A funded trader has a different question than a retail buyer. Not whether an EA makes money, but whether the account survives the rule book long enough to pay anything out. Every audit we publish answers that with the same replay. We take the daily balance path of the measured run, divide each daily return by a position size factor, and walk it against a drawdown floor. Then we repeat that at seven sizes. Two numbers come out per size, how often the account died and what it would have paid out per year. Across the 26 default audits in the catalog the answer is uncomfortable in both directions. Almost nothing dies, and almost nothing pays.
The profile is IQ Capital Classic in its funded form. Our copy of it is dated 2026-08-13 and carries a 6 percent end of day floor, a 0.5 percent single position loss limit and a 30 percent consistency rule. The seven sizes are full size, 1/1.5, 1/2, 1/3, 1/4, 1/6 and 1/8. Dividing every daily return by that factor is our model of trading smaller, and each audit prints all seven steps rather than a single verdict.
Three mechanics decide how those numbers read. First, the account starts at the tested deposit and every gain above it is swept out as a payout, so the model never compounds. Second, because the balance is pushed back to the start on every gain, a floor that trails the peak and a floor fixed under the start are the same line, which is why one calculation covers both rule shapes. Third, a death does not end the walk. The account resets to the starting balance and continues, so eleven deaths mean eleven fresh accounts rather than one long decline. That is exactly the view from the trader seat, where the firm absorbs the loss and the trader loses the seat.
At full size, 23 of the 26 default audits record zero deaths against the 6 percent floor. The three exceptions are Scalping Robot Pro with 11 deaths, Pulse Engine with 20, and Logan with one. Pulse Engine sits apart for a reason we name below, its audit comes from a vendor report on another broker with a 1,000 USD deposit rather than from our own test setup.
A floor almost nobody touches sounds like good news and is mostly an artifact of the test frame. Our protocol runs a 100,000 USD deposit at 0.01 base lots, which is a deliberately deep frame built to show the complete trade sequence rather than to imitate a funded account. Read the deaths column as a statement about the strategy shape, not as a promise about the account a firm would hand you.
Logan is the clean case for the standard advice. It dies once at full size and keeps dying once at 1/1.5, 1/2, 1/3 and 1/4. At 1/6 the deaths drop to zero and stay there at 1/8. The advice held. Now read the payout column of the same run. It falls 0.42, 0.28, 0.21, 0.14, 0.11, 0.01 and 0.00 percent per year across those same seven steps. The first size that survives the floor hands out 0.01 percent per year, which on the tested 100,000 USD frame is around ten dollars annually. Survival was available the whole time. It cost the entire reason for being there.
Scalping Robot Pro is the case where scaling down does not even buy that. Its deaths run 11, 7, 6, 4, 3, 2 and 1 across the ladder, so cutting the size to one eighth still leaves an account death inside a run from 2003.05.05 to 2026.08.19. The withdrawal column reads 0.00 percent at every one of the seven steps. Nothing was ever above the starting balance for long enough to be paid out. The audited run finishes at -53,234.68 USD after costs, and the sweep is the same story told from the payout side.
For the 22 audits measured in our own world on the 100,000 USD frame, the full size withdrawal figure runs from 0.00 to 0.54 percent per year. The top of that range is Gold Trade Pro at 0.54, followed by Quantum Queen X, Quantum Queen and Quantum Athena X at 0.51 each, Gold House at 0.44 and Logan at 0.42. The bottom is Smart Gold Hunter and Scalping Robot Pro at 0.00, with Lizard and OilVector X at 0.01.
The four remaining default audits print much larger figures, 318.06, 21.85, 17.54 and 1.70 percent per year. They are not better strategies. They are Pulse Engine, Market Anomalies EA, Waka Waka and Quantum Emperor, four audits built from vendor reports on a 1,000 USD deposit at a different broker. A percentage per year is a ratio against the capital base, and a 1,000 USD base is one hundred times smaller than ours. Those four numbers belong in a different column, not at the top of this one.
The single position rule caps one trade at 0.5 percent of the starting balance, and in our funded profile the second breach is the hard one. At full size 22 of the 26 audits record zero breaches. The exceptions are Logan with 15, and the three vendor report audits Pulse Engine with 360, Market Anomalies EA with 27 and Waka Waka with 7. On a 1,000 USD deposit the same rule caps a trade at 5 USD, so those counts say as much about the deposit as about the strategy.
The consistency rule is the one that catches good runs. It asks whether the best single day made up more than 30 percent of the year it belongs to, and 19 of the 26 audits breach it at least once. Smart Gold Impulse breaches in 14 years, The Gold Reaper in 12, ThunderGold Scalper in 11 and Gold Atlas in 10. Even Prop Firm Gold EA, an EA sold for exactly this use, breaches in 7 of its 24 years while never touching the floor and never breaking the single position limit.
There is a gap in that check worth knowing before you rely on it. The share is only defined when the year made a profit, so a losing year cannot breach the consistency rule in our model at all. Lizard shows both halves. Its 2010 breached with a best day of 32.56 USD against a year profit of 56.02 USD, which is 58.1 percent. Its years 2003 to 2007 were all negative, so they carry no share and no breach, however lopsided the days inside them were.
The replay walks closed trade balances, one point per weekday, carried forward on days without a close. A position held open and deep underwater produces no balance movement at all until it closes. That matters most for the strategies that pass most comfortably, because holding through adverse moves is exactly what a grid does. The daily loss rule of the FTMO profile shows the size of the blind spot where we can measure it. On Pulse Engine the end of day count is 43 breaches and the equity sampled count is 56, and on Waka Waka the end of day count is 0 while the sampled count is 1. A breach that only appears in the equity view is still a breach.
Scalping Robot Pro has no equity view at all. Its audit records the section as skipped with the reason in plain text, the reconstruction failed validation because a sampled floating drawdown of 53.3 percent exceeded the tick based report head drawdown of 29.2 percent, so the implied pricing model does not fit that report. The run with the most deaths in the sweep is the run whose intraday picture we refuse to publish. We would rather print the gap than a number we cannot stand behind.
The prop fit lamp for every audited EA is public on the category page, and the full seven step ladder sits inside each audit. Your own tester report? The browser check is free.