Topic: What size account to test?

I'd love to hear your thoughts on what size account you test new systems on?  I understand most people will test with micro lot size - but what account balance do you have?

Let me know your thoughts smile

Re: What size account to test?

DoCZero wrote:

I'd love to hear your thoughts on what size account you test new systems on?  I understand most people will test with micro lot size - but what account balance do you have?

Let me know your thoughts smile

Cent Account i find not good, spread to big.

For Real in Test i use 500 Dollar Account with 0.01

smile

Re: What size account to test?

Hey Rantampla,

Thanks for the reply - thats the same amount i'm testing with at the moment (with a 50:1 leverage) , but i'm concerned about running out of margin.  I'm currently testing 10 systems side by side , but have more waiting (haha).

How many systems to you test concurrently?

smile

Re: What size account to test?

DoCZero wrote:

Hey Rantampla,

Thanks for the reply - thats the same amount i'm testing with at the moment (with a 50:1 leverage) , but i'm concerned about running out of margin.  I'm currently testing 10 systems side by side , but have more waiting (haha).

How many systems to you test concurrently?

smile

Ich check all on Leverage 1:400 - that is more better for the small Account i think.

First i have use 100 EA, was to much big_smile

Now i test ~10-20 on 500 Dollar Account smile

Re: What size account to test?

I'm here to debunk some of the misconception some people has on demo testing.

The whole ideal of testing is to create a similar trading environment as close as possible to the real trading.

If the ideal of using nano account gives you a sense that it's "real" just because you are using real money, then you may be very wrong. 

Firstly, if all you ever going to trade is in nano/cents account (for the rest of your life), yes by all means use nano account.

However if you are ever thinking of retirement whereby your Forex income is going to supplement your stable income, then trading in nano account isn't going to be the answer.  So in this case, testing in nano account is redundant because it will not provide you that kind of "real" trading environment you need whereby you will ultimately trade in, std account.

Most people think demo account aren't that "real" and yes, it all depends on your broker.  But yet demo account (from reliable broker) is the next real thing you ever going to get.

I've been using demo accounts for testing all these years and I can "replicated" my success from demo accounts to live account.  And If I've any doubt about replicating this success, another possibility is to copy trade from demo account to live accounts.  From my experience, I don't have any problem and as long as your broker is reliable and has many LP, you won't have any missed trades.

I've tested 2 brokers, JFD and Tickmill.  Their demo and live accounts feed are identical and I do demo copy and live trade to test out and I'm satisfied with the results.  Do your own test yourself, if a broker you can trust, then open a live account and run your EA in both demo and live to see the results and whether it meets your standard of requirement.

When you start to seriously develop EA, you will start having hundreds and thousands of them.  Testing them with real money isn't going to be practical.

Re: What size account to test?

hannahis wrote:

I'm here to debunk some of the misconception some people has on demo testing.

The whole ideal of testing is to create a similar trading environment as close as possible to the real trading.

If the ideal of using nano account gives you a sense that it's "real" just because you are using real money, then you may be very wrong. 

Firstly, if all you ever going to trade is in nano/cents account (for the rest of your life), yes by all means use nano account.

However if you are ever thinking of retirement whereby your Forex income is going to supplement your stable income, then trading in nano account isn't going to be the answer.  So in this case, testing in nano account is redundant because it will not provide you that kind of "real" trading environment you need whereby you will ultimately trade in, std account.

Most people think demo account aren't that "real" and yes, it all depends on your broker.  But yet demo account (from reliable broker) is the next real thing you ever going to get.

I've been using demo accounts for testing all these years and I can "replicated" my success from demo accounts to live account.  And If I've any doubt about replicating this success, another possibility is to copy trade from demo account to live accounts.  From my experience, I don't have any problem and as long as your broker is reliable and has many LP, you won't have any missed trades.

I've tested 2 brokers, JFD and Tickmill.  Their demo and live accounts feed are identical and I do demo copy and live trade to test out and I'm satisfied with the results.  Do your own test yourself, if a broker you can trust, then open a live account and run your EA in both demo and live to see the results and whether it meets your standard of requirement.

When you start to seriously develop EA, you will start having hundreds and thousands of them.  Testing them with real money isn't going to be practical.

Good Info smile

I think Cent / Nano Account is not Real, the Spread is to big.

But my thinking for Demo is: The Spread is the same all time or not?

In live the Spread changing fast, but in Demo all time the same Spread or i am wrong?

I have last 1,5 months 2 Portfolio with Demo with 100 on every Account run.

Now i have for 1 Week the best changed to live with 0.01 real, not Nano / Cent.

First week are good.

Re: What size account to test?

I'm not sure about your demo,

But my demo accounts will have identical data feed as the real, ie. spread, price quotes etc

I even run my EA simultaneously in both demo and live and there is very slight difference such as a few cents difference overall for the total trades (for all EA, not just 1 position). 

So that's good enough for me to "tolerate" the difference between demo and live. 

Do do your own observation cos I can't speak for all brokers (especially those unreliable ones cos they make a living out of your failures, so they will do all they can to make sure you fail).

Re: What size account to test?

hannahis wrote:

I'm not sure about your demo,

But my demo accounts will have identical data feed as the real, ie. spread, price quotes etc

I even run my EA simultaneously in both demo and live and there is very slight difference such as a few cents difference overall for the total trades (for all EA, not just 1 position). 

So that's good enough for me to "tolerate" the difference between demo and live. 

Do do your own observation cos I can't speak for all brokers (especially those unreliable ones cos they make a living out of your failures, so they will do all they can to make sure you fail).

mhhhh ok, i have this never real checked, i have this for many years back seen, and i have thinked it is every time the same...

but when you say, i must check now, much years later now, maybe the broker have changed it....and it was in cfd trading, no forex, where i have see it wink

Thank you smile

Re: What size account to test?

Hi Rantampla & Hannahis,

either demo or nano account - i guess effect is the same (or fractional difference in execution).. But the real wisdom , in my opinion, for this thread is:

When you start to seriously develop EA, you will start having hundreds and thousands of them.  Testing them with real money isn't going to be practical.

This is exactly the issue - I will run out of margin to test the amount of systems I am developing. I think Demo is the only way..

Thank you for both your input smile

Every day i'm learning wink

Re: What size account to test?

I am using Virtual hosting to test - what are your thoughts then in using another broker (with free virtual hosting) to test, or should I bite the bullet and rent a full vps for demo testing?

Re: What size account to test?

DoCZero wrote:

I am using Virtual hosting to test - what are your thoughts then in using another broker (with free virtual hosting) to test, or should I bite the bullet and rent a full vps for demo testing?

When you only the EA Studio Portfolios check, then enough a VPS

When you will use many MT4 Instances, then better you use a Dedicated Server.

I have a VPS for fast scalping to Broker with 0.3 ms

and a dedicated Server, for normal Strategies with 20ms to Broker.

I have good Experience with that smile

12 (edited by hannahis 2017-08-30 19:10:15)

Re: What size account to test?

Hi DoCZero

Save your money on VPS especially when you are demo testing.

How often do you encounter power failure?

I run all my terminals on my computer and only use VPS to run my "additional" trialing EA.

So that when there is power shut down, what are you most afraid of? Those open trades right? (not un-open trades)

As long as you always have your SL set right from the beginning when you placed your trade, then when power shut down, you aren't afraid that your trades will close in SL, but you will then be "afraid" you lose the opportunity to make money and close it at the good price (while your are experiencing power shut down).

Thus to run an "additional" trialing in my VPS that happened to be the same trailing I'm running for all my open trades in my computer, I am ensured that all my opened trades are well taken care of.  Using this method, I save lots of money in VPS becos I only need to subscribe to the most basic package (in my case, I got my VPS free from my broker).

In summary, subscribe to VPS only when

1.  You are running live account
2.  When you are always facing power failure that really impact your testing results.

Otherwise, save your money cos if you are using micro lot size to trade in the beginning, you will realised that VPS will eat up most of your monthly profit.  I have been testing for years, if I were to add up all these money saving, it would be thousands of dollars by now cos I run 20 to 30 MT4 terminals and that would required a large VPS package/system.

Re: What size account to test?

A VPS in Germany give it for 12 Euro / Month. It is good for 2-4 MT4 Terminals.

12 Euro i think all EA must make this smile

Re: What size account to test?

Hello All,

The VPS isnt so much about reliability - but for me, more about system resources.  I'm doing many different things with this PC and building systems is only one part of it.   Testing also would probably not be the best way.  So the options I have is to build a dedicated PC for trading testing / live systems.   

A better option for me is to rent a VPS - I can get a 5 core 3gb ram VPS for around 30 usd / month - its cheaper than building the rig (on a yearly basis).

I went this way in the end - ping to broker is 3ms which is good enough to me smile

Thanks for the inputs!

Re: What size account to test?

DoCZero wrote:

I can get a 5 core 3gb ram VPS for around 30 usd / month - its cheaper than building the rig (on a yearly basis).

Can you please share the vendor/provider of your VPS. Thanks

Re: What size account to test?

Hello Araza,

Im using VPSserver.com - but there are many providers. I setup the server in New York as its close to my broker.

If you would like some help in setup ect, just PM me and i'm happy to discuss further smile

Re: What size account to test?

I set it up with forexvps and they are linked to my pepperstone account. The specs are not as good as yours for the price so was asking.
Thanks will definitely pm you in case help needed or when decided to switch.