How do you validate that an EA is truly profitable? - page 3

New comment
 
Ryan L Johnson #:

0.1 standard lots is a mini lot, 0.1 mini lots is a micro lot, 0.1 mini lots is a micro lot, and 0.1 micro lots is a nano (cent) lot─assuming that you're trading OTC FX or CFD instruments.

So... Assuming that the Tester is bottomed out at a micro lot, adjust the tick value to 0.01 to compensate for a cent lot.

Again, this assumes that your tick size is correct. If not, change it to match your live chart's price decimal digits as well, e.g., 0.0001

Thank you for this. Very helpful. 
 
Yengkok #:
Thank you for this. Very helpful. 

You're most welcome.

To be perfectly honest, I don't even have access to nano/cent lots in my location... but I've been around long enough to see enough about it.

 
Ryan L Johnson #:

You're most welcome.

To be perfectly honest, I don't even have access to nano/cent lots in my location... but I've been around long enough to see enough about it.

But your suggestion worked. Thank you bro for the help :)
 
Osmar Sandoval Espinosa #:
How long you like waiting until consider moving to a bigger account?
5-6 month's atleast depends on the EA, if its martingale/grid I'm just flipping accounts never use it on more than 2k account.
 
I would validate an EA in stages: backtest for logic sanity, out-of-sample for robustness, demo forward test for execution, and then a very small live check for the final reality test. The main thing to watch is whether the drawdown shape and trade distribution stay similar as you move from synthetic to live conditions. If those change sharply, the model may depend too much on the test environment. What matters most to you at this stage: win rate, drawdown, or expectancy?
 
Hesham Ahmed Kamal Barakat #:
drawdown or expectancy

drawdown and expectancy are the mosti importanat performance indicators.

expectancy (R) comes from win rate and Risk reward ratio

 
Well first I would use some meta filters given by the MQL5 itself to identify the key metrics of the EA LIVE signal.

They have a very interesting tool here that discriminates a lot for who takes the time and knows how to read it.

https://www.mql5.com/en/signals

And then I use these filters that are realistic of a sound risk management.

◾️ Using Robots
◾️ Max leverage: 1/100 (higher is not safe...)
◾️ Monthly Profit: no limit
◾️ Max DD: Max 15%

If it is not there this is a huge red flag.

Chris
Trading signals for MetaTrader 5
Trading signals for MetaTrader 5
  • www.mql5.com
A showcase of successful trading strategies with automated copy trading
 
Backtesting through different market regimes, forward testing then out of sample testing, looking at stat metrics then if you are confident, go live with a small account or demo
 

If an EA has a stable profit of more than 10% per month, it's a "suspicious" flag. If it has more than 25% profit per month, it's a red flag. If it has more than 50% profit per month, it's a pure scam.

A 50% monthly profit turns $10k into $1.3M within a year. If someone has such an EA, it's stupid to put it up for sale, as it can be reverse-engineered and lose its ability to make money.

123
New comment