A channel that posts majors all month suddenly names a coin nobody has heard of, with no stop and no target. The coin is not a discovery, it is a selection: the book is thin enough that a small amount of coordinated buying moves the chart, and thin books are exactly where your exit stops existing.
For weeks the feed is ordinary. Majors, a few large perpetuals, entries and stops laid out in the usual format. Then, at some specific minute, a different kind of post: a ticker you have never seen, no entry range, no stop, no target, and one line about being early.
Nothing about that message is a discovery. It is a selection, and the property it was selected for is the same property that decides what happens when you try to sell.
Why that coin, and not one you have heard of?
Because a liquid market will not move for the money available to move it.
An empirical study of coordinated Telegram pump groups, first released in November 2018 and later expanded for publication at the 28th USENIX Security Symposium, traced the message history of over 300 Telegram channels and identified 220 pump events organised through them between July and November 2018 (figures from that first version; the expanded paper covers 412 events over a longer window). It states the mechanic plainly: large exchanges carry "high bid and sell walls on the order book" which "prevent the price from fluctuating significantly even at coordinated pump-and-dump events", while smaller venues "tend to host esoteric coins with low liquidity, whose price can be more easily manipulated compared to mainstream coins". Half of the coins pumped in that dataset had a market capitalisation below 100 BTC at the time it was measured.
A larger 2023 analysis from the School of Computer Science at the University of Bristol, covering 765 coins across a dataset of 10,687 pump events, describes the same choice as a deliberate trade-off between two approaches. The "quality" approach picks "coins with lower liquidity, and therefore lower value" and produces a high percentage price increase; the "quantity" approach picks more liquid coins and accepts a smaller move. The same paper notes earlier research finding that the number of exchanges a coin is listed on correlates negatively with how well a pump works.
The regulator language matches. The US Commodity Futures Trading Commission's customer advisory tells people to avoid pump-and-dump schemes "that can occur in thinly traded or new 'alternative' virtual currencies", organised anonymously in chat rooms and messaging apps with thousands of members (Beware Virtual Currency Pump-and-Dump Schemes). The US Securities and Exchange Commission's investor education material describes the general shape: fraudsters "spread false or misleading information to create a buying frenzy", and "once the fraudsters dump their shares and stop hyping the stock, the stock price typically falls and investors lose money" (Investor.gov).
Thin is not an unfortunate side effect of a small coin nobody noticed yet. Thin is the specification.
Who is already holding it when the message arrives?
Someone is, and the price often shows it before you can read the post.
That study reports unusual return signals "even an hour before the announcement of the pumped coin", and attributes them to organisers using their own information "to purchase the to-be-pumped coin before the coin announcement". The CFTC advisory puts the consequence in one sentence: "it is the people pulling the strings who get out first making the most in the scheme."
Then there is the ordering inside the audience itself. Where a service sells tiers, the paid tier receives the name first and the free channel receives it afterwards. That is not a bonus feature, it is a queue, and everyone standing behind you in it improves your position only by arriving later than you did. The softer version of the same structure is familiar to anyone who has looked at a paywalled feed, where the public post carries the excitement and the paid post carries the fields that would let you check anything. For the wider map of who gets paid by whom in these arrangements, we set it out in where the money in signal channels actually comes from.
You are not early. You are being sold to by people who were.
What actually happens in the first minutes?
The same study dissects one event tick by tick. The first buy order was placed and filled "within 1 second after the first coin announcement", fast enough that the researchers assume automation. Eighteen seconds later the price was already at its peak. Three and a half minutes after the start, it had fallen below its opening price, and afterwards trades happened only sporadically.
Buy volume in that event exceeded sell volume, measured both in the coin and in bitcoin terms. The reading of that gap is the important part: many participants ended up "stuck" with a coin "which they are unwilling to liquidate at the low market price". Not wiped out by a stop, not closed at a bad level — simply holding, because the price at which they would agree to sell is no longer offered by anyone.
The CFTC advisory describes the same compression from the outside: these schemes "can be over in a matter of minutes", and in the example it gives, "the buy and sell cycle was over in less than eight minutes."
Why is the exit the problem and not the entry?
Because buying and selling are not symmetrical operations on a book this thin.
Entering is trivially easy, and easy for a reason: at that moment there is a crowd doing the same thing and somebody very willing to sell to all of you. Exiting requires the opposite — resting bids, below the current price, large enough to absorb what you hold. On a market chosen for having very little of that, they are not there.
The mechanics have a name and every exchange documents them. Slippage, as Binance Academy explains, "happens when a trade executes at a different price than expected", and shows up in "markets with low liquidity or high volatility"; the page notes that "in some liquidity pools with low liquidity, slippage can exceed 10% for larger trades". Its three suggestions are the honest ones: use limit orders, split large positions into smaller pieces, and prefer "assets with higher trading volume and deeper order books".
Read those against a low-liquidity call and the bind becomes obvious. A limit order at the price you want does not fill, because nobody is bidding there. A market order fills, by walking down every bid until it has taken enough, which is you moving the price against yourself on the way out. Splitting the position into pieces means the exit takes time you do not have, because the whole event lasts minutes. A position that took one click to open can take an hour to close, and by then the number that made you happy on the screen was never money — it was a mark against a book with nothing in it.
This is a different failure from arriving late at an ordinary signal, though both can happen on the same day. That one is about entering a liquid market at a worse price than the post described, and it has its own mechanics in chasing a signal after the entry is gone. This one is about not being able to leave at all.
Does waiting it out work?
It is the reflex — do not sell into the collapse, hold and let it come back — and the evidence on what these coins do afterwards is not kind to it.
The Bristol study measured the pumped assets against the wider market and found the long-term impact negative: "an average 30% relative drop in price a year after the pump event", with prices already more than 25% below the wider market at 90 days. The venue itself is part of the exposure too. As of 15 November 2018, one in five of the coins pumped on a single small exchange during that study's four-month window had already been delisted from it.
A position you cannot sell today, in an asset that on average underperforms for the next year, on a venue that may stop listing it, is not a hold. It is a receipt.
What can you check before you touch a low-liquidity call?
None of this is a view on whether any particular coin is worth owning. These are the checks that sit entirely inside your control, and each takes less time than the event lasts.
Where does it actually trade? Look up the coin's markets before anything else. One venue means one order book, one operator's uptime and one listing decision standing between you and any exit at all. The listing-count finding cuts both ways: markets that exist in only one place are the ones these events select for.
What does the depth look like a few percent from price? Open the order book and add up the resting bids within, say, two or three percent below the last trade. Compare that number with the position you were about to take. If what you intend to buy is a meaningful fraction of the bids standing under it, you are not taking a position in that market — you are the market, in both directions.
Did the call come with any level you could be wrong at? A name and a sense of urgency is not a trade. Without an invalidation there is no point at which you exit by rule, so you exit by feeling, which on a chart that peaks eighteen seconds in tends to mean not at all. A call with no stop and no target also cannot be checked afterwards by anyone, including the person who made it.
Does this post look like the rest of the feed? A channel that spends the month on majors and then names an unlisted micro-cap has not changed its opinion, it has changed its product. It is worth knowing what the channels you follow normally call, which is rarely what subscribers assume — see what coins signal channels actually call.
Whose rules are you standing inside? Exchanges prohibit this explicitly: Binance's Terms of Use bar using its services for market manipulation, naming pump and dump schemes, wash trading, self-trading, front running, quote stuffing and spoofing or layering among the practices covered, regardless of whether local law covers it, and using their chat services to manipulate a market can cost you access to them. The CFTC is blunter still: "Do not participate in pump-and-dump trades; market manipulation is against the law." Being on the buying side of somebody else's coordination is not a free option.
Is there an earlier that is safe?
No, and this is the part worth being explicit about, because the natural reaction to everything above is to try to solve it with speed — a faster alert, a bot watching the channel, the paid tier instead of the free one.
The first fill in that case study landed inside one second and the top was in eighteen. There is no tier of the same product that puts you ahead of the people who chose the coin, because the position that pays in this structure is the one that already existed before the message was written. Moving up the queue does not reach the front of it, and the front is not for sale. The CFTC's practical version is one line: "Don't purchase digital coins or tokens because of a single tip, especially if it comes over social media."
What is actually in your control
Not the chart, and not whoever is on the other side of it. What you control is whether you take a position in a market you have not looked at, and whether you check the exit before you need it.
Three questions do most of that work. Where does this trade, what is bid a few percent under the price, and what would tell me I was wrong. A call that survives them is a call you can leave. A call that does not is not an opportunity you were too slow for — it is a position built to be easy to enter, and that is the whole of it.