You cannot audit someone else's exchange account, and you do not need to. A profit screenshot makes three claims that can be checked against a price chart, and the arithmetic of leverage is unforgiving about which ones fit together.

What does a profit screenshot actually claim?

More than people notice, which is what makes it testable. A typical shared position card carries a pair, a direction, an entry price, a mark or exit price, a leverage multiple, a return percentage, and sometimes a timestamp.

Those are not independent facts. Given the entry, the exit and the leverage, the return is determined. Given the pair and the date, the entry and exit either existed in the market or did not. A screenshot is a small system of equations, and a fabricated one usually fails to satisfy it.

What the picture does not claim is equally important, and we come to that below. First, the checks you can run on the image itself.

Can you check anything without access to the account?

Yes, three things, and none of them requires cooperation from whoever posted the image. You need a price chart for the pair and a calculator.

Check What you compare What a failure means
The arithmetic Return against price move times leverage The numbers were typed, not captured
The prices Entry and exit against the pair's actual range on that date The trade was not available at those levels
The timing The screenshot's timestamp against when the channel published the call The call came after the move, or after the entry was gone

Each of the three is decisive on its own. A screenshot that passes all three is still not proof of anything - it is simply a claim that has not yet been contradicted, which is a lower bar than most people apply.

Does the arithmetic hold?

Usually this is the fastest check and it fails most often. On a perpetual futures position, the return an exchange displays is the profit measured against the margin you put up, so it scales with leverage.

The relationship is close enough to exact for a screening check:

return on margin  =  price move in percent  x  leverage     (approximately)

Run it backwards. A card showing +240% at 10x implies the price moved about 24% in the direction of the trade while the position was open. A card showing +240% at 50x implies a move of about 4.8%. Pull up the chart for that pair and that date and see which of those happened, if either.

Three failure patterns show up repeatedly.

  1. The move never happened. The implied price change is larger than anything the pair did that week.
  2. The leverage does not fit. A modest, real price move is paired with a return that would need three times the stated leverage.
  3. The pair is wrong for the claim. A large percentage is claimed on a deeply liquid major, at a size and speed that pair did not move.

Fees and funding make real returns slightly worse than the formula, never better. So a screenshot whose numbers only work if you ignore costs is already failing, and what those costs do to a small target is worked through in Following crypto trading signals: how it actually works.

Were those prices ever available?

This is the check people skip, and it is the one our own data says matters most. An entry price that appears in a screenshot has to have existed in the market, at a moment after the call was published, for the trade to have been takeable.

In our corpus of 3,167 replayed outcomes as of 7 August 2026, the single most common result is that it was not. 1,554 outcomes - 49.1% - are cancelled: price reached the profit target before the published entry ever filled. The concentration caveat travels with that figure everywhere we quote it: one high-volume channel supplies 1,197 of the 1,554, and across the other sixteen publishers the rate is 21.9%. The breakdown is in Half the trades never happen.

Read that against a screenshot. Even for calls published in full and in advance, with an entry anyone could see, the most frequent recorded outcome is a real move that no follower could have traded. A screenshot showing the profitable end of such a move is a picture of something that happened to somebody, possibly, and not evidence that the channel's subscribers were in it.

So the price check has two parts. Did the pair trade at the entry price on that date at all, and did it do so after the call went out?

What does the image deliberately not show?

The five things that would let you judge the result rather than admire it.

  1. Position size. A 300% return on $40 and on $40,000 produce identical cards. Nothing in the image distinguishes them.
  2. The rest of the account. One winning position says nothing about the nine losing ones open beside it, or the balance the account started with.
  3. Fees and funding. Both are real costs of holding a perpetual and neither appears in a position card's headline figure.
  4. Whether the account is real. Demo and testnet environments produce screenshots that look exactly like live ones, which is covered as its own flag in 10 red flags of a signal channel.
  5. Whether the trade was ever published as a call. A closed position is a report. It becomes a track record only if the entry was announced before it filled.

That last one converts the whole problem into a single question, and it is the question the rest of this article is built around.

Does a screenshot from a copy-trading page settle it?

It is better evidence than a Telegram image and it still needs reading carefully, because exchanges do not compute these numbers the same way.

Binance calculates a copy-trading lead's win rate as profitable closed positions divided by total positions, and states that partial closes are not counted as closed positions (Binance copy trading performance indicators). OKX, under the same phrase, divides days with profit by days leading trades (OKX Learn). Two numbers, one word, no comparison possible between them.

A live page has the property a screenshot never has: it updates whether or not the trader wants it to, and it covers every position rather than a selected one. That is the whole difference. What the percentage on it means is a separate problem, worked through in Win rate: why 90% accuracy can still lose you money.

Why is selection the real problem rather than forgery?

Because most screenshots are genuine, and a genuine screenshot of a chosen trade tells you nothing about the ones that were not chosen. Forgery is the easy case to worry about and the harder one is the honest picture of the best day of the quarter.

Regulators who supervise the licensed version of this activity write rules about exactly this. The CFTC's advisory on trading systems sold online tells readers to "be alert for the possibility that the system promoter manufactured results by selecting historical trades that would have yielded the greatest returns" (CFTC). In 2023 the SEC charged nine investment advisers in a single sweep over marketing-rule violations involving performance claims (SEC, September 11, 2023). The National Futures Association, having watched firms attempt to fix presentation problems with disclaimers, found that "the use of the mandated disclaimer has not prevented recurring abuses" (NFA Interpretive Notice 9025).

None of those bodies regulates a Telegram channel, and a channel owes you no disclosure at all. That asymmetry is the point: the presentation problem is serious enough to be regulated where it can be, and signal channels operate where it cannot.

What does high leverage do to these numbers?

It makes impressive percentages cheap and survival expensive, which is why the leverage figure in the corner of a screenshot deserves more attention than the return in the middle.

Of the 3,227 signals we have parsed as of 7 August 2026, 2,965 state a leverage multiple, and the highest is 200x. At that setting the whole margin behind the position is a move of one two-hundredth, which is 0.5%. Liquidation arrives a little before that point in practice, because the exchange closes the position at its maintenance margin rather than at zero. The same multiple that turns a 1.2% move into a +240% screenshot turns half a percent the other way into a liquidation, and only one of those two events gets photographed.

So a run of high-leverage winners is not evidence of skill in the way it appears to be. It is evidence that the account survived the sample you were shown. What that survival looks like when it ends is set out in What account blow-ups really mean.

The ten-minute check

Work in this order. Most screenshots fail at step 2 or 3 and the rest of the list becomes unnecessary.

  1. Read the card properly. Write down the pair, direction, entry, exit, leverage, return and any timestamp. If two of those are missing, stop: there is nothing to check, and the omission is the finding.
  2. Do the arithmetic. Divide the claimed return by the leverage. That is the price move the trade needs. Does it look plausible for that pair?
  3. Open the chart. Find the date and check that the pair actually traded at the entry price and reached the exit price. A wick counts, an approximation does not.
  4. Find the original call in the feed. It must be timestamped before the entry filled and must name the entry. A results post with no call in front of it is a report about the past.
  5. Check the entry was still reachable when the message landed. If price had already run past it, the follower's version of this trade never opened. That is our most common recorded outcome, not an edge case.
  6. Count what is missing. Size, account balance, fees, funding, the other open positions. Every absent field is a degree of freedom the poster kept.
  7. Ask for the thing that would settle it. A read-only exchange key, a public copy-trading page, or a third-party record of the calls as published. Read-only access cannot place trades or move funds, so refusing it on security grounds is not a reason, and the refusal itself is informative.

If a channel passes steps 3 to 5 on three consecutive screenshots, you have found something rare: results attached to calls that existed beforehand. The full pre-purchase sequence is in How to verify a crypto signal channel.

What this does not prove

Passing these checks does not make a channel good. It makes one screenshot consistent with one trade, which is the smallest possible unit of evidence, and a channel can produce a hundred consistent screenshots while losing money overall. Sample size and expectancy are separate questions.

Failing them does not prove forgery either. Numbers can be inconsistent because the card was cropped, because the position was scaled in and out at several prices, or because the poster is repeating a friend's trade. What a failed check establishes is that the image does not support the claim being made with it, which is all a reader needs to decide.

Our figures describe our index as of 7 August 2026: 3,227 parsed signals, 3,167 resolved outcomes, 2,965 with a stated leverage. That index is 50 channels and is not a random sample of the market. The cancelled rate quoted here is dominated by one publisher, and every figure moves as the replay catches up.

The arithmetic relationship between return, price move and leverage is a screening approximation. Fees, funding, partial fills and scaled entries all move a real result away from it, always downward on costs. Use it to find the screenshots that cannot be right, not to certify the ones that survive.

Nothing here alleges misconduct by any channel.

The practical read

Three questions turn a screenshot from an impression into evidence, and none of them requires anyone's cooperation.

Does the return match the price move times the leverage? Did that price exist, on that date, after the call was published? And where is the call?

A picture that answers all three is worth looking at. A picture that answers none is worth exactly what it cost to make. Current ratings for the channels we can score are at Signal Providers, and the rules behind them are on the Methodology page. Nothing here is financial advice.

Sources