Do Forex Robots Work? An Honest Answer With Data
Do forex robots work is a question with a measurable answer, and it is not a clean yes: a handful of expert advisors have years of verified real accounts behind them, most of what is sold shows demo accounts only, and the ones that last do it with drawdowns nobody advertises.
What verified accounts show
Two opposite designs have years of real, third-party-readable accounts behind them, and almost nothing else does.
One is the grid with recovery: it keeps working for years, and it does so with falls of 25 to 85 per cent from the peak. One verified account read on 6 December 2025 had run since 2018 and showed a 66.54% drawdown with a profit factor of 1.72 over 5,043 trades.
The other is gold day trading with a stop loss on every trade. A signal read on 18 September 2026 turned €1,000 into €7,436 over three years with no deposits or withdrawals and a 12.2% equity drawdown.
What breaks most robots
Undersized accounts, not broken logic. A grid sized for €8,500 and run on €1,700 produces five times the drawdown on the same trades, and the account is liquidated by a margin call rather than by the strategy being wrong.
The second cause is a record that never existed. Of the vendors whose pages we read, one had five accounts and all five were demo, and another had deleted all three of its public signals by the time we looked.
The third is the growth percentage itself. Signals with deposits and withdrawals of fifteen to a hundred and sixty times the initial balance report percentages that mean nothing, which is why drawdown is the only column worth ranking on.
What Kestrel shows, and what it cannot yet
Kestrel is a grid expert advisor for MetaTrader 4 — what that means, in full, is on Kestrel Grid EA for MT4: Expert Advisor With Controlled Lots. It has a sixteen-year simulation and a forward test that has been running on demo since September 2026. It has no verified real-money account, and no figure on this site is presented as one.
The simulation made €26,228 on EURUSD Patient over 2010 to September 2026 — +309% of the €8,500 capital that record recommends — and it floated −€2,788 against the account at its worst instant, on 21 August 2024 (simulated, fixed lot). Like the kestrel that hangs still in the wind, it holds through that kind of episode and closes the basket once; the floating loss is the cost of the hold.
The method behind those numbers, including everything the run leaves out, is on Kestrel Expert Advisor Backtest 2010-2025: Full Report.
The question worth asking
Not whether forex robots work, but how much capital this one assumes and what its worst historical episode cost. Both answers are published per portfolio and per level on Is Kestrel a Safe Expert Advisor? Drawdown and Worst Float.
If the record convinces you, the choice is short: your pairs (1 to 4 of EURUSD, GBPUSD, EURCAD and AUDCAD), Patient, Active or both for each — Patient adds to a basket less often and holds a smaller worst case; Active adds more often, returned more in the backtest and reached a larger worst case — and the level whose recommended capital fits your account — €999, €4,999 or €9,999 per year, VAT included, on Kestrel Forex Robot Price and Licence Levels.
Frequently asked questions
Do any forex robots have a verified long-term record?
A few do. A verified Myfxbook grid account read on 6 December 2025 had been running since 2018 with a 66.54% drawdown, and an MQL5 gold signal read on 18 September 2026 had multiplied its account by 7.4 over three years with a 12.2% equity drawdown and no deposits or withdrawals.
Why do so many expert advisors blow up?
Usually because the account was too small for the strategy, not because the strategy stopped working. The same settings on five times less capital produce five times the drawdown, which is where the sixty and eighty per cent figures come from.
Is a high win rate a good sign?
On its own, no. Across the signals we read, win rates of 77 to 92 per cent always came with an equity drawdown of 31 per cent or more, because a high win rate is the signature of a system that holds losing positions open instead of closing them.
Last updated: