Guides · Due diligence

Due diligence before buying a signal: a working process, not a hunch

ConfirmedTrades team · 18 May 2026 · 8 min read

A repeatable process for any signal or copy-trading offer: verify the source, size the risk, stress the record, check costs, test small, set an exit rule.

Signals are different from EAs

With an Expert Advisor you get the logic and run the trading bot yourself. With a signal you get the trades, after the provider has taken them, with whatever latency and slippage sits in between. That changes the due-diligence questions. The provider's discipline matters more, because you cannot inspect the logic. Execution matters more, because every trade is a copy of a copy. And the relationship matters more, because you are trusting a person to keep doing what they did.

This process treats those three things explicitly, whether the signal provider is a person or an automated system. It assumes the provider has a ConfirmedTrades page; if they do not, ask them to connect one, which is free.

Step 1: confirm the source

The record must be at least live-synced and ideally broker-verified with the trading-privileges check. The check matters especially for signals: it proves the person selling the signal controls the account that produced the record, rather than republishing someone else's investor view. Confirm the account type is Real and note the base currency. Verified vs unverified explains the labels.

Step 2: read the risk before the return

Open the Glance hub and read Max DD (balance) and Max DD (equity). A large gap between them means the provider holds losers; for a signal that is a serious problem, because when the basket is underwater you will be holding it too, on your margin. Read the Score breakdown, especially the risk-control component. Then open the AI review and look for any high-severity flag: rarely uses stop-losses, martingale-style lot increases, or average loss far larger than average win. Any one of those on a signal you intend to copy is a reason to stop.

Assume the next drawdown will be deeper than the worst one on record, and decide now whether you could sit through one and a half times it.

An equity curve above its underwater drawdown plotEquityBelow the previous peakdeepest point -17.2%Time
The underwater plot beneath the equity curve shows how far the account sat below its previous peak, and for how long. Depth is only half the story — time spent down is the half that makes people quit.

Step 3: stress the record

If Custom Analysis is enabled, use it. Remove the best month and see what is left. Remove the best symbol. Split the record into halves and compare. Filter to the hours you would actually be able to copy, if you plan to copy manually. A record that only works whole was a good period, not a strategy. If Custom Analysis is not enabled, ask the provider to turn it on; a confident one will.

Check the sample size against how long before a track record means anything. If the record is too short, follow the account and wait.

Step 4: audit the costs twice

The provider's Execution panel shows spread and slippage in points at their broker. You will pay your broker's spread plus copy latency plus your own slippage, and on a signal you also pay the subscription. Work out the provider's average win in points and compare it with your total round-trip cost. Scalping signals are usually uncopyable for this reason alone, however good the provider's own record is. Broker cost audit has the method.

Step 5: check behaviour, not just numbers

Scroll the trade history. Do lot sizes change suddenly? A provider who doubles size after a good month is managing their own equity, not yours. Are there manual interventions on an account described as automated? Is the reveal delay set, and is it consistent with the holding time? Does the provider hide sections that would be unflattering? None of these are disqualifying alone, but together they describe how the person behaves when no one is asking.

Ask one direct question: "What was your worst week, and what did you do during it?" A good provider answers with numbers and a specific action. A poor one answers with reassurance.

Step 6: test small and measure the gap

Subscribe with the smallest size that makes the numbers meaningful, and connect your copying account to ConfirmedTrades. After a few weeks, put it side by side with the provider's account. The difference between the two is the real cost of the signal: latency, slippage, your broker, and anything the provider does that does not transmit. If the gap eats most of the edge, you have learned that cheaply.

Step 7: write down the exit rule before you start

Decide, in advance and in writing, what would make you stop: an equity drawdown beyond a fixed percentage, a change in lot sizing, a high-severity flag appearing in the AI review, the provider going quiet, or the provider's account stopping syncing. The point of writing it down is that during a drawdown you will not be able to think clearly, and the provider will be telling you to hold.

A one-page summary

  • Source: broker-verified, trading privileges, Real account.
  • Risk: equity drawdown, Score risk component, no high-severity flags.
  • Stability: survives Custom Analysis without its best month and symbol; sample is large enough.
  • Costs: provider's edge minus your costs and the subscription is still positive.
  • Behaviour: stable sizing, honest disclosure, sensible reveal delay.
  • Test: small size, side-by-side comparison, measured gap.
  • Exit: written rule, decided before the first trade.

Where to look

Start on Rankings with Verified only ticked, or Explore for the live feed. The free calculators cover position size, pip value, margin and risk of ruin. If a provider is not on ConfirmedTrades, a free sign-up and the connector will put them there in an afternoon; the Help Center covers the rest.

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.

All guides