Forum Replies Created
-
Posts
-
Hi Petko,
I traded the strategy on a dukascopy demo account for a month along with many other strategies (Dukascopy demos expire after 2 weeks but I requested one that doesn’t expire, when I opened a live account with them :-).
It only made a single trade for a small profit but that was to be expected from the numbers shown in my previous post (Average trade of 483/108 = 4.47). Reading this into EA studio also confirms that this profit was made with very little drawdown as shown below (No doubt there was a bit of intrabar drawdown but EA studio shows zero drawdown which is good).
Yeah gave it a large stop loss level as you say. I just wanted to give the indicators room to do their job and from looking at the journal really there were very little trades closing via stop loss or take profit but rather just from the indicator signals which was the intention (and also what happened with the trade on demo).
I’m no expert on all the indicators but I do like to have a very quick scan to make sure there are no super obvious red flags in the strategy, like buying long when it looks to me like the entry rules would better suit a short position for example (for ranging strategies). EA studio of course generates sensible strategies but sometimes I like to just target ranging strategies so like to remove anything that obviously looks like a trending strategy and vice versa, also if a strategy is too complicated or doesn’t make sense to me at a reasonably quick glance I’d rather avoid it even if it probably is fine.
As I said this one was a super simple un-exciting strategy created manually. But this week I will add one that is showing a much more impressive equity curve and has been generated and tested though the reactor that can hopefully result in a few more trades and a larger profit.
Of course there are never any guarantees especially with a signal strategy as it could be a great strategy but just going though a losing phase for a few weeks, that’s why trading lots of strategies is so important.
Exactly Petko :-),
If its clear that I am getting more and better EA’s for certain pairs and timeframes and it makes sense to play to that!
Glad to hear you feel similar about not being too pedantic about keeping things perfectly balanced in this way.
Likewise if a certain asset always seems to be performing poorly generally in backtests compared to others then its seems okay to leave it out of the portfolio rather than include it just the sake of maintaining the pair balance, provided there is not too much exposure to a single asset and also enough diversity in assets.
At the end of the day the goal is to include a wide range of uncorrelated strategies over several assets in order to increase my chances of a smoother equity curve with more controlled account drawdowns.
That’s been my learning too Petko.
Aiming to keep things simple is always best and if what ever you are trying starts to feel too complex then take a big step back and do some thinking!
Also fallen victim to trying to manage far too many demo accounts. Finding the golden middle ground as you say with regards to simple systems and the amount of tests you are running at once is really important otherwise it can all get too messy and its easy to end up learning nothing because the lessons are all to jumbled. Less haste is more speed as they say :-)
Hi Mario,
What reason have you decided you need so much data? Is there a specific type of strategy that you are targeting?
From above my understanding is that you are playing around with just the stop loss and profit taking ranges in order to obtain strategies that trade only a few times a month?
Presumably your logic is that in doing so you are targeting better entry and exit points that happen more rarely? Are you trying to target trend following strategies?
As Petko says with what you are doing you can easily experience long stagnations or losing periods. If you are determined to stick with such a long time frame and low frequency trading strategy then I would definitely suggest doing so on many assets (30+). I also think its sounds like an unnecessary amount of data though.
I agree with Petko that 1-100 is a pretty wide range for the Reactor/Generator to do its work, would be interesting to understand more about what exactly you are targeting though?
It definitely sounds like you have something clearly in your mind of what you want which is a good thing!
Hi Dirk and Burak,
Good to see you have joined the forum, looking forward to seeing how your results go!
Happy to help how I can Petko :-), and couldn’t agree more about continually trying to improve with testing and experimenting!
Hi Ekene,
This is copied form an earlier post in this topic:
”Hello, you are absolutely right, Jacpin. The best thing would be to test a Demo account with higher leverage, and you can do it 100% same as your live account. This way you will see what will be the different. basically as Petko said, there will be no difference in the performance, but you will see more free margin to trade with.”
On a 25k account with 100 leverage you have 2.5 million dollars of total buying power but you don’t have to use all that buying power (and definitely shouldn’t), but there is nothing wrong with having it and not using it.
Opening for example 20 0.1 lot positions simultaneously requires 200k of buying power. So on a 25k account you could open all these orders with 1:8 leverage and would have zero free margin remaining (ie you couldn’t open any more trades) but with 1:100 leverage you will still have over 90% of your free margin remaining to open more trades with if you wanted to.
Bottom line 1:100 leverage isn’t risky but trading without good money management is.
I’ll add one, see if I can get the ball rolling ;-)
Its a SUPER simple one, I created it manually (i.e. no generator) and each indicator setting has default settings. Optimization didn’t result in huge improvement so I left it unoptimized.
I ran a monte carlo with 40 tests on changing indicator parameters only and they all passed (Repeated this 3x), I also checked that it performed well on several uploaded data from different brokers over the period above which it did.
I created it using the end date of about 6 months ago (Doing the steps described above) and then tested it on the most recent 6 months that I had kept up my sleeve and again it performed well on all the brokers data (5 other brokers data including Premium and Metatrader data).
Its definitely not making loads of trades but I think since its been checked on several different brokers data that can consider the trade count in term of how many trades it performed well on, considerably higher then the 108 over the 5 years as shown?
For the premium data shown spread is 10, commission is 7 and swap is -5 and -2.
I’ll aim to add something more exciting soon, just thought I’d try and kick things off. Looking forward to seeing some strategies others have created :-)
Hey Jaroslav,
Great to hear form you :-)
It definitely sounds like things are improving a lot which is great!
Your first question is around the timings. I personally think the more the better if you have time to update the EAs so if you can make it a quick and easy daily task to do then why not. Its important to tweak your system around your current schedule and make it manageable. BUT how often is necessary is dependent on the timeframe and frequency of how often the EAs trade. For example if your EA’s are H1 and M30 and typically opening trades 1-2x a week then changing a couple of times a week is likely to work well, but if you have M15 and M5 that are trading say 3 or more times a week typically then I would probably favor updating every day or two.
What broker are you using? Demo and live should be very similar so to check this for yourself why don’t you try the following:
1 – Upload the live data and demo data to your EA studio account and then see how the results compare when you switch from using the live to demo data. Also you can check with Mt4 back tester.
2 – If you want to convince yourself of how live and demo compare then why not pick a few single EA’s each for a different pair and trade them each with the smallest lot size on both the demo and live account to compare the results. Of course this could cost you a small amount if they lose but it really shouldn’t be a costly lesson (You could even cheapen the lesson by reducing the stop loss and profit taking from typical max=100 to 50 when generating EA)
Definitely a smart move to run more than one demo at once, just make sure to track everything properly so you learn from all the things you are trying but will hugely speed up the learning.
Keep doing what you are doing. I really think you are going in the right direction, just keep doing what you are doing and learning from the results.
I can understand wanting to get onto live as soon as possible but once you have sufficiently consistent demo results then this will translate to good live results. Whenever you go live, to start you could also put the same EA’s on a demo initially to further build your confidence in this “live ~= demo” equation.
Hi Guys,
First of all I like what you are doing in collaborating and learning together as a group of 3!
In a nutshell what is seems you are doing is the following:
1. Using Petkos approach presented in the Professional Course but with some modifications
2. Rather than using Petkos pool you are looking to generate your own pool and apply the approach to that
This sounds like a sensible enough starting point to me.
I am guessing that you have formed a collective in order to both share costs such as software, hardware and trading capital etc as well as knowledge and effort?
I’ll try to answer a few of your questions below but first off I would like to ask what type of timeline and progress path do you intend to take towards trading with live funds and what checkpoints along the way seem sensible?
These are my thoughts on some of your questions, will be interesting to hear from Petko also for sure (The website is going through some work and he is no doubt swamped so I expect he will be a little later to respond than usual due to that btw)
3.How can you avoid that a strategy in LIVE goes directly into a loss phase.
You cant 100% avoid this this only becomes a big problem depending on the profit window you are aiming at. For example if you have a strategy that has ups and downs tending to last about a week but is generally going up over the course of a month say, then if you start it during a drawdown but keep it on for a month then likely all is well. Of course if you are changing robots weekly then you may just get the drawdown for a week and then remove the EA and never get the profits that were coming. It sounds like you are aiming to change robots monthly which is sensible so take a look at how long the drawdowns tend to last I suppose and wether it starts in a drawdown or profit is partly down to luck but be more concerned with not how it starts but how it finishes if you give it appropriate time.
So simply put if a strategy looks good but it has had long stagnations lately or drawdowns lasting more than a month then I would try to pick a strategy that has shorter stagnations and drawdowns lately, that way at least if you do hit a drawdown initially it is less likely to last long. Also at this point of refinement I’d personally always favor a small sample of most recent data only (i.e. the most recent couple of months)
4.When the trend changes, our strategies are also adjusted. Or do we stick to a 1 month rhythm?
Create a plan and stick with it. A month is a sensible rhythm if you interfere and try to adapt too much then this is gonna get harder and harder to backtest.
5How often do you add new strategies to the pool? And how do you avoid duplicate strategies in the pool itself.
You can add strategies to the pool as much as you want. Duplicates are not a problem in the pool as when you pass them through the validator any duplicates will be removed. If the pool seems to be getting too large then you can easily put them through the validator with to strip some out. For example lets say you grow your pool to 3000 over 6 months or so but you would rather trim that back to 500 you could just load half of your collections to the validator and rerun with your preferred settings and then save the new collection and repeat for the other half. That way you can get the top 300 according to what you prefer (profit factor, nett profit, R/D for example).
If there are a lot of strategies being discarded purely because excessive strategies are passing the criteria but the collection only holds 300 then you can rerun the validator with an either an increased criteria (say increase pf>1.2 to pf>1.4 for example), or a stricter robustness test such as increase monte carlo for 80% to 90% or reduce the correlation coefficient from 0.98 to 0.95 for example to give less correlated strategies.
There is a lot to learn and it take time, so get lots of practice using demo accounts until you are ready for live. You can also of course try several different systems on several separate demo accounts to save time in learning. Make sure to keep track in a spreadsheet you can can learn from what you have already done
Keep us posted on progress, I’m sure others will chime in soon :-)
Sounds good,
I still think the calculated risk is a little low, you are essentially expecting a Return/Drawdown of 8 (Nothing wrong with being optimistic though!). But your time in demo trading will give you more of a feel for what to expect (which is one of the reasons its so important).
I would definitely recommend you keep close track and record these important parameters in your demo trading to build a ‘feel’ for things (which really only comes with time and practice). Don’t be too surprised or disappointed if your drawdowns are more than expected and you can adjust for the next month with new insight (continually :-).
Also you say you are always searching for lowest drawdown, just make sure this doesn’t mean lowest profits also! The return/drawdown may be more appropriate I think.
Happy trading, looking forward to seeing upcoming results and discussion!
Hi guys,
Well done on your selection of software and community btw. It took me a while to find EA studio and this community and I strongly feel it is the best software out there and the best place to learn it.
So in a nutshell you are using a modified method of Petkos Professional Course using a a self generated pool of EAs?
I’ll give my thoughts on some of your questions and definitely look forward to hearing from Petko also. There is actually so much information in your above post it almost opens up too many questions. I’ll take a look tomorrow and add a few extra points but for now I’ll just stick to answering a few below.
3 – How can you avoid that a strategy in LIVE goes directly into a loss phase.
There is always going to be a risk of this happening. However its significance depends on your trading rhythm (as you put it). For example, say you are trading an EA for one month that typically has shown in backtesting to have waves lasting about a week but generally trending up. If you are trading this EA for a month then starting on a loss is not so much of a problem, however if for example you EA has shown to have waves lasting a month or more then that could be a problem.
You are going to be trading 36 EAs so make sure as a final check to have a look at the typical cycles in the last 6 months and favor ones that have shorter stagnations and drawdown periods to lower the risk. You can take other measures such as taking a look at what factors might typically lead to the reversal of a drawdown but I’d say starting on drawdown or not can just come down to luck sometimes but it is not how the month starts that is important or skillful but how it finishes.
4 -When the trend changes, our strategies are also adjusted. Or do we stick to a 1 month rhythm?
Don’t interfere with your EA’s based on emotion. You should decide on a rhythm and a plan and stick to that. Backtesting a system based on a 1 month cycle of changing EA’s is a good timeframe and once you have decided on a good system you need to stick to the plan and keep emotion out of it. Always on demo of course until you are consistently profitable.
5 – How often do you add new strategies to the pool? And how do you avoid duplicate strategies in the pool itself.
Don’t worry about duplicate strategies in the pool as when you run the validator they will be filtered out. This can be reduced by decreasing the correlation coefficient though.
New approaches:
1 We have chosen to reach the 20% target. we have chosen 3 strategies instead of 2. The risk is still within limits. How many strategies per time unit and asset do you use? (See table below)
This sounds fine and is only a small tweak to the system in my opinion. Just lower the lot size traded as you increase the number of strategies. Is your calculated risk simply a sum of the drawdowns given by EA Studio? Over what period was this risk and profit shown in the table calculated (OOS, IS, COMPLETE?). For any of these taking these values seems very optimistic to me and looks like you have simply taken an average for the drawdown risk? For example if EA studio has calculated the OOS max drawdowns to be 2% for 10 EAs on different pairs then I would be expecting a risk of at least half of those experiencing a drawdown at the same time which would be 10% but it looks like you are just taking an average and assuming a risk of 2% for the portfolio of 10 EAs?
2 We are currently testing with 1 strategy, with 7 weeks instead of 4 weeks, on the grounds that we think the performance is more stable if we choose a wider time period for filtering. We tested this and found that the sweet spot is somewhere between 6 – 8 weeks. What are your experiences with rebalancing the portfolio, what intervals do you have? (See table below)
This seems a little bit narrow in terms of testing and I think you need to increase your sample size significantly before deciding that this is the sweet spot, its simply the sweet spot for this particular strategy over the trading period you have chosen. its highly likely if you repeat the exercise for a few different strategies a different sweet spot will appear.
One last thing is a big apology if I seem over critical in my thoughts and super to the point. I actually think you’s are getting off to a GREAT start and am very interested in seeing how you guys progress :-). Its just late here and I am tired but I wanted to reply. (Also I know Petko is very busy as the site is going through some big updates at the moment and so I expect he wont be as quick to respond as normal right now).
Keep us posted with your progress :-)
Hi Selim,
About your 20% monthly target. I think it is a good target. For reference, on my first month of demo after taking the professional course I increases a 3000 account to 4500, so a 50% increase. I used Petkos pool and just followed his instructions as closely as possible. Easy to then rush into live trading due to impatience but consistency is everything!
A good rule of thumb is make sure you are happy with your profits and consistency over a 3 month period before moving to live.
The great thing about EAs (Or one of the many great things) is that emotion should be removed. So more than manual trading it should be manageable to remain consistent if you follow your system with no interference. Even so if that first month of live the first week starts with a drawdown emotions can be tough to handle when its real money so build your confidence and understanding in demo. If you have three months of profit in demo then you would have been profitable in live also with the same broker if you did the same thing (The results wont be identical but they will be very similar).
Also, why don’t each of you three get on here as a separate voice? You could each post your individual results and discuss with each other and Petko and the community can join in? I think you would each learn more that way and get more benefit from the forum.
You could also each try a slightly different approach inline with your overall philosophy too and see what works best over a couple of months (For example one of you could run the system using Petkos pool, one with your own pool and one with say M30 and M5 from either Petkos pool or your own) OR even just all generating your own different pool with identical settings and identical approach OR each using the same pool (either Petkos or you own) but using different brokers. Can be a definite BIG benefit to a team of 3, use it to its full potential :-)
I can see three separate fxblue account links but not much is happening on them? Its not very clear to me what is being shown here? Is the intention to each run the system separately and share results in this way?
I’d personally love to see this topic progress on the forum and hope you’s stick with it. I like the organization and planning even if I don’t fully understand 100% of it, its definitely a big key to success in this so off to a good start.
That sounds like a pretty good plan forward Matthew. I’d say increase the number of bars to around 35000 for the H1 timeframe though which is about 5 years. Maybe try the same for the M30 or M15 timeframe also for 4-5 years.
15 months is too short for H1 period.
Also do this using several pairs as discussed. This is also going to be important in limiting your account drawdown.
Keeping track of everything in a spreadsheet is a really good idea also, not just a quick sheet but something that you will maintain for months.
“I’d like to find a way to make consistent profits every month with this and I’m gonna get there!” With this attitude you WILL get there!! Keep pushing, every time you try something and if it doesn’t work you will learn something and develop deeper insight and this will lead to success over time. Fall 7 times get up 8 as they say – I’ve fallen a hell of a lot more times than 7 but I never stop getting up;-).
Enjoy the challenge :-) Looking forward to hearing how you get on.
Hi Petko,
I am wondering how long you intend to pause trading?
I have paused also but my plan is to run some backtests over the past 2-3 weeks soon and see how things would have performed and then provided results are okay then start again, does that seem a sensible approach?
Also 100% agree with your advice above, definitely a rough patch for trading but regardless Jarolsav it sounds like you have been stung by being impatient (as I have been several times so not a criticism at all :-).
BUT you clearly said that you traded demo and lost 20% and then moved straight to live from that point? That is the main issue I see really, take more time to practice and once you have made some profits in your demo account then that is a sensible point to try moving to live (start with small money though)
Can’t think of many things worthwhile that take only 2 weeks to master, keep practicing and I’m sure your results will improve though ;-)
Yeah my original thought re generating for complete but then normalizing for 30% was as you say, optimization has changed the strategy and then you check it’s still okay OOS.
But if you want more bars for generation then I think just set the bars higher, but keeping the OOS completely reserved for a final check so that it hasn’t been used at any point in the strategy development is the way to get the most out of this robustness check.
