How to vet an EA before you buy: the full checklist
ConfirmedTrades team · 4 May 2026 · 9 min read
Before paying for an Expert Advisor: verified record, real vs demo, drawdown, martingale and grid flags, stop-losses, sample size, costs and seller conduct.
Why a checklist
Most Expert Advisors are sold on a single big gain number and a screenshot. That number is the least reliable thing about a track record. A checklist forces you to look at the things that predict whether you will keep your money, in an order that catches the worst problems first. Most of it takes minutes on a verified profile.
1. Is the track record verified, or just typed in?
Anyone can type numbers into a spreadsheet. Look for proof the data came straight from the broker: a Broker-verified badge, which means the platform confirmed the history with read-only access, beats a self-reported record every time. A trading-privileges check adds that the seller actually controls the account and is not publishing someone else's investor view. If the record is only self-reported, ask why; the connector is free. Verified vs unverified explains each rung.
2. Real account or demo?
Demo results ignore slippage, requotes and the psychology of real money. A forex robot that shines on demo can fall apart live, and scalping or news strategies are the worst affected. Every account page shows its type and currency and flags demos. Be sceptical of a seller who only shows demo, and of a Contest account, which may have been traded with nothing to lose.
3. Drawdown and risk of ruin, not just gain
A +300% gain with a 70% drawdown is a blow-up waiting to happen. Read both Max DD (balance) and Max DD (equity) in the Glance hub; they come from a ledger that cannot be reset by deleting and reconnecting the account. Then run the numbers through the risk of ruin calculator at a risk per trade you would actually use. If a realistic risk gives a meaningful chance of ruin, the gain is irrelevant. Maximum drawdown explained covers why equity drawdown matters more than balance drawdown.
4. Watch for martingale and grid
A very high win rate paired with an average loss many times larger than the average win is the signature of martingale and grid systems: they win small, often, then give it all back in one event. Check the average win against the average loss, and whether lot sizes increase after losses. The AI review does both checks automatically and raises high-severity flags; a strategy type of Martingale or Grid in the review should end the conversation unless the seller has a very good explanation of how the tail is capped.
5. Are there stop-losses?
No stop-loss means one bad move can wipe the account. If most trades carry no stop, the equity curve looks smooth right up until it does not. The AI review reports the share of trades with a stop-loss and a take-profit. Some legitimate strategies use basket or equity stops instead of per-trade stops, so a low figure is a question to ask rather than an automatic rejection, but the answer has to be specific.
6. Sample size and duration
Fifty trades over two weeks proves nothing. Look for hundreds of trades across many months, ideally spanning different market conditions. A short, pretty curve is easy to manufacture and even easier to stumble into by luck. The monthly calendar shows how many months the record covers and how many were red. How long before a track record means anything goes deeper.
7. Concentration
One symbol, one session, one setup: the more concentrated the record, the more it may be a lucky streak rather than a durable edge. The AI review raises a watch flag when trading is heavily concentrated in a single symbol, and the Score's consistency component penalises concentration. Diversification across symbols and time is a good sign. Where Custom Analysis is enabled, remove the top symbol and see whether anything is left.
8. Costs: will it survive your broker?
The Execution panel shows spread and slippage in points for each symbol the account trades. Compare the typical spread with your own broker's quotes at the same hours, and compare the average winner with the round-trip cost. If the edge is a few points and the costs are a few points, the EA's result belongs to its broker, not to you. Broker cost audit walks through the numbers.
9. Does the record match the sales page?
Sellers often describe an automated trading system one way and trade it another. Check that the symbols, timeframe, trading hours and lot sizing on the verified record match what the sales page says the EA does. A "low-risk" EA whose verified account shows 0.5-lot trades on a $1,000 balance is not low-risk. A "set and forget" EA whose record shows manual interventions (trades with no magic number, or comments that change) is being babysat.
10. The seller's conduct
Look at how the seller handles the parts of the record that are not flattering. Are sections hidden? Is the trade history exposed with enough columns to check magic numbers and comments? Do they answer direct questions about drawdown with numbers or with reassurance? A seller who publishes a full verified record with its bad months intact is telling you something a screenshot never can.
After you buy
Run the trading bot on a demo or a small real account of your own, connect that account to ConfirmedTrades, and compare your result with the seller's over the same weeks. Any gap is either your broker's costs or something the seller is doing that the EA alone does not. You can put your own account side by side with an account you follow for exactly this purpose.
Do it in one place
On ConfirmedTrades, every published account shows its verification badges, an AI strategy analysis that flags martingale, grid and missing stop-losses, the full drawdown and risk stats, and execution costs per symbol — so you can vet a strategy before you trust it, or prove your own.