Koryu
← Research library
Research · 研究 · 07 · Evaluation

The feed can be edited. That is the whole review.

13 Jul 20266 min readEvaluationKoryu Research

This article evaluates a category, not named individuals, and describes incentive structures rather than accusing any specific channel of fraud. Some honest signal publishers exist; the argument below is that the format makes them nearly impossible to distinguish from the rest, which is a different and in some ways sadder problem. Nothing here is investment advice. Decisions are yours.

The direct answer: as a category, no. Telegram crypto signal groups are structurally unaccountable, independent tracking repeatedly finds their real performance far below their advertised claims, and the business model pays for audience size and trading volume, not accuracy. The longer answer is worth your time anyway, because understanding exactly HOW the format fails teaches you what verification means everywhere else.

What the format cannot prove, even in honest hands

Strip away intent and look only at the medium. A Telegram channel is an editable, deletable, unaudited feed controlled entirely by its admin. Messages can be deleted without trace for subscribers who join later. Entries can be edited after the fact. Vague setups can be posted wide ("watching this zone") and retroactively claimed as precise calls once one edge of the zone wins. Multiple channels can run contradictory calls under one operator, with marketing budget flowing to whichever history looks best. None of this leaves evidence a subscriber can detect from inside the feed. The medium offers none of the three properties a track record requires, immutability, completeness, and unforgeable timestamps, so every performance claim from inside it, however sincere, reduces to "trust me." In a market with real money at stake, "trust me" is not a track record. It is a pricing strategy.

What independent tracking finds

The gap between claimed and tracked performance is not a secret; it is documented by the industry's own review ecosystem. Trackers who log posted signals in real time and mark them against exchange prices, rather than accepting a channel's self-scored dashboard, report the same pattern year after year: advertised accuracy in the nineties, tracked accuracy near a coin flip. Review sites that exist to AFFILIATE for signal groups, hardly hostile witnesses, themselves estimate that the large majority of paid Telegram groups are scams, with commonly cited figures above 80%. When the sell side of an industry publicly concedes that most of the industry is fraudulent, the base rate has spoken. The mechanics of how a coin-flip feed gets dressed up as 95% accurate deserve their own anatomy, which we give them in the 95% lie.

Follow the money: the referral engine

The dishonesty is not a moral accident; it is what the revenue model selects for. Signal channels monetize three ways, and it pays to work through the arithmetic of the biggest one. Exchange referral programs pay the referrer a share, commonly in the tens of percent, of the trading FEES generated by every user who signs up through their link, for as long as that user trades.

Now do the math from the channel's side: a follower who trades $20,000 of volume a month at 10 basis points round-trip generates roughly $20 a month in fees, and a meaningful cut of that flows to the channel, forever, regardless of whether the follower profits. A thousand such followers are a five-figure monthly annuity that depends on exactly two things: audience size and trading frequency. Accuracy appears nowhere in the formula. Add paid promotions from token projects, effectively advertising sold as conviction, and VIP-tier subscriptions whose renewals depend on excitement rather than audited results, and the incentive landscape is complete: the format rewards entertainers with a trading vocabulary, and selection pressure does the rest.

The pump-group subspecies, on a timeline

A meaningful share of "free VIP signals" channels are not even playing the ambiguous game above; they are exit-liquidity operations with a standard choreography. Day one through twenty: organizers quietly accumulate an illiquid small-cap coin whose entire daily volume they can dominate. Announcement day: the "signal" fires to tens of thousands of followers, urgency language attached, entry "before it runs." The following minutes: follower buying spikes the price into the organizers' resting sell orders. The following days: the coin bleeds back below the entry, the channel posts its next opportunity, and the cycle repeats with the same audience funding each round. The tell is structural rather than behavioral: calls concentrated in thin, obscure pairs, urgency as a feature, and an entry price that only makes sense if someone needed buyers at exactly that moment. If you cannot identify the exit liquidity in the room, it is you.

The honest-minority problem

Here is the part that deserves more sympathy than it gets: genuinely skilled, honest publishers exist on these platforms, and the format ruins them too. Their audited-quality intentions are indistinguishable, from the outside, from the theater around them, because the medium cannot carry proof. They compete for attention against operators unconstrained by truth, which is a contest honesty loses on charisma. And their rational move, the one we would advise any of them to make, is to leave the format: publish a complete ledger somewhere append-only, commit hashes at publication time, disclose methodology, and let the record speak, per the attestation explainer. The tools are free. A provider who understands them and declines has made a choice, and the choice is information.

What legitimacy would look like, itemized

For any signal service, on any platform, the bar is concrete: a complete public ledger including every loss; entries provably timestamped before outcomes, by cryptographic commitment rather than platform metadata; a written scoring rule for what counts as a win; expectancy and drawdown net of realistic costs; methodology disclosed deeply enough to be criticized; economics disclosed, with referral conflicts named; and a record spanning at least one full bear market, because bull-phase brilliance is the cheapest commodity in crypto, per the regime-vs-skill piece. That list is demanding, and it should be: it is roughly what we demand of ourselves, and the fuller version with pass/fail phrasing is in seven verification checks.

The playbook, and the bottom line

Practically: treat any unverifiable track record as marketing copy, price the category's conceded base rate of fraud into every claim, assume urgency is a red flag rather than an opportunity, and remember that the burden of proof is cheap for anyone actually honest, publishing a hash costs nothing. When a seller who profits from your belief declines to make belief unnecessary, that refusal is the most informative data point they will ever give you. The category's answer is no. The individual exceptions are exactly the ones who can prove it, and proving it is precisely what the category refuses to do. How the format compares, object for object, with a measurement board is drawn out in the category comparison. Decisions are yours.

Related reading
EvaluationThe 95% win rate lie: how signal sellers manufacture accuracy6 min readEvaluationHow to verify a crypto signal provider isn't lying: seven checks7 min readEvaluationIs Token Metrics worth it in 2026? An honest verdict6 min read
Frequently asked

Are Telegram crypto signal groups legitimate?

As a category, no. The format is structurally unaccountable: feeds are editable and deletable, timestamps are not independently verifiable, and independent trackers repeatedly find real hit rates near coin-flip against advertised rates in the nineties. Honest publishers exist but cannot be distinguished inside the format.

How do signal groups fake their win rates?

Standard techniques: counting a first take-profit inside daily noise as a win regardless of the stop-out that followed, deleting losing calls, running multiple channels and promoting the lucky one, and posting entries without stops so no trade can ever officially lose.

How do free signal channels make money?

Exchange referral kickbacks that pay on subscribers' trading volume, paid promotions from token teams, and upsells to VIP tiers. All three reward a large, excited, frequently trading audience; none reward accuracy. Some free channels are exit-liquidity operations outright.

What would a legitimate signal service look like?

A complete, uneditable public ledger including losses; entries provably timestamped before outcomes, ideally by cryptographic commitment; disclosed methodology; results net of costs; and a record spanning at least one full bear market.

Can copy trading replace signal groups?

Exchange copy-trading at least has venue-verified records, which beats screenshots. But it is following a person, not auditing a process: leaders churn, overfit to bull phases, and blow up. The verification standard is the same either way, and few pass it.