Completely different results with the same expert on four BROKERS

New comment
Please help me understand: I am running the tester strategy with an expert of the MT5 market on gold with four different brokers Dukascopy, Swissquote, FP-Market and Ava-Trade and I get completely different results.
In particular with Swissquote and Dukascopy the result from 01.01.24 to today is the same for both at about 4 million euros with an initial deposit of 10,000 euros.
With Ava Trade and FP Market it is 100 times lower at about 35,000 euros.
The leverage is 1:400, the Drowdown is the same and also the number of operations carried out about 300 and the lost operations about 40.
I get the same result using real ticks of Dukescopy downloaded from TickStore on the four experts.
Please help me understand how it is possible for this to happen and which of the results should I consider real??
Thank you so much to anyone who helps me because I'm racking my brains to understand but I can't find reasonable answers on my own.
Maurizio Lucini:
Please help me understand: I am running the tester strategy with an expert of the MT5 market on gold with four different brokers Dukascopy, Swissquote, FP-Market and Ava-Trade and I get completely different results.
In particular with Swissquote and Dukascopy the result from 01.01.24 to today is the same for both at about 4 million euros with an initial deposit of 10,000 euros.
With Ava Trade and FP Market it is 100 times lower at about 35,000 euros.
The leverage is 1:400, the Drowdown is the same and also the number of operations carried out about 300 and the lost operations about 40.
I get the same result using real ticks of Dukescopy downloaded from TickStore on the four experts.
Please help me understand how it is possible for this to happen and which of the results should I consider real??
Thank you so much to anyone who helps me because I'm racking my brains to understand but I can't find reasonable answers on my own.

What is the seller saying ?

It's hard to be sure without checking myself, and without the code, but that seems like a bug. Can you check the lotsize of the trades on the different brokers ?

Hello,

The answer is very simple:  If the inputs different: the outputs (results) will different too.

Reasons:

1: Brokers  use different price feeds/liquidity providers

2: Brokers like to manipulate the prices, ticks/bars of course.

3: Brokers use different fees as spreads, commissions.

and so on.

These differences can cause different results.

On other hand: please keep in mind: backtests are not same as real forward test. Because the backtest never simulate the real trade situations.

In Backtest trades are executing without delays that is never happen in the real world. Delays is one of the most enemy of all scalpers, hft, arbitrage and many other systems that perform very well on backtests or demo, but never win on live accounts. 

Many developer try to sell useless robots, that can work well on demo or backtests but all of them fail  when you try it on live environment.

The biggest true is in forex: more easier to sell useless stuffs for unexperienced traders than to earn money with trade!

Quite easy to write a robot that perform well on backtests or on demo. But very hard too write a robot that perform well on live account for long period. And these type of robots are never sale!

So before you buy any robot: check the live performance with your eyes first. Ask investor access to the live account to see the valid performance.

If the buyer/product is correct he will allows it. If not: avoid it because it is a fake and you will throw your money to the trash.

Finally: be aware from these type of robots that produce big differences at different brokers. Sensitive robot = small chance to win. Robust robot = much more chance.

Best Regards,

AlgoGURU

Maurizio Lucini:
Please help me understand: I am running the tester strategy with an expert of the MT5 market on gold with four different brokers Dukascopy, Swissquote, FP-Market and Ava-Trade and I get completely different results.
In particular with Swissquote and Dukascopy the result from 01.01.24 to today is the same for both at about 4 million euros with an initial deposit of 10,000 euros.
With Ava Trade and FP Market it is 100 times lower at about 35,000 euros.
The leverage is 1:400, the Drowdown is the same and also the number of operations carried out about 300 and the lost operations about 40.
I get the same result using real ticks of Dukescopy downloaded from TickStore on the four experts.
Please help me understand how it is possible for this to happen and which of the results should I consider real??
Thank you so much to anyone who helps me because I'm racking my brains to understand but I can't find reasonable answers on my own.

The discrepancy in the results between different brokers you're experiencing can stem from several factors, even when using the same expert advisor (EA), real ticks, and leverage.
One of the main things that get missed and it's crucial for any test is to check the "Quality of Data" for each broker. That by itself can result in different outcome between different brokers.
Also you might want to check the market data feed, different brokers may have different data feeds, meaning the tick history or price movements they provide could vary slightly, even if you are using the same historical data for testing.
Execution and slippage is another reason you might need to consider especially in strategy that involves high-frequency trading.

Maurizio Lucini:
Please help me understand: I am running the tester strategy with an expert of the MT5 market on gold with four different brokers Dukascopy, Swissquote, FP-Market and Ava-Trade and I get completely different results.
In particular with Swissquote and Dukascopy the result from 01.01.24 to today is the same for both at about 4 million euros with an initial deposit of 10,000 euros.
With Ava Trade and FP Market it is 100 times lower at about 35,000 euros.
The leverage is 1:400, the Drowdown is the same and also the number of operations carried out about 300 and the lost operations about 40.
I get the same result using real ticks of Dukescopy downloaded from TickStore on the four experts.
Please help me understand how it is possible for this to happen and which of the results should I consider real??
Thank you so much to anyone who helps me because I'm racking my brains to understand but I can't find reasonable answers on my own.

I suspect a difference in symbol definitions. Gold is not equal Gold, it has a typical setup as a symbol, but dont expect it to be all the same at all brokers.

Sounds to me like one has a smaller unit in their contract (CFD).

Could be one of those examples, where you need to dig a bit deeper to find the difference.
Maurizio Lucini:
Please help me understand: I am running the tester strategy with an expert of the MT5 market on gold with four different brokers Dukascopy, Swissquote, FP-Market and Ava-Trade and I get completely different results.
In particular with Swissquote and Dukascopy the result from 01.01.24 to today is the same for both at about 4 million euros with an initial deposit of 10,000 euros.
With Ava Trade and FP Market it is 100 times lower at about 35,000 euros.
The leverage is 1:400, the Drowdown is the same and also the number of operations carried out about 300 and the lost operations about 40.
I get the same result using real ticks of Dukescopy downloaded from TickStore on the four experts.
Please help me understand how it is possible for this to happen and which of the results should I consider real??
Thank you so much to anyone who helps me because I'm racking my brains to understand but I can't find reasonable answers on my own.
Hi Maurizio,

i know this one is old, but its back on top now. Same number of trades, same losing trades and the same drawdown means the EA most likely took the same trades on all four. So the difference is the lot size. Going from 10,000 to 4 million in 9 months only works if the EA raises the lot with the balance, and then small things compound. A lower maximum lot, a different lot step or a different contract size (Dominik's point) on two of the brokers caps the growth at some point, and from there the results split by a factor of 100.

I'd compare the lot sizes of the last trades, not the first ones, and look at the symbol specification on each broker (contract size, volume max, volume step).

As for which result is real, I'd say neither. On a live account the broker's volume limit and the liquidity would stop that growth long before 4 million.

Best regards
Chris
New comment