Findings · 2026-11-12
Telegram signal calls arrive after the move
We timed 22,632 calls from Telegram signal channels against one-minute prices. By the time the message lands, most of the move it points to has already happened.
The claim. “Get in the moment a signal channel posts and you ride the move.”
What we found. The move comes first, the post second. What is left after the post is smaller than the fees.
What we measured
We capture the messages of the Telegram signal channels on our panel as they are posted. For this test we took every long or short call from 17 June to 10 November 2026 on a coin liquid enough for a small order to fill: 22,632 calls. Each one was matched to one-minute futures prices around the moment the message appeared.
Before the post
In the hour before a call was posted, the coin had already moved 1.87% in the direction of the call on average (the typical call: 0.95%). In the last 15 minutes before the post it moved 0.74%, and in the last 5 minutes 0.30%.
Channels are not calling moves before they start. They are posting coins that are already running.
After the post
In the first minute after the message the average move was 0.04%. After 15 minutes it was 0.05%, and for the typical call it was zero. Getting in and out costs about 0.157% in exchange fees and spread. The push that followers supposedly ride is about a third of what it costs to trade it.
The test, locked before we looked
We wrote the rule down first: follow every call 60 seconds after we capture it (we see a message a median of 12 seconds after it is posted) and hold for 15 minutes. On the part of the data the rule had never seen, 11,415 trades since 1 September, the average result was a loss of 0.153% per trade after costs. One trade in three made money.
It lost in every month, in every liquidity band, and for each of the six largest channels.
Betting against the calls does not work either. Taking the other side from 30 minutes after each post and holding four hours lost 0.343% per trade, which was worse than picking the direction with a coin flip.
What this does and does not say
This is a test of the first minutes after a post, the part where "be fast" is supposed to pay. It says nothing about whether a channel's calls make money over days. That is what our panel measures, over 72 hours per call, and you can run the same measurement on any channel with an audit.
How it was tested
- 22,632 calls, first call per channel and coin in any six hours, 17 June to 10 November 2026.
- One-minute Binance futures prices. Costs: 0.107% taker fees plus 0.05% spread, plus funding.
- Rules locked 10 November 2026 before any result. Data split at 1 September: the figures quoted are from the later, unseen part.
- Uncertainty measured across days, not across trades, because calls on the same day move together.
Research on past data, not financial advice. Every test we run, failures included, is on the Proving Ground.