Are Telegram Trading Signals Legit? How to Audit a Channel

Some Telegram trading signal channels are run honestly, but almost none publish a record you can verify, and US regulators name messaging-app groups as a recurring route for investment fraud. Even an honest record shrinks for its followers: in our illustrative month, a claimed +60 pips becomes -15 pips once spread, message delay and two missed winners are counted. A trading signal (a posted buy or sell instruction with an entry, stop and target) is only as good as the time-stamped, broker-connected record behind it, losers included.

Key takeaway

No public data shows what share of Telegram signal channels are honest, but the FTC recorded $1.1 billion of reported 2025 losses to investment scams that began on social media, and the CFTC says testimonials in groups you are added to uninvited are fake. Honest records shrink too: an illustrative +60 pips over 20 signals nets a follower -15 pips after 1.0 pip of spread, 1.5 pips of delay and two missed winners. Only a broker-connected record of hundreds of signals, losers included, deserves an audit.

What do regulators say about Telegram trading signal groups?

US regulators treat messaging-app groups as a known fraud route, UK and EU regulators say unauthorised online promotion or advice can break the law, and the EU markets authority says a disclaimer does not protect the poster. The CFTC's 2024 messaging-app advisory notes that by default anyone with your number can add you to a group, and warns: "Scammers use this vulnerability to add random or targeted phone numbers to WhatsApp groups or Telegram chats."

The advisory describes pump-and-dump schemes (organised buying that lifts a price so insiders can sell into it) and promises of 100, 500 and 1,000 percent profits. On the testimonials inside such groups it is blunt: "You might also see testimonials from other group members. It's all fake, lies designed to steal your money." It describes groups you were added to uninvited, not every channel you choose to join.

The CFTC's 2019 social media advisory lists schemes that tout trade signals, which it says "often come with hidden costs or don't perform as promised." The FCA's FG24/1 guidance (March 2024) says unauthorised influencers who promote regulated products without an authorised firm's approval "may be committing a criminal offence." ESMA's finfluencer factsheet says telling people what to invest in can count as investment advice needing a licence, and adds: "Disclaimers such as 'This is not investment advice' will not protect you in these cases."

No regulator or researcher publishes the share of Telegram signal channels that are fraudulent or profitable, and lawful signal publishing exists.

How much money is lost to social media investment scams?

People reported $2.1 billion lost to scams that started on social media in 2025, and $1.1 billion of it went to investment scams, according to the FTC's April 2026 data spotlight. In the FTC's words: "In fact, $1.1 billion, more than half the money reported lost, was to investment scammers." Nearly 30% of people who reported a loss said the scam began on social media.

Reported social-media losses grew about eightfold, from $261 million in 2020 to $2.1 billion in 2025. By platform, the FTC's chart puts Facebook at $794 million, WhatsApp at $425 million and Instagram at $234 million, and lists Telegram at $111 million among the smaller platforms.

Picture three bars on one baseline: $2.1 billion for all scams that began on social media, $1.1 billion (52.4%) for investment scams, and $111 million (5.3%) for every scam that began on Telegram, with a thin line beside them climbing from $261 million in 2020. The third bar covers every scam type that started on Telegram, not signal losses, and the FTC breaks out neither Discord nor paid signal subscriptions.

Investment scams are more than half of what people reported to the FTC losing to scams that began on social media; the Telegram bar counts every scam type started there, and all of it is a floor.

These figures are floors. The FTC cites a study in which only 4.8% of mass-market fraud victims complained to a Better Business Bureau or a government body, and notes: "And since most scams aren't reported to the government, the real losses are likely much higher."

How do signal providers make money if their followers lose?

A signal channel can earn in three ways that pay whether or not its followers profit: subscriptions, broker rebates on follower volume and, in pump groups, selling into the followers' buying.

Subscriptions are paid up front. Under MetaQuotes' MQL5 Signals rules, as of October 2026, the provider receives each fee "less 20% commission retained by the 'Signals' service", and a cancelled paid subscription is not refunded; how copy fees really add up shows what that does to a small account. No primary source publishes a typical price for a Telegram VIP group.

Free channels that ask you to open a broker account through their link can be paid as an introducing broker (a referrer the broker pays on its clients' trading). One broker's partner page, as of October 2026, offers "up to $8 in rebates for each trade that your referred clients execute." That is a standard model and implies nothing about the broker; the rebate is the same on your losing trades.

The third route is the pump. Hamrick and colleagues joined crypto pump groups for a 2021 study and identified "952 pumps on Discord and 2,469 on Telegram over six months in 2018". The data are crypto and dated. Organisers buy first, the channel calls the coin, and the followers' buying is their exit.

Picture a follower on the left and three pipes running to the channel on the right: the subscription, cut 20% at the marketplace and not refunded on cancellation; up to $8 per trade from the broker, on winners and losers alike; and the market, where the organiser buys early and sells into the followers' buying.

All three income streams reach the channel whether the follower wins or loses, which is why a free channel is not free of conflict.

What is a claimed signal record worth after spread and delay?

Much less than it shows, because a follower pays costs the posted result never does: in our illustrative month, a claimed +60 pips becomes -15 pips after 1.0 pip of spread, 1.5 pips of delay slippage and two missed winners. A follower who misses nothing keeps only +10 of the 60.

No public dataset measures Telegram message delay or follower fills, so the inputs are illustrative. The month has 20 EUR/USD signals, 12 winners and 8 losers, each with a 15-pip stop and a 15-pip target, so the channel claims 12 x 15 minus 8 x 15, or +60 pips. Slippage (the gap between the price you expected and the one you got, explained in where slippage comes from) is 1.5 pips a trade, and spread is 1.0 pip.

Picture one signal on a clock: the provider enters at 0 seconds, the message posts at +20 seconds, you read it at +60 seconds and your order fills at +75 seconds, by which time price has moved 1.5 pips against you. The seconds are illustrative; only the 1.5 pips feeds the arithmetic.

The provider's fill and yours are different prices; the first 20 seconds of delay happen before the message even exists.

The table walks the claimed record down to the follower's account, one cost at a time.

Step	Calculation	Pips	Running pips	At 1 standard lot
Claimed record	12 x 15 - 8 x 15	+60	+60	+$600
Two missed winners	2 x 15	-30	+30	-$300
Spread on 18 traded signals	18 x 1.0	-18	+12	-$180
Delay slippage on 18 traded signals	18 x 1.5	-27	-15	-$270
Follower net	10 x 12.5 - 8 x 17.5	-15	-15	-$150

Computed by Phoenix Technologies from illustrative inputs; method: spread and delay charged only on the 18 signals actually traded; 1 pip on EUR/USD is $10 per standard lot (100,000 units x 0.0001). Your 10 winners pay 12.5 pips each (+125) and your 8 losers cost 17.5 each (-140), and if the two misses are random rather than winners you expect +9 pips instead.

Picture the same numbers as a waterfall on one baseline: a +60 bar, a -30 step for the missed winners, a -18 step for spread and a -27 step for delay, landing on a -15 bar below the line. Beside it stand two small bars for the kinder cases, +9 for random misses and +10 for no misses. In all three, costs eat 83% of the average claimed signal: 2.5 of its 3 pips.

Missing two winners and paying spread and delay on the signals you did take turns a winning posted month into a losing followed one.

A common shortcut charges spread and delay on all 20 signals and then also subtracts the missed winners; that double-counts, because a signal you never took cost you nothing to enter. And the margin is thin: with a 3-pip average signal and 1.0 pip of spread, delay that costs 2 pips a trade takes the follower to exactly zero.

US rules for registered advisers already name this gap. CFTC Rule 4.41 requires hypothetical results to carry a disclosure that says: "Unlike the results shown in an actual performance record, these results do not represent actual trading." A channel's posted results are unexecuted in that sense, and the rule's warning about market factors such as liquidity is your spread and slippage. The rule governs commodity pool operators and commodity trading advisors under US law and may not reach a given channel, but it names the same sim-to-live gap that breaks backtests.

One figure in search summaries claims that an independent analysis of 50 or more Telegram channels found a 65% claimed win rate against 38% real. We found no dataset, method or primary source behind it, so treat it as unsupported. The leader-to-follower gap itself is general to copying, as what followers actually earn shows.

How many signals does it take to trust a win rate?

Hundreds at least: a 60% win rate over 20 signals is statistically compatible with anything from about 38.5% to 81.5%, and even 1,000 signals leave a margin of 3 points either way. A win rate also means nothing without the payoff and the costs.

On the 15-pip stop and target with 2.5 pips of cost a trade, the follower breaks even at a 58.3% win rate (17.5 divided by 30). A posted 80% or 90% win rate can still lose money when stops are wider than targets, as the risk-reward math in R multiples shows. The table shows how wide the uncertainty around a posted 60% stays as a record grows.

Signals in record	95% margin	Plausible range for a posted 60%
20	21.5 points	38.5% to 81.5%
50	13.6 points	46.4% to 73.6%
100	9.6 points	50.4% to 69.6%
250	6.1 points	53.9% to 66.1%
500	4.3 points	55.7% to 64.3%
1,000	3.0 points	57.0% to 63.0%

Computed by Phoenix Technologies; method: normal-approximation 95% interval, 1.96 x the square root of (0.6 x 0.4 / n), with independent signals and an illustrative 60% win rate. Telling 60% apart from the 58.3% break-even takes about 3,320 signals, or 166 months at 20 a month. The channels that advertise long records are also the ones whose luck held, the selection effect behind why top-ranked records mislead.

How do you verify a signal provider before paying?

Put five questions to the record and check what kind of evidence answers each; only a record that passes all five is worth testing with your own money. Screenshots and member testimonials answer none of them.

Was every signal posted before entry, with a timestamp you can check? Edited or deleted messages fail.
Is the record pulled from a broker account, not typed by the provider?
Does it include every signal, losers included?
Did every signal carry a stop when it was posted?
Does it cover at least 100 signals?

Screenshots fail for a documented reason. The CFTC's 2019 advisory says fraudulent profiles brag about huge win percentages and post screenshots of winning trades or huge balances, and warns: "Fraud offenders can create many online profiles or anonymous identities, and use them in a coordinated way to lure people into schemes." Under Rule 4.41, even a US commodity trading advisor's testimonial must disclose that the "testimonial may not be representative of the experience of other clients".

A broker-connected record is the strongest evidence a provider can offer. On MQL5, signals can only be created from a real account, and "Signals based on demo, contest and cent accounts are not allowed." Verified still does not mean profitable for you: a broker link proves the provider's fills, not yours.

Picture a grid with the five questions down the side and four kinds of proof across the top: a screenshot, a spreadsheet, a broker-connected tracker and your own broker statement while following. The screenshot column is almost all fails, passing at most the stop question, and the spreadsheet passes only what it claims. The tracker passes the broker and completeness rows for the provider's account. Only your own statement shows what you get.

A broker link proves the provider's fills; only your own statement while following proves what the signals are worth to you.

What if a signal provider asks for your account password?

Stop. A provider who wants your login, or offers to trade your account, is no longer publishing signals but directing your account, and that changes the rules. In the US, one condition of the registration exemption in CFTC Rule 4.14(a)(9) is that the adviser is not "Directing client accounts" or giving advice tailored to particular clients.

A broadcast, non-tailored channel may be lawfully unregistered in the US, so being unregistered does not prove fraud. But the CFTC's 2024 advisory says advice from unregistered people online, or trading with unregistered firms with no US presence, can "substantially increase your fraud risk", and its 2019 advisory says to check NFA or FINRA registration and disciplinary history.

A 2019 CFTC complaint alleged that a Texas-based individual defrauded more than 140 people, partly by soliciting subscriptions to a forex trading education and signals service, and that he lured clients primarily through social media, including Instagram and WhatsApp, while touting large forex profits; those are allegations. The country-by-country picture is in what regulators say about copy trading, and nothing here is legal advice.

Can you copy Telegram signals into MT5 or a funded account?

Into your own personal account, yes, at your own risk; into a prop firm account, usually not, because many prop firms ban third-party signals and account sharing. The firm rules, and how firms detect copied signals, are in why paid signals breach prop rules.

Automating the entry does not repair the record. A signal copier can cut the time between a message posting and your fill, but not the 20 seconds between the provider's entry and the post on our illustrative clock, and it verifies nothing about the provider. Thor is built for a different job: it replicates trades from one master trading account to follower accounts, so the decision and the accounts stay with the same person and your record is your own broker statements. The difference is set out in signals versus copy trading versus PAMM.

Is a VIP signal group worth paying for?

Only if its record passes all five audit questions and your own small-size test still shows a profit after costs. Otherwise the fee buys information about the channel, not an edge.

Use these rules before you pay:

If the record is screenshots, a typed spreadsheet or member testimonials, do not pay.
If it holds fewer than 100 signals, read its win rate as plus or minus about 10 points or worse.
If the average claimed signal is not well above your own spread plus delay (2.5 pips in our example), the follower's edge is gone before you start.
If the channel wants you to open a broker account through its link, assume it earns on your volume, win or lose.
If you were added to a group you never asked to join, leave.
If anyone asks for your password or offers to trade for you, stop and check registration.
If it passes, follow at the smallest size your broker allows for 100 signals and compare your statement with the posted record before sizing up.

The honest tradeoff: a channel that fails the money test may still teach you how someone reads a market, and that can be worth a fee. Price it as education, not income, and size any copied trade knowing ESMA's factsheet warns that forex, futures, CFDs and volatile crypto can mean losing 100% of the capital you invest.

Go deeper

Frequently asked questions

Are Discord trading signals more trustworthy than Telegram ones?

No evidence says so either way. The FTC does not break out Discord losses at all, and the 2021 Hamrick study found pump groups on both apps: 952 pumps on Discord and 2,469 on Telegram over six months in 2018. Apply the same five-question audit to any chat app.

Are WhatsApp investment groups a scam?

Groups that add you uninvited often are, according to the CFTC, which says the testimonials in them are fake. The FTC describes WhatsApp groups of supposed successful investors steering people to fake platforms that show fake profits, and reported $425 million in 2025 losses to scams of all types that started on WhatsApp. A group you joined deliberately still needs the same audit.

How much does a Telegram VIP signal group cost?

No primary source publishes typical prices, so treat any quoted range with caution. What is documented is where a fee goes: on MQL5 the service keeps 20% and a cancelled paid subscription is not refunded. Judge any price against the record it buys, not the profits it promises.

Can a demo-account track record be trusted?

Only as a hypothetical. MQL5 deletes signals based on demo, contest and cent accounts, and CFTC Rule 4.41 requires simulated results of US commodity trading advisors and pool operators to carry a disclosure that they are designed with the benefit of hindsight. A demo record has no real fills, so it hides the spread and slippage a follower pays.

Do crypto pump signal groups make money for members?

The research says the price effect is small and fading. The 2021 Hamrick study found pumps "modestly successful in driving short-term price rises", an effect that diminished over time, and organisers buy before they call the coin. The CFTC's 2024 advisory describes the cooperative trading projects pitched in messaging-app groups as pump-and-dump schemes.

What should I do if I lost money to a signal group?

Report it; in the US, the FTC's loss figures are built from consumer reports to its Consumer Sentinel Network. Most victims never report, and the FTC cites a study in which only 4.8% of mass-market fraud victims complained to a Better Business Bureau or a government body. Every report makes the public picture less of an undercount.

Is Telegram riskier than Facebook for scams?

Not by reported losses: the FTC's 2025 chart puts Facebook at $794 million and Telegram at $111 million, for all scam types. Those figures show where reporting victims were first contacted, not the risk per user, and the FTC breaks out neither signal channels nor Discord. The audit, not the app, decides whether a record deserves trust.

What does a pip of slippage cost in dollars?

On EUR/USD, 1 pip is $10 per standard lot, because a standard lot is 100,000 units and a pip is 0.0001. In the illustrative month, 1.5 pips of delay slippage on each of 18 traded signals costs $270 at one standard lot. Scale it by your own lot size before you judge a signal's claimed pips.