I think that the controls you have installed re margin etc, are going to allow me to work with fewer strategies, This week I will try to cut down to 50 or so, or at least generate better quality strategies that will allow that.

I am more interested in having a few that work well consistently, easier to manage position sizing.

That being said....... I am quite disappointed that you omitted a happy face. lol

daveM

Hi Steve,

A million thanks to you for such wonder software, a great time and effort saver.

I have set up 41 MT4 Terminals and run more than 3000 EA in a matter of hours. 

Without Portfolio Maker, I wouldn't even want to attempt to run such large number of EA,

imagine setting up more than 3000 EA, one MT4 chart at a time!!! That's madness. 

But with Portfolio Maker, all I need to do is just install 41 MT4 charts, what a breeze

Here is just some suggestions, if you would kindly consider

1. Add BE, SL, TP inputs, so that we can test the results if we were to change any of these settings. 

Often I'm interested to test what's the outcome would be like if I were only change the BE or SL settings and re-test these EA again to see any possible improvement.

Hence, instead of going to each individual EA to make these changes, if you were to include it as part of the changes we can make the changes across all the EA in the Portfolio, that would be great.

2. Add the option to use Trailing Features, a simple trailing is suffice if it is not too much a trouble.

I understand you may want to let Portfolio Maker be where it is now. 

But if you are willing to consider again.  Do hope you would take a look into this.

A lot of FSB users wish we have more than 1 Closing/Exit Strategy.  In the past there were lots of hope for the Profit Protection Features we were discussing and for some reasons, this project was shelf aside.

In addition to our exit conditions, we often realised that if we were to add in a trailing exit, we could have a better and more dynamic exit (if for some reasons, our original closing conditions were too slow to respond to dramatic changes in the market).

3. Of cos there are lots more features we can imagine.  And whenever you are open to discuss ways to make Portfolio Maker an even more excellent tools for FSB Traders developmental needs, I believe a lot of us will be very willing to come forth to contribute more ideas.

Here is an overview of what an EA development Process is like

1. Strategy Creation - FSB software

2. Strategy Testing - Portfolio Maker (by Steve)

3. Strategy Analysis - MT4 Tracker (by Steve)

4. Strategy Analytic Trade Management - (by Steve) Future, upcoming product when Steve going to use performance metric to determine which EA to turn on and off.

Do you realise Steve, you are going to own the remaining EA development process after we are done with Stage 1 Strategy Creation?

So do take some time to think about it and hopefully you realised that you could turn this whole project of yours into a very power and useful software package.  And I believe you probably has already begin your planning process and just waiting to surprise us again and again with your awesome products.

Best Regards
Hannah

Hi Alamehmazen,

I've recommended a software in the post here https://forexsb.com/forum/post/47885/#p47885

It make changing file names in a folder or bulk a breeze.

Hi Steve, I am having problems when trying to compile the EA.  I get numerous errors - I have checked the correct MT4 folder as I have about 4 versions of MT4 available.  I have attached a screen shot of the Meta Editor.  I think I have followed your instructions correctly.

FSB-Pro-generated strategies bundled in a portfolio EA (using Portfolio Maker) trade less frequently than their twin strategy in its own chart.  alamehmazen123 and I reported this observation previously.  Also, I've posted about this before and also added an FAQ to the mt4tracker.com website.  This does NOT apply to EA Studio-generated portfolios -- just those created by Portfolio Maker.

The image above comes from MT4 Tracker.  In this particular test there were 20 strategies.  The even-numbered strategies were bundled in a portfolio EA created by Portfolio Maker.  The odd-numbered strategies were attached to their own chart.  For example -- magic numbers 11000 and 11001 are the same strategy, where 11000 is bundled in a portfolio EA and 11001 is attached to its own chart.  11002 and 11003 are also twins -- with 11002 bundled in a portfolio EA and 11003 attached to its own chart.  And so on...

As you can see, the even-numbered strategies (bundled in a portfolio EA) trade less frequently than their twin strategy attached to its own chart.  Furthermore, as you go down the column, the trading for both declines.  I recently posted my theory as to why this happens (https://forexsb.com/forum/post/48265/#p48265) -- I'm just speculating -- I do not know for sure.  Briefly, each chart runs in its own thread and performs all its calculations and (and sometimes trading) with each OnTick it receives.  There are a limited number of CPU time-slices available and MT4 has more important things to do (e.g. download data from the broker, send / receive orders).  I think the strategies generated by FSB-Pro are more complex and require more processing time than OnTick was designed to provide and its time slice is cut short.  There are other threads with higher priorities.

To summarize:

1. FSB-Pro-generated portfolio EAs can be run in MT4's Strategy Tester and analyzed with MT4 Tracker to determine which are the most promising strategies.  This seems to work well.

2. Running that same portfolio EA in a trading account (demo or live) misses trades.  There is no question about it.

I've been using portfolio EAs in live accounts the past couple of months and the results are mediocre.  And these were using strategies with excellent statistics.  So, for now, I will recommend not to use this type of portfolio EA in a trading account.

I apologize this doesn't seem to be working out.  I thought it was a good idea, but since MT4 is basically a "black box" and no one except MetaQuotes knows what its limitations are, then there was no way of knowing in advance that a portfolio EA might run into performance problems.

Again -- this does NOT apply to EA Studio-generated portfolio EAs.  In my hands, those work great.

Hi Sleytus

Thanks for your reply.  I think i was using the correct MetaEditor.  I couldn’t decide which was the correct one, so I opened it up by clicking on the mqh file.   I then checked that the .exe file ended up in the correct version of MT4.  That said, what you say makes sense in that the errors are referring to a different instance of MT4.  I’ll have a look.

ATB

George

I think there are a few users who will be thinking about the difficulties and proposing tentative solutions.

One developer of a tick based scalping strategy recommends that for his product there be only one instance of the program and only one chart in order to get things correct. That corroborates the apparent difficulty that we are experiencing.

Hi Steve,

Thanks for the update. 

I understand that if I want to use the Portfolio EA for MT4 Backtesting, I have to switch to On Tick version, am I right?

Lastly, is it possible to have the "Description" section from the previous version whereby I can add a description onto my PM EA so that I can describe what's the portfolio consist of, as I have many PM EA running and this description helps put in more information such as 1) where these EA from from which batches of testing 2) When (date) I started testing this PM EA.

Thanks once again for your awesome work, PM makes such a big impact to anyone who wants to speed up their work.

Regards
Hannah

I don't believe that would solve the problem.  When MT4 calls OnTick it creates a new thread -- and that thread gets cancelled when the next tick arrives -- regardless whether the previous OnTick has completed.  Within OnTick, functions *are* called one-after another -- but there is also network activity going on (e.g. when you open or close a position) that temporarily blocks execution.  So it becomes a timing issue as to when / where a logjam would occur.

OnTimer is not called by MT4 and, hence, is not subject to cancellation by MT4.  It appears to solve both the logjam problem and thread cancellation problem.

As I mention in the FAQ, you may not necessarily suffer from the OnTick problem since it seems to depend on:
1. the number of charts that are opened
2. whether the charts are trading the same or different symbols
3. the machine's memory and CPU (and whether other tasks are running)
4. network bandwidth
5. ????

Hello Hannah,

Good point.  There are a couple of quick fixes:
1. Your license key is good for 3 activations, so the license server in the cloud won't complain unless that number of installations is exceeded.  So, you should be able to install MT4 Sidekick on 3 machines.
2. I can send another license key for Sidekick if you need more than 3 machines.

I will also check into building a version of Portfolio Maker that does not require Sidekick.  Even if Portfolio Maker doesn't use Sidekick there is code that is part of the build that still looks for the Sidekick DLL when the *.ex4 is loaded into memory.

I've found that Sidekick works very well so I just assumed everyone would want to use it whenever they used a portfolio EA -- but since you brought it up I can understand there may be times you want to test on multiple machines.

Give me a few days and I'll try to come up something.

Hello Hannah,

Trading the top 10 or 50 breaks from the "barometer" behavior that I think makes Sidekick a unique and powerful tool.  The top 10 or 50 strategies are determined from historical stats.  You can use MT4 Tracker to prune a portfolio for the best strategies.

I think it was you who pointed out in a previous post there really is no such thing as a bad strategy.  And that stuck with me and I think it is true.  Strategies perform differently depending on the current data stream.  How a strategy performed the past month does not guarantee how it will perform in the coming month.

The purpose of Sidekick is to provide a "barometer" of how a strategy is performing NOW.  It is mostly based on observations that strategies frequently exhibit consecutive streaks of wins and losses.  MT4 Strategy Tester even includes the metrics "Average Consecutive Wins" and "Average Consecutive Losses" -- so even MetaQuotes has noticed this.  As a side note -- it is interesting to me that no one ever takes advantage of that observation, as far as I can tell.

With Sidekick, I don't care how a strategy traded in the past -- i.e. maybe it performed great or maybe it performed poorly.  All I care about is how it is performing NOW.  And using a Demo Account (or a Real micro account) to create a shared database that the Live Account can use provides a real "edge" -- at least, that is what I'm finding.

Sorry Correction

2. How to use Exit only EA or EAs, made by FSB pro and to be used for ONLY one Symbol i.e. USDJPY with Portfolio Maker?
    Possibly I can combine NONO indicator and give a magic number name for EA in produced MQ4 by FSB pro file or what?
    Suppose that, this or these EAs to be applied to ALL EAs inside a specific portfolio made EA