Topic: Expert Advisor Studio - Wish list

Dear Traders,

Please post here your ideas and requests for improving Expert Advisor Studio.

Trade Safe!

Re: Expert Advisor Studio - Wish list

Spread level protection. On other topic i wrote but adding here also to keep all wishes under this topic. Also can you tell me please when to expect this function ready approximately? Without this function actually i dont want to run created strategies as i will not get the ideal results.

3 (edited by santaferes 2016-07-27 21:55:57)

Re: Expert Advisor Studio - Wish list

1-Report/Indicator Chart - Fast scroll left-right with scroll under chart or with page up/down buttons. Scrolling chart with only mouse now and it is too slow.

2- On charts when mouse over chart better to see dates

3- For created strategies stagnation duration i couldnt see. Better to ad it also.

4- Option to add our own custom indicators if possible.

5- Multi currency tester for strategies.

6- Monte Carlo test function for strategies.

For now thats all smile

Re: Expert Advisor Studio - Wish list

Acceptance criteria: with option to include/exclude OOS results.

Re: Expert Advisor Studio - Wish list

How able creating "Stealth" stop orders function?

Re: Expert Advisor Studio - Wish list

Dear Popov,

Could you please add Dukascopy tick data as addtional historical data source just like FSB pro?

Regards.

I’m working on improving my English and trading skill

Re: Expert Advisor Studio - Wish list

Hi
Export to Fsb.

Thanks Pit

Re: Expert Advisor Studio - Wish list

Forget 50,000 bars History

Thanks Pit

Re: Expert Advisor Studio - Wish list

Please add "max stagnation" on the sort record.

I’m working on improving my English and trading skill

Re: Expert Advisor Studio - Wish list

Hi Popov, could you please add "risk & reward" ratio on the sorting option?

Regards.

I’m working on improving my English and trading skill

11 (edited by alsilva86 2016-10-23 22:56:20)

Re: Expert Advisor Studio - Wish list

Acceptance criteria: with option to include/exclude OOS results.
Forget 50,000 bars History

An option to create not only an EA but a XML, so later we can manage it on FSB desktop.

Additional dynamic profit things:

Leave position if signal is reverted but profit > X

12 (edited by Freesby 2016-10-24 09:25:59)

Re: Expert Advisor Studio - Wish list

- By clicking on a row of the Journal, directly opens Indicator Chart window to the exact point of the selected trade.

- Conversely, clicking on the candle or the right menu of the indicator Chart window, opens to the Journal

- Put automatic function (selectable in the on/off preference) to create lists of repositories with automatic name (first compound) eg:
Cross = EurUsd
Time Frame = 4H

==== EurUsd_4H

Re: Expert Advisor Studio - Wish list

I ask to add the line "Custom Order Comment" in the input table of the exportet EA (the same as the FSB exported EA's have). This would make life much easier when several EA's are running parallel and it so is possible to give them individual names that I can see in the MT4 Comment-column.

Re: Expert Advisor Studio - Wish list

Insert an arrow, to clear the entire column of additional statistics, portfolio, explorer without having to close and reopen the columns individually.

http://s22.postimg.org/89r99vwjh/Schermata_2016_10_27_alle_09_56_53.png

Re: Expert Advisor Studio - Wish list

Thanks for providing this software for so long, I have had tremendous fun and learned a lot by using it over the past several years. I'm not trading, so the pro version doesn't make sense for me, and the old 2014 3.8.2 version is as much as I need for playing around, though sometimes I'd like to be able to do contingent strategies and strategies whose parameters change in response to other signals, and I'd like to specify different strategy ranking criteria, e.g. Sortino rather than Sharpe ratio (Sortino uses the downward variance only, Sharpe penalizes rapid gains as being upward volatility).

One big thing missing from EA Studio is even primitive money management. The ability to set all the parameters that FSB version 3.8 had under "Strategy Properties", sub-frames  "Handing of Additional Entry Signals" and "Trading Size" is really essential. The stop loss and take profit are much less important, in fact I never had a strategy improve performance when using them. Setting lot limits is useful, it would be even better with a soft limit that reduces trade size more and more as one approaches the lot limit.

Being able to have the next opposite signal reduce one's position allows constructing the absolute simplest strategies that scale in and out of positions and never actually close at all (actually, a "never close" closing point condition should be added - I fake it by using a condition that never happens). For the pro version, at least, having Kelly Criterion money management with settings for initial "edge" and confidence in that edge, and for fraction of Kelly amount desired. This allows the system to bet less in market conditions when it's lost its edge. Martingale multipliers less than one should have a similar effect, but less sophisticated.

I'd really like to see the ability to make custom value criteria for judging strategies. One I was thinking of would compute for the account equity on each successive candle the average rate of return since the start. This would optionally be weighted by a window function such as a simple ramp, so that rates of return count more towards the end of the period.  The weighted rates of return for each candle are then averaged (geometrically would likely be the right way) to produce the figure-of-merit used to judge the system's quality. It automatically penalizes risk without over-penalizing it. Using different window functions would be fun to play with, too.

I'd like to be able make a top-level strategy that senses market conditions, perhaps taking human input into account, then based on that allocates funds among a virtual portfolio of strategies tailored for those conditions, for instance ones for upward, downward or neutral markets, that do better in high or low volatility. The output of the strategies would be weighted by the amount of the portfolio invested in each strategy, and summed to produce a unified order flow. This could be extended to a portfolio of such strategies for each of many currency pairs or other assets, then having a portfolio of portfolios, (with correction to investment weights for correlations between different pairs, etc.). When combined with Kelly criterion portfolio allocation this could be a very robust and profitable technique. (Ed Thorpe wrote a paper or two on  Kelly criterion portfolio allocation, but the leveraged case isn't well understood, by me at least).

Another thing I'd like to see is specifying the parameters for indicators in time units rather than candles, which makes porting a strategy to different timescales easier. Candles are really a bit of an artificial view, though it does cut down on the computing cycles. Having time-denominated rather than candle-number parameters is also useful in minute-candle strategies where the cycles one wants to react to may be 12 or 24 hours, way above the 200 candle maximum.

Re: Expert Advisor Studio - Wish list

in the "optimizer", instead of only the only 5 options to reset indi values, why not adding an option to set the max and the minimum of each value of each indi!( ex: 0 & 200)

Re: Expert Advisor Studio - Wish list

E.Harris

Are you referring to the online Expert Adviser Studio or to the desktop version.

For all of these additional calculations, you might need to be using MatLab for testing. Not all users have the size computer required to make all of these calculations..

I can assure you that even without these additional tools and sophistications, that FSBPro does a nice job on creating strategies for Forex and Stocks and Indexes.

Personally, I have made a portfolio of many strategies that I am using constantly and while I do not use a lot of complex techniques I do get what I consider to be a nice return.

I actually appreciate the simpler strategies as they are easy to optimize.

My 'secret' goal is to push EA Studio until I can net 3000 pips per day....

Re: Expert Advisor Studio - Wish list

Hi E.Harris

I prefer EAStudio generated EA to be simple, just like this is totally okay. If I want to make more complex strategy I'll just use FSB Pro

What I really like from EA Studio is very fast generating and backtesting process, the generated EA is also flawless and very light.

I’m working on improving my English and trading skill

Re: Expert Advisor Studio - Wish list

Blaiserboy wrote:

Are you referring to the online Expert Adviser Studio or to the desktop version.

For all of these additional calculations, you might need to be using MatLab for testing. Not all users have the size computer required to make all of these calculations..

I can assure you that even without these additional tools and sophistications, that FSBPro does a nice job on creating strategies for Forex and Stocks and Indexes.

Personally, I have made a portfolio of many strategies that I am using constantly and while I do not use a lot of complex techniques I do get what I consider to be a nice return.

I actually appreciate the simpler strategies as they are easy to optimize.

The lack of ability to have the next signal add and next opposite signal reduce the position is in EA Studio, the old freeware version could do this, and it made for extremely simple, robust strategies using a channel or bands -type indicator - see my recent post in the trading strategies section.

The other ideas were for FSB Pro most likely, though it would be great to have them in EA Studio.Yes, no doubt FSB Pro does even better than the old freeware version, but I'm not trading, so it isn't in my budget.

I strongly agree strategies should be as simple as possible if they are likely to continue working, on the other hand there are strategies which work very well in certain market conditions, but don't totally die in other situations, everybody has lots of strategies they could be using, like golfers have lots of clubs, or even better like fishermen have lots of lures. If you can choose your strategies to match upcoming market conditions with even a bit of statistical edge, you can do much better than with a single strategy with constant settings.

Money management is more than half the profit in a strategy, and complexity in deciding how much to bet is much less likely to be curve fitting than complexity in the indicators. (Of course if you used indicators directly in money management you could be curve-fitting, but basing bet sizes on the track record of the strategy, especially the recent track record, is not actually adding any signals, it's only modulating what the basic strategy itself is signalling. Usually this means in practice that when the strategy makes lots of bad trades in a row the strategy progressively reduces its trade sizes or even suppresses some signals when the optimal trade size is below the minimum allowed trade.  ( bad trades: more exactly, when correlation between the most recent trading signals and later price movements falls, the trade-oriented view may obscure what's really happening in an ongoing, but fluctuating position - in fact the equity is the only thing that matters, the balance isn't important).

I have a strategy with a profit factor of 29, a Sharpe ratio of 1.1, yet it had 50 losing trades in a row. If I could tell it to cut its bet by 10% after every losing trade and to increase again after winning trades up to some maximum, it should perform even better with lower risk.

The Kelly criterion is the mathematically proven bet-sizing rule, but it needs to know the edge and the odds - the better the edge and the odds, the more you should bet, but these are not perfectly known or even stationary quantities in trading, so betting a certain fraction of the Kelly amount from 1/3 to 2/3 is usual.

Re: Expert Advisor Studio - Wish list

Good morning Miroslav,
as you know this is my first month using Ea Studio, I didn't run any strategies on a demo forex account but I spent many time on EAstudio, and then backtested strategies with MT4.

Your tool for me has high quality strategies generation process in a very short time. Anyway, here's my improvements list, the first 3 should be more important for me, thanks for reading:

1. It would be great if your tool had an optimization function like this: improve strategy by adding a new entry rule, and then it improve strategy just by browsing the best additional indicator and the best settings for it. This is very time-saving! Given that there are around 40 indicators with many entry rules for each of them, adding manually an additional entry rule is very boring and time-consuming.

I saw many times that if Ea Studio creates a good EA with one entry rule even after optimization, net balance increases a lot by manually adding an additional entry rule. I remember a day net balance was around 12-13.000$ and by manually adding a 2nd entry rule it arrived at 16-17.000$.

2. increasing "maximum data bars" inside "data horizon" section. For timeframes M1, M5, M15 importing 200.000 bars in my opinion is too small because initial data is May 2016 for M1 strategies, March 2014 for M5 ones and October 2008 for M15 data. If starting period could be previous to these, surely it will require more time to create a better strategy but mostly for M1 and M5 they will be more accurate (6 months period and 2 years period seems too short to me, in fact I run some M1 and M5 strategies in Backtest on Mt4 on previous years but they didnt' work well, lot of drawdowns and high stagnation.

Instead, a strategy created on H4 timeframe worked very similar in MT4 Backtest since 2001.

I don't know but maybe 5-6 years of data for M1 timeframes (around 2 millions bars) and 10 years for M5 (around 1 millions bars) could create better strategies in the long term.

3. knowing what are best settings for generating strategies the first time, if for best "net balance", or share ratio, or profit factor and so on, given a specific timeframe and currencies couple.

4. saving some EAs directly into the online account, if not unlimited but at least one for every currency couple and timeframe (one for EurUsd M1, one for EurUsd M5, one for EurUsd M15 and so on);

5. the button "Share strategy" doesn't create a functioning link: here's one of mine but it open forexsb.com homepage:

https://forexsb.com/expert-advisor-stud … yKji3MueyM;

6. the opportunity to change the "initial account amount" directly inside "Expert Properties" window, above "entry lots";

7. additional leverage values inside "settings" area: 1/600, 1/700, 1/800, 1/900, 1/1000

8. adding page1, page2, page3 inside "collection" menu for browsing all the strategies;

9. inside "Transaction Log" window under "Report" menu, the opportunity to order the journal based on "Profit" column, so that we can check best and worst trades made from the biggest to the smallest.

10. inside "balance chart", having time data on the orizontal axis (Weeks, Months and/or Years), so that it's possibile to see balance increasing in which period.

Sorry for mistakes in writing and congratulations for this very interesting online tool,
Angelo

21 (edited by yonkuro 2017-01-15 13:12:34)

Re: Expert Advisor Studio - Wish list

I wish for an option to search and optimize for the best stagnantion and balance line deviation.

Also please add a feature to delete and rename the historical data.

I’m working on improving my English and trading skill

Re: Expert Advisor Studio - Wish list

Label the strategy collection with a unique identifier. Perhaps autosave collection every few minutes to protect against power failure.

One day, in some circumstance, make compatability with FSBPro.........., not a high priority.

My 'secret' goal is to push EA Studio until I can net 3000 pips per day....

Re: Expert Advisor Studio - Wish list

Enhance Workflow capability by allowing it to work with collected strategies. The point would be to maintain selected strats from collection over wider period of time, essentially managing/updating/improving the strat portfolio. It should do all those things it does now except generation. In other words, having Workflow additionally in the collection.

Re: Expert Advisor Studio - Wish list

Would it be possible to have a 'Delete Button' on the 'Editor'?

Reason being.... as we go through the collection to see if the strategies are suitable for export, it would be handy to be able to delete the unsatisfactory ones.

Thanks

My 'secret' goal is to push EA Studio until I can net 3000 pips per day....

Re: Expert Advisor Studio - Wish list

Would it be possible to have a 'Delete Button' on the 'Editor'?

You have it:

http://s23.postimg.org/ym3hehkpn/screenshot_2017_02_10_at_12_41_38.png