Does a Higher Price Buy a Better Gold EA? 43 EAs, 11.5 Years, Sorted by Price Band
2 October 2026, 15:29
0
23
In July we backtested 43 popular paid gold (XAUUSD) EAs from the Market over 11.5 years, all on default settings. This post looks at one question only: did paying more get you a better result? We sorted the results by price band and by review count. No product names here; the numbers are aggregates.
Test conditions
- 43 paid gold EAs, list price $50 to $1,999 at test time (median $299)
- XAUUSD M15, 2015.01.02 to 2026.07.01 (11.5 years), MetaQuotes-Demo history, 99% quality, 1-minute OHLC model
- $10,000 deposit, 1:500, every EA on its default inputs (no vendor set files)
- Every drawdown figure below is the maximum equity drawdown over the same 11.5 years
The overall result
37 of the 43 produced a usable report (5 did not finish within our time limit because of very heavy logging on default settings, and 1 does not trade on defaults). Of the 37:
- 17 excellent (PF above 2 and max DD below 15%): median PF 4.37, median max DD 4.1%
- 11 profitable with max DD below 40%
- 1 profitable but with a 75.2% max DD, 1 with an unusually high PF of 31.55 (max DD 1.9%) that we flag for caution
- 7 blew up: PF 0.23 to 0.93, max DD 62.8% to 102.2% (median 96.6%)
Price did not sort good from bad
Median list price: excellent group $299, blown-up group $499. By price band (37 measurable EAs):
- Under $200: 14 EAs, 7 excellent, 2 blew up. Median max DD 7.55%, worst 96.6%
- $200 to $399: 8 EAs, 3 excellent, 1 blew up. Median max DD 3.9%, worst 99.8%
- $400 to $699: 7 EAs, 3 excellent, 2 blew up. Median max DD 8.0%, worst 102.2%
- $700 and above: 8 EAs, 4 excellent, 2 blew up. Median max DD 9.9%, worst 99.0%
Every band contains at least one account-killer. The rank correlation between price and PF was 0.10, and between price and max DD 0.09: close to nothing.
Review count did not either
- Fewer than 10 reviews: 18 EAs, 9 excellent, 4 blew up
- 10 to 29 reviews: 9 EAs, 4 excellent, 1 blew up
- 30 or more reviews: 10 EAs, 4 excellent, 2 blew up
Reviews are mostly written in the first weeks. The fatal drawdowns in our test sat years deep in the history.
An OHLC caveat, measured
A 1-minute OHLC model can flatter scalpers with thin take-profits. We retested the EA with the highest PF on 100% real ticks. OHLC, 2015 to 2026: PF 11.35, max equity DD 5.30%. Real ticks, 2020 to 2026: PF 3.00, max equity DD 5.60%, 703 trades. Still profitable, but the PF fell to about a quarter. Treat any PF above 10 from an OHLC test as unconfirmed. The blow-ups point the other way: they failed even under the optimistic model.
Three months later
On 2 October 2026 we rechecked all 43 product pages. 11 no longer open (5 of them from the excellent group). All 7 that blew up in our test are still listed. Of the 32 still listed, 12 raised the price, 5 lowered it and 15 kept it. Our figures are for the versions we tested in July; later versions were not retested.
What to check instead of price
1. A backtest of 10 years or more on real data, with the period stated.
2. Drawdown measured on equity, shown next to the profit for the same period.
3. Your own default-settings run from the free demo. It costs nothing.
4. A real-tick retest for any PF above 10.
Method notes: defaults only, one data feed. Your broker's spread and history will move the numbers. Five EAs could not be measured and we make no judgment about them. This is research, not investment advice. We sell gold EAs ourselves, so read our numbers with the same checklist.
Our EAs and their reports: https://www.mql5.com/en/users/app.develop.sk/seller
The full study, with the per-EA table, is on our site (English): fxea365.com/en/blog/gold-ea-43-price-quality-review
