How to judge a forex signal provider in ten minutes, ours included

Telegram has thousands of signal channels and almost none of them have a record you can audit. That is not an accident. A channel is a stream of messages the owner can edit and delete, so the default state of a signal record is "whatever the owner says it is". A provider worth paying for has built something that takes the record out of their own hands.

Six ways a record gets better than it was

1. Losses deleted. The oldest trick. A signal stops out, the message is deleted, and the month's recap counts the ones that remain. Telegram lets owners delete any message at any time with no trace. A page that cannot be read by a parser (screenshots, zones, mixed commentary) is not proof of this; it just means you have to audit it by hand.

2. Losses edited. Subtler. A stopped-out signal's message is edited afterwards so the stop was "actually" wider, or the entry was "actually" better. Telegram shows an "edited" mark, but few people look, and recaps are written from the edited version.

3. TP1 counted as a full win. The trade reaches TP1, then reverses and hits the original stop. The recap says "TP1 hit, win". A member who held for TP2 lost the full stop. Scoring only the best moment of a trade produces win rates in the high nineties from a strategy that loses money.

4. Zone entries claimed at the best fill. "Entry 2410 to 2415" lets the provider score the trade from whichever end makes it a win. Members who filled at the other end got a different trade.

5. Two channels, opposite calls. Post a buy in one channel and a sell in another. One of them was right. Recruit from that one.

6. Screenshots of accounts that are not the signals. A profit screenshot from an account is not evidence the posted signals produced it. Different trades, different sizing, or a demo.

None of these require malice at the start. A provider who is honest but scores by hand will drift toward all six under the pressure of a bad month, because each one is a small, defensible decision made about one trade.

The questions that expose them

Ask, in writing, and keep the answers:

• Where is the full history, and can I see it without joining? If the answer is a Telegram channel, the history can be edited or deleted with no trace. If it is a page on a website that updates by itself, ask the next question: what is the method, can a person correct a record, and is a corrected record marked?
• How is breakeven scored, and how is a trade that reached TP1 then stopped out scored? Any provider who has thought about it can answer in one sentence. One who has not will hesitate.
• How many signals were posted last month, and how many hit the stop? Not the win rate. The counts. A win rate without counts can be produced from three trades.
• Do you post single entries or zones?
• Is there a way for me to run last month's signals with my own take-profit rule? If the record is real, this is a feature they will be proud of. If it is not, it is a threat.

What checkable looks like

For our own record, the answers are: every signal since 2023 is on the results page, with entry, stop, three targets and the outcome, written by the system as trades close. The desk can correct a record by hand (a wrong entry price, a breakeven the system cannot see), and a record corrected that way is locked so the automatic scoring does not overwrite it. Automatic updating does not make a record impossible to edit; it makes the default state of the record the market's version, with corrections the exception. Breakeven scoring is stated on the verify page: once TP1 is reached a signal can no longer be a loss, it finishes at the highest target it reaches or at 0, and breakevens count as wins in the win rate. We say that so nobody has to discover it. The counts for the 30 days to 21 September 2026, replayed from the channel's own messages: 85 signals posted, 79 resolved, 3 stopped out before any target, 96% reached TP1, with the per-signal file published. Entries are single prices. The backtester runs any date range with your rule, and the Analyzer replays a channel candle by candle.

That is the standard. Hold everyone to it, including us, and if you find a signal on Telegram whose outcome does not match the page, send it to support with the message link.

The ten-minute check

1. Open the provider's last 30 days of messages. Count signals posted. Count "edited" marks.
2. Pick five signals at random. For each, find the outcome message. If there is none, it probably lost.
3. Find one signal that stopped out. If you cannot, the record is curated.
4. Check whether entries are single prices or zones.
5. Ask the breakeven question and the counts question.
6. If the provider has a results page, pick two signals from Telegram and find them on it. They should match to the pip.

Ten minutes. Most channels fail at step 3.

What a real record looks like

Real records have losing weeks. Ours has stops every month, and you can find the bad weeks in the history. A provider with no losing periods is either not showing you the whole record, has curated it, or has not been trading long enough for you to judge. The point of a checkable record is not that it is perfect. It is that you can see how imperfect it was, before you pay, under a method the provider has written down and cannot change quietly without the numbers moving in front of you.