Eusebiu Dascalu / Profil

Someone mentioned in the reviews that the bot only works in backtesting; for live trading, it's a waste of money.

test it!

That's a good result. I’m looking forward to good results ahead.. In the case of a prop firm, the spread is high, so I’m curious about how much difference the results will make..

it take a good insight to build a logic EA great work sir

thank you, really appreciate it

A few users specifically asked me for a PROP Firm version, but you can use the version that suits you best

OK, let me try another way: what makes those things preferable for prop firms?

The idea was to target higher profits while keeping risk as low as possible. I chose these settings with prop firm risk limits in mind, but they can also be used on regular accounts

TP slippage - the resulting market order gets filled at a different price, even without gap/news. With a good broker, this price is usually better than requested, just like with SL it's usually worse than requested. We can optimize for this by preferring to close e.g. breakouts on dynamic TP than on trailing SL. For strategies with fixed SL/TP, it's a reason win rate matters. Luckily, to some extent this is reproducible even in real ticks backtests. In real execution, the slippage/overshoot may be larger such as due to larger lots and limited liquidity.

Oh, didn't realize you also posted this as a forum thread - added my reply to there.

A large number of EAs here in the Market kind of pass backtests when run in "Every tick" mode (fake ticks) or presumably also on demo account ticks, but fail on "Every tick based on real ticks" when logged into a real account. So yes, presumably many EA authors who don't actually trade make this mistake... and I'm afraid some others do it on purpose.

You're absolutely right!

yes

StarTrader

thanks!

jForex

It is still 99 on our end

Interesting results, but a few numbers stand out to me and I'd be curious how you'd explain them: 1. Sharpe of 20.21 and LR Correlation of 1.00. An LR correlation of 1.00 means the equity curve is mathematically a straight line. Combined with a Sharpe of 20, that's not a return profile that exists in live trading — a very strong real system sits around Sharpe 2. 2. The win/loss asymmetry. Average win is $3.13, average loss is $17.18 — a 5.5:1 ratio against you. At a 97.82% win rate that's profitable, but the break-even win rate is 84.6%. That means the system stops working the moment the win rate slips 13 points. Systems with this shape don't degrade gradually; they hold the line and then break in one regime shift. 3. Maximum consecutive wins of 316. That's 43% of all 733 trades in a single unbroken winning streak with zero losses. The worst moment in the entire test was two consecutive losses. The system has never actually been under pressure, so the drawdown figure tells us very little. 4. 777,051 M1 bars is roughly 2 years. That's one market regime. There's no evidence here of behaviour across a rate cycle, a crisis, or a volatility shift. 5. Fixed lots, not compounding. Drawdown stays at ~$58 while the account grows from 1,000 to 2,947 — so the 5.64% DD figure is measured early and becomes ~2% later. It reads better than it is. None of this means the system doesn't work — it clearly made money over those two years, and to your credit there's no grid or martingale in the deal count. But the reported metrics are describing an untested curve rather than a validated edge. If you have a 10+ year backtest, an out-of-sample period, or a Monte Carlo on the trade sequence, that would say far more than these numbers do. Genuinely interested if you have them.

that looks like a clean backtest to me :)