having a sliding window of OOS that the user can pick.