MQL5 Market Census of 15,032 EAs | 1,050 Gone in 8 Weeks

A Full Census of 15,032 EAs on the MQL5 Market — 1,050 Disappeared in 8 Weeks

"How many EAs are there on the MQL5 Market, what is the going price, and how far can they be trusted?" To answer this simple question, we used our own crawler to scan every page of the MQL5 Market's MT5 EA category and obtained the list of all 15,032 products listed as of September 3, 2026. We then compared it against a snapshot of 13,950 products taken about 8 weeks earlier (July 2026) to measure market turnover as well.

All data in this article is primary data collected by our site. This is not a performance evaluation of individual products; it shows the structure of the market as a whole in numbers.

Survey Overview

Statistics on review counts, ratings and supported symbols are based on the detailed collection as of July 2026 (13,991 products); statistics on prices, totals and turnover are based on the full collection of September 3, 2026.

Finding 1: 1,050 Products (7.5%) Disappeared From the Market in About 8 Weeks

Matching the July snapshot (13,950 products) against the September 3 full list (15,032 products) by product ID:

In about 8 weeks, 1 in 13 products vanished from the market, and an even larger number were newly added. Extrapolated naively, annual turnover runs into the thousands. "An EA that was on sale yesterday no longer exists today" is everyday life in this market. Buying an EA without being able to check a long track record, only to find its sales page gone a few months later, is not an exceptional accident — statistically, it is a commonplace event.

Finding 2: The Median Is $149.99, Yet 1,562 Products Cost $999 or More

Distribution of displayed prices for the 13,346 paid products (as of September 3, 2026):

Products by price band:

Half can be bought for $150 or less, yet 1,562 products cost $999 or more, pulling the mean ($629.78) up to more than four times the median. The most common price was exactly $30 (1,419 products).

What matters is that, in our own measurements, we have found no correlation between price and performance. In our mass test of 43 gold EAs, a $1,999 EA wiped out its account over the test period, while the highest PF came from a $199 EA. At least from our testing, there is no data supporting "expensive means good."

Finding 3: 85.6% Have Zero Reviews

Of the 13,991 products with detailed data as of July 2026, only 2,019 (14.4%) had at least one review. The remaining 11,972 (85.6%) had zero reviews.

Looking further at the 2,019 products that do have reviews:

• The median rating is 4.69/5.0, with 56.8% at 4.5 or above
• The median number of reviews is 2

In other words, it is a world where "almost every product with reviews is highly rated, but with only a handful of reviews." Poorly rated products disappear from the market before they get bought (Finding 1), so the review distribution of surviving products is inevitably skewed toward high ratings. The conclusion the numbers point to is that star ratings barely function as a basis for choosing an EA.

Finding 4: Over 40% Are Gold (XAUUSD) EAs

Supported symbols tallied from the detailed data as of July 2026 (where identifiable):

Over 40% of products whose symbol could be identified are gold EAs, more than 3.5 times the second-place EURUSD. Supply is concentrated in gold, whose high volatility makes it easy to produce "flashy backtest curves." More supply also means wider variation in quality. In fact, we have summarized the results of testing 43 gold EAs on real ticks under identical conditions in this article.

Choosing an EA in This Market

Here is what the census of 15,032 products tells us.

1. Market turnover is intense (7.5% gone in 8 weeks). The "length" of a sales track record is itself a rare signal.
2. Price is not a proxy for quality. In a market with a median of $149.99 there are 1,562 products at $999 or more, but no correlation between price and measured performance has been confirmed.
3. Review sample sizes are too small. 85.6% have zero reviews, and those that do have a median of 2. Look at test data rather than stars.
4. Gold EAs are oversupplied. The more choices there are, the more essential comparative testing under identical conditions becomes.

On this basis, our site evaluates EAs not by the numbers on sales pages but by backtests under uniform conditions on Dukascopy real ticks. See our Testing Methodology for details of the evaluation method and the EA Ranking for the test results of our own EAs. Measured reviews of individual EAs are published in series such as the 10-year real-tick test of Ox7F.

Survey data: our own crawl (full collection of 15,032 products on September 3, 2026; detailed collection of 13,991 products in July 2026). The original data has been preserved. Figures in this article are based on public information at the time of collection and may differ from current listings.