Search for gold signals and you will find hundreds of Telegram channels, most of them promising very high win rates. You cannot check a provider's skill in advance, but you can check whether its claims are testable at all. These are the eight questions we would ask before paying anyone for XAU/USD signals, including us.
1. Does every signal have prices and a timestamp?
"Gold looks bullish, buy the dip" is commentary, not a signal. A signal you can evaluate states the direction, an entry price, a stop-loss and at least one take-profit, and it is published before the move. Without all four, nobody can tell afterwards whether it won or lost, and the provider can describe almost any outcome as a success.
2. Can you see the settled record, losses included?
Screenshots of winning trades prove nothing. Ask for the complete list of past signals with their outcome, in a form you can download and check against a price chart. If the provider trades its own signals on a live account, third-party verification services such as Myfxbook or FX Blue, connected to that account, are a stronger standard than any self-reported table.
Look at how unresolved signals are handled. A record that quietly drops trades that never reached the target or the stop flatters itself.
3. Is the win rate backed by the size of wins and losses?
A high win rate is easy to manufacture: use a tiny target and a wide stop. What matters is expectancy, the average result per signal. Suppose a service wins 70% of the time, its average win is $10 per 0.1 lot and its average loss is $30:
0.70 × $10 − 0.30 × $30 = $7 − $9 = −$2 per signal
That service loses money while winning seven trades in ten. Ask for the average reward-to-risk ratio, the profit factor (gross wins divided by gross losses) and the number of settled signals. Fewer than about a hundred tells you very little.
4. How deep was the worst losing run?
Maximum drawdown, the largest fall from a peak before a new high, tells you what following the service would have felt like at its worst. Combine it with the position size you plan to use. A drawdown of 40 signals' worth of risk is survivable at 0.5% per trade and fatal at 5%.
5. Is the record about gold, and does it cover rough markets?
A blended record of forex, crypto and indices says little about XAU/USD. Gold's daily range in 2026 has been roughly double what it was in early 2025 (see how far XAU/USD moves in a day), so check that the history includes a volatile stretch such as February and March 2026, not only quiet months.
6. Can you act on the signal in time?
Gold can move several dollars within seconds of a US data release (see when gold moves most during the day). Ask how signals are delivered, how quickly, and whether the published entry is realistic once your broker's spread and slippage are included. A signal that only works at a price you could never have been filled at is not a signal.
7. Who holds the money?
A signal subscription should never require you to send trading funds to the provider. Your money belongs at a regulated broker in an account in your name. Managing other people's money generally requires a licence from the local regulator, such as Bappebti in Indonesia or the FCA in the UK, and an unlicensed "account management" offer is a red flag on its own.
Some free signal groups are paid by a broker for every client who opens an account through them. That is not wrong in itself, but it changes the incentives. Check the broker's regulation yourself rather than relying on the channel's description of it.
8. What happens if you want to stop?
Read the cancellation terms before you pay. Monthly plans you can cancel at any time are easier to test than prepaid lifetime access. Promises of guaranteed returns, "no losing weeks" or pressure to upgrade before you have seen a full month of results are reasons to walk away.
A five-minute test
- Pick any five past signals from the provider's record.
- Open a chart at the publication time of each one.
- Check that the entry price was actually available, and which of stop or target was reached first.
- Compare your result with the provider's.
If you cannot do this because the record does not exist in a checkable form, you have your answer.
How MindAIGold answers these questions
Every MindAIGold signal carries a concrete entry, stop-loss and take-profit, and its outcome is settled automatically when price reaches one of them. Settled results, including losses, can be inspected with a free account in the track record, with CSV export. Orders go to your own MT5 account; trading funds stay with your broker, and paid plans can be cancelled at any time. None of this guarantees profit, and we would rather you run the five-minute test above on us as well.
This article is general information, not investment advice. Trading leveraged products carries a high level of risk. See our risk disclosure.