How to manage the monthly EAs?

[Anonymous]

Hello!

I bought a good amount of courses but I haven’t seen any course that mentioned how to manage, switch or kill EAs after a month. Petko only shows how to pick EAs for the very first time.

For example, on June 1.

– I ran 100 EAs on my demo account and let them run for 2 weeks.

– After 2 weeks, on June 14, I compare the past 7 days to the past 2 weeks’ results. Pick the EAs that survived both in the past 7 and 14 days, and move them onto the LIVE demo. (Based on IIan’s recommendation/video).

– After 3 weeks, on June 21, I compare the past 7 days to the past 3 weeks’ results. Pick the EAs that survived both in the past 7 and 21 days, and move them onto the LIVE demo. Kill the ones that aren’t on the list from the past week.

– After 4 weeks, on June 30, I compare the past 7 days to the past 4 weeks’ results. Pick the EAs that survived both in the past 7 and 30 days, and move them onto the LIVE demo. Kill the ones that aren’t on the list from the past week.

 

Around week 4-5 or July 1, Petko updated the monthly EAs. My questions are:

– What should I do with the first batch of EAs? Let them run or turn off and replace them with a new batch?

– How do I know that the first batch and the second batch don’t include the same EAs?

– Petko recommended running these EAs at least 1 month before moving onto a live account, but he updated them every month so the new parameters change every month. So I’ll have to keep testing the new monthly EAs? When would be a good time to move them onto a live account?

Tanya

[Anonymous]

I’d do as Petko suggests. Each month you will be moving the robots that you had running on demo from the previous month.

Then you could put the new months robots that he has updated on a second demo account and when they have been demo trading for a month you can switch to moving the robots form that account.

Essentially have two demo accounts and alternate between the two each month for copying EAs from. Does that make sense?

You are only using 1 month results maximum to select your EAs so there is no need for running EAs for any longer on a demo account.

Hope that makes sense and sounds like a simple plan ;-)

 

[Anonymous]

Do you mean we will delete the old EAs after running the new EAs for a full month?

That means, the parameters of EAs that are uploaded for the current month, for instance, July, will be used in the real account next month which is August? Is that right?

[Anonymous]

Yes that is my recommendation.

So each set of robots would run for two months total. Month 1 is purely trading on demo to get her statistics  and then the second month is moving trades from that demo account to your live account (or mock demo).

If you have two demo accounts then you will always have a demo account that’s got a months worth of stats before you trade.

Hope that makes sense?

 

[Anonymous]

That makes sense!

It’s been 3 weeks since testing the EAs on demo accounts. So far, no luck yet. Still unprofitable from picking winning EAs in the past 7 days AND in the past 30 days (IIan’s method).

I will keep them running for another month and see if they will perform better. If not, I might have done something wrong haha

[Anonymous]

Excellent glad that all makes sense :-)

Profits won’t come straight away, it takes time and practice to improve and work out the nuances.

I have a few questions:

What EAs are your running?

How many are you allowing to trade on your mock demo account?

What is the balance between symbols an timeframes on your mock demo account?

Also what is the size of your mock demo account and what Lot size are you trading?

[Anonymous]

Thanks again Samuel for your response. Truly appreciate it :)

To answer your questions:

What EAs are you running?
– Forex Trading with 30 Strategies: Walk Forward Optimization (June) – EURJPY, GBPJPY, USDJPY
– Forex Trading with 30 Strategies: Walk Forward Optimization (July) – EURGBP, GBPJPY, EURJPY
– 100 Forex Strategies + Portfolio Expert Advisors Monthly (June) – EURUSD
– Cryptocurrency Algorithmic Trading + 50 Robots Monthly (June) – BTCUSD
– Cryptocurrency Algorithmic Trading + 50 Robots Monthly (July) – BTCUSD
– Top 5 Gold Trading Strategies in 2022 + 10 Robots Included (April) – XAUUSD
– Algorithmic trading with Dow Jones Expert Advisors + 10 EAs (June) – US30
– Forex Algorithmic Trading + 100 GBPUSD Trading Strategies (July)
– Automated Forex Trading + 99 Expert Advisors Every Month (July) – EURJPY, EURUSD, USDJPY

How many are you allowing to trade on your mock demo account?
– Only 3 sets of EAs that past 2 weeks test, the rest of them are still in the incubator period (first 2 weeks)
–  Total EAs that past PF 1.4 and more than 3 trades are 10 EAs
– Out of 10 EAs, I chose 7 EAs to balance out the portfolio. I make sure they have equal weight. And I picked them based on net profit (Imitate IIan’s method here)

What is the balance between symbols an timeframes on your mock demo account?
– All EAs I got from Udemy course are M15
–  The symbols are GBPJPY, EURUSD, BTCUSD (All M15)

Also what is the size of your mock demo account and what Lot size are you trading?
– It’s a 10,000 USD starting balance. The minimum lot size is 0.5.

Any recommendations?

[Anonymous]

You are welcome Tanya :-)

The above is all pretty sensible which is great, although ballpark I would probably be aiming for about 12 EA’s with a 0.1 lot size on a 10k demo account. If only trading 6 then could increase to 0.2 and likewise if trading 24 EAs then trade 0.05 lots per EA.

You will get a feel for the kind of profits and drawdowns with a bit more experience and can tune how you want but this is just a recommendation to start out.

Also M15 and H1 timeframes tend to combine well so you could also try including some H1 in your portfolio for the same assets.

Other than that I would say be very clear in your tracking of what you are doing and the results you are getting. That way over the course of a few months you will build valuable data for moving forward.

 

[Anonymous]

Let me recap…

– The ballpark for 10K account is 12 EAs @0.1 lot size

– M15 and H1 tend to combine well for the same assets

For picking the winning EAs, how many days of performance do you look back before moving them into the real account? What are the criteria do you like to see? (I know IIan and Petko have their methods, I just wanna hear yours :) )

[Anonymous]

Hey Tanya,

Yeah good recap.

For selecting the EAs we are doing a couple of things when we trade them on demo. First is we are actually applying a final filter that removes strategies that MAY have been overfit and also then trying to catch the EAs in their profitable phase.

I think it’s important to wait the full month personally before moving anything onto mock demo with over 10 trades and good performance, also I like a week as a shorter filter too.

For the short filter I am happy with a few trades and a profit factor of 1.1. For the longer filter I like to see more trades and a profit factor of 1.1 along with a drawdown value that aligns with what was shown in the backtest.

A longer filter would also be smart in my opinion (say 3 months) but it’s a balance between managing all the accounts also so I like 1 month and 1 week for simplicity.

I would suggest that you think of a few things you would like to try and then have three mock demo accounts and compare the results over a few months to see what works best (remember and record everything in a spreadsheet).

Of course you have to find the balance with time and effort that works for you. I don’t have a lot of free time so I adjust my systems to suit.

[Anonymous]

Hey Samuel,

– So, you use both short filter (1 week) and long filter (1-3 months) to pick the winning EAs. My question is, isn’t 1 week filter too short, let’s say for H1? Maybe 1 week is good for M5 and M15, but not for H1 and H4 due to the number of trades?

– Speaking of time and effort, I myself have a lot of free time every day, however, I don’t want to trade M1 or M5 if the results aren’t as consistent as higher TF (maybe because of the news or other factors). I noticed on my demo account that when I picked M1 and M5 using 1 week and 1 month filters and put them on a live mock account, they didn’t perform as well as the previous week. I’m not sure if it’s because the market condition changes too fast or something else.

– Another question I have is that you mentioned it’s ideal to run a full month on a demo before moving the winners to a live account. After 1 month, do you run Reactor to generate complete new EAs (let’s say 10 portfolios with 30 EAs each)? OR do you optimize the parameters of the previous month’s EAs?

– How often do you generate new EAs using Reactor?

– Just curious, what is FSB for? When to use them? Do I REALLY need FSB when I already have Expert Advistor Studio? How do we put both of them into use and benefit each other?

[Anonymous]

Hey Tanya,

Good question re timeframe. It would depend on number of trades really, if your H1 is making 3 trades in a week on average then it’s fine but like you say if it’s only making 1 or two then two weeks would more appropriate for H1.

I tend to stay away from M5 and M1 personally, the lower timeframe the more of an impact noise is gonna have which is totally unpredictable so I stick to M15 and H1 mostly (H4 and Daily also).

For how often to add new EAs it would depend on how many on your demo account are still trading well generally (ie have a profit factor of 1.1 over the complete time on demo would be good ones to keep). In do tend to just put a complete new set of EAs up once a month but it’s not the only way.

Do you have FSB also? It is great but you don’t need it. In my opinion it’s best when you have a particular trading strategy that you are trying to create. Both EA studio and FSB pro have their strengths and weaknesses but I wouldn’t be overly concerned about rushing to learn FSB pro unless you want to do something that you can’t in EA studio.

[Author]

Posts