A single-position gold breakout, measured across six market regimes before it was allowed near your challenge.

Yes. Every report is the actual MetaTrader Strategy Tester output for a $100,000 account trading XAUUSD, produced by the shipped binary with the shipped preset loaded unchanged. FundedNext was run on its own server where its history reaches and on a labelled proxy for 2023 and 2022; FTMO was run on its own server in all six regimes; Blueberry Funded was run on its own server where its history reaches and on a labelled proxy for 2022; The5ers exposes no demo server, so every one of its runs is on the FTMO feed as a labelled proxy, and the guard charges the firm's commission of about $4 per lot per side inside each run, so the days and floors are net of it while the report's profit figures stay gross; Funded Trading Plus exposes no demo server, so every one of its runs is on the FTMO feed as a labelled proxy, and the guard charges the firm's commission of about $7 per lot per side inside each run, so the days and floors are net of it while the report's profit figures stay gross; Eightcap was run on its own server in all six regimes. All 48 reports ship in your download, so you can open the same files and read every trade in them. They are backtests on past data, not a promise of future results.

Because a gold EA fitted to one year is the easiest thing in the world to sell and the hardest to keep. Every shipped preset was run, unchanged, through the 2026 trend, the choppy second half of 2025, the first half of 2025, the slow year 2024, the quiet year 2023 and the whipsaw year 2022. The settings were chosen on 2022 to 2025 and checked on 2026; the years 2006 to 2021 were never used at all and are reported year by year in every edition report. A curve-fitted system passes one window and falls apart in the others.

No. At the shipped Standard risk it reached target on every firm in 2025 H1, 2024 and 2023; in the 2025 chop none reached target, and none breached; and in 2022 none reached target and none breached. Speed is the other honest number: 2026 took 20 to 22 days across the five firms that reached it (one did not), while 2023 took 143 to 318. Nothing breached a floor in any of these runs, which is the claim this product actually makes.

Only in a trending gold market, and the page does not pretend otherwise. The lead phase reached its target in 20 to 22 days across the five firms that reached it (one did not) in the 2026 window. In 2023-type gold the same engine takes months, and in the 2025 chop it mostly waits. The regimes table is there so you can see what the market you are in is likely to give you.

Because it is the number that decides whether an engine is worth risking a challenge fee on. The same configuration, guard off, was run over every year from 2006 to 2021 on the deepest feed here, none of which was used to choose a setting; the edition reports print that table year by year, positive years and negative ones, and where a firm's own feed serves only a few of those years at minute resolution, both records are shown.

It cannot be. The engine holds one position at a time, in one direction, from one pending stop order, with its own hard stop. It never adds to a position, never holds both directions, and the lot size is a fixed percentage of equity at each entry; where a firm counts any lot increase after a loss as martingale, the guard caps the next lot at the size that just lost. Where a rung breached in testing it was because the market moved, not because the EA doubled down, and that rung was not shipped.

Each edition carries the rule its firm actually has. Minimum trading days, and profitable days where the firm counts those, are counted before the target is allowed to latch; a minimum hold keeps every ticket open at least as long as the firm requires; the engine works the hourly chart, so across the reports on this page the average hold runs from 8h 43m to 14h 53m. Of the 2456 position closes in all 48 reports, 1 were inside a minute.

That depends on what your firm requires, and each preset already matches its programme. Where a firm requires a flat book before the weekend the preset flattens on Friday at the guard's GMT hour and takes no new trade until Sunday; one input switches it off if your dashboard permits weekend holding. The news filter is off in the shipped presets because none of these programmes forbids trading through releases on an EA account; the input is there if yours does.

FundedNext: EAs require FundedNext's paid EA add-on and are not available on Free Trial or competition accounts. Grid is prohibited FTMO: EAs allowed, no fee, no time limit Blueberry Funded: EAs allowed. Any lot increase after a losing trade counts as martingale, so the guard caps it; grid, hedging and hyperactive scalping are banned The5ers: EAs allowed. Minimum hold 60 seconds; no tick scalping, HFT or news bracketing. Bootcamp requires a stop on every trade, which this build always sets Funded Trading Plus: EAs and news trading allowed; no stop-loss mandate, no minimum days, no time limit. Grid, martingale and cross-account hedging are banned Eightcap: EAs allowed. Minimum hold 60 seconds per trade; no grid, martingale or hedging. The Day Trader family is out of scope. Firms revise these terms, so confirm on your dashboard before you buy the challenge.

Neither firm offers a demo server, so their editions were laddered on the FTMO feed with every caption saying so, and their commissions were not left out: the guard charges itself the firm's published commission per lot per side inside every run (The5ers at about $4 per lot Funded Trading Plus at about $7 per lot), so every verdict, every day count and every floor on this page is net of it. The tester's own profit figures are printed gross, and the edition report says which is which.

Open your firm's folder, then load the file that matches your programme, phase and account size. A START-HERE text file in every edition names the exact file for each programme and size, and the PDF manual walks through it. Risk, the daily-loss basis, the reset clock and the floor are already set for that programme. When you move to phase two, load the phase-two preset so the guard re-anchors to the new balance.

MT5 editions for all six firms, plus MT4 for FundedNext, FTMO, Blueberry Funded. The MT4 ports are the same engine written in MQL4 and they ship one rung below MT5, validated on two regimes rather than six because that is all the MT4 gold history holds. Use MT4 only if your account offers no MT5 option.

Treat it as required rather than optional. The EA works from pending stop orders and places a hard stop on every trade, then manages it as the market moves, so it has to stay connected. A home computer that sleeps, restarts for an update or drops its connection cannot manage an open position.

Each edition is built for the symbol its firm actually quotes, so the reports above trade XAUUSD and XAUUSD.i. Attach the EA to whichever XAUUSD chart your terminal shows, on any timeframe, and it reads the hourly chart itself. If your firm renames the symbol, email support and I will tell you the one setting to change.

Switch to the funded preset for that firm, which every edition includes. Those run at a rung capped by the challenge rung, because on a funded account the job changes from hitting a target to not losing the account, and the funded runs on this page held their floor for their whole span. Where a firm restricts weekend holding on funded accounts, the funded presets keep the same Friday flatten.

They are named, not hidden. The ceiling table above shows the rung at which each firm's lead phase first breached a floor, and the list under it names every programme that was tested and not shipped, with the reason, and every rung below a shipped one that breached where the shipped one passed. A preset that cannot pass is worse than no preset, so a programme that breached at every rung simply has no file in the download.

It happens, and it is the reason this is six separate editions rather than one EA with a dropdown. When a firm changes a daily-loss basis, a reset time or a floor rule, the affected edition is rebuilt, re-laddered and the new build goes onto the same download page you were given, at no extra charge.

No. The download holds compiled EAs, the presets, the manuals, the six-regime backtest report for each edition and the 48 tester reports. If you need source for a custom build, that is a separate conversation.

Because this is an instant digital download of compiled software, it is sold as final. That is precisely why 48 full tester reports across six regimes are on this page instead of a highlight reel, and why the misses are on it too. If you are unsure which edition fits your account, email [email protected] before you buy and I will tell you honestly, including if the answer is that it is not a match.