Pump and dump has moved to Telegram
Coordinated pump schemes now run through messaging groups with tiered membership, where organisers accumulate before announcing a target and each tier sells to the tier below. The mechanic is identical to a 1920s stock pool; what changed is that the entire plan is now recorded in a timestamped, attributed transcript.
The delivery mechanism for pump and dump schemes has moved from tip sheets to telephone boiler rooms to fax to email to newsletters to message boards to social platforms to messaging apps.
The mechanic has not changed once. Accumulate cheaply, promote, sell into the demand created, stop. Exchange Act § 9(a)(2) was written in 1934 to stop the stock pools of the preceding decade, and the drafting still fits without amendment.
What each new channel changes is the economics and the evidence. The current channel changed both dramatically, and in opposite directions.
The tiered structure
A modern chat group pump has an organisational feature that older versions did not: it sells the sequence in which participants learn the target.
A group recruits members, frequently with paid tiers. At a scheduled time, the group will collectively buy a named asset. Paying members receive the ticker seconds or minutes before free members. Free members receive it last.
This is advertised as a benefit. It is a description of who the exit liquidity is.
Organisers accumulate before anyone is told. Paid members buy in the first seconds and sell to free members buying in the following minute. Free members arrive after the peak and hold.
The arithmetic is not uncertain. In a thin asset, the price frequently peaks within ninety seconds of the announcement and returns near its starting level within ten minutes. Every tier’s profit is the tier below’s loss, and the bottom tier is by construction the largest.
Members join understanding this, believing they will be fast enough to sell to someone slower. Most cannot be. That is not a failure of the members’ judgement so much as a property of the structure they are joining.
Why it works better on tokens than on stocks
The technique migrated to digital assets for reasons that are structural rather than cultural.
Liquidity is thinner. A token with a $340,000 liquidity pool can double on a few hundred thousand dollars of buying. A quoted equity with even modest volume cannot.
There is no filing system. A promoted microcap can be checked against its filings — a transformative contract that exists appears in a Form 8-K. A token has nothing to check against.
Trading is continuous and permissionless. No market hours, no account opening delay, no suitability check between seeing the message and executing.
Much of it is outside any perimeter. Where the token is not a security, the securities antifraud provisions do not reach it at all.
That last point is less protective for organisers than it sounds, and this is the most important legal observation about the modern form. Wire fraud does not require the asset to be anything. Section 1343 requires a scheme to obtain money by materially false pretences over interstate wires, which describes a tiered pump exactly. Most successful criminal prosecutions in this area have proceeded that way.
The equity version did not disappear
It is worth resisting the idea that this is purely a crypto phenomenon. This library’s chat group pump facet includes equity matters — the NanoBit action records over $1.1 million.
And the older forms remain substantial. The paid stock promotion facet is one of the largest in this library. The Jammin’ Java action records over $26 million; Notis Global and Empires Consulting record $6 million each.
What has changed in equities is the distribution cost. A newsletter campaign required paying publishers. A messaging group requires an audience, and audiences are cheaper than advertising.
The evidence problem, inverted
Here is the part that should worry anyone running one of these schemes, and it is the single biggest difference from every historical version.
A boiler room left almost no record. Telephone calls, scripts destroyed, salespeople who could each testify only to their own conversations. Building a case required customer complaints, months of interviews, and inference.
A messaging group produces a complete transcript. Timestamped, attributed, retained by the platform. Recruitment pitches, tier announcements, organisers discussing timing, members complaining afterwards.
Prosecutions in this area have rested substantially on the participants’ own messages, which are frequently explicit about what is happening and to whom. In several matters the organisers described the tiered structure in writing, to strangers, as a selling point.
On-chain, it is worse still for them. Pre-announcement accumulation is permanent and public. Anyone with a block explorer can identify wallets that acquired a token before an announcement and sold during the spike — and can then check whether the same wallets did it before.
This is, by some distance, the most self-documenting fraud on this site.
Where the wash trading fits
Coordinated pumps frequently sit alongside exchange wash trading, because the volume statistics that make a token look active are themselves manufactured. The Coinbase action in this library is a venue-level matter rather than a pump, but it illustrates the general point that reported volume in this market is not always a measurement.
A pump run in an asset whose baseline volume is fabricated is a scheme layered on a scheme, and it is why the social media ramp and the chat group pump so often appear together in the same records.
What actually protects people
Not much, honestly, and it is worth being direct about that.
The structural protections that work in equities — filings, suitability rules, trading suspensions, broker gatekeeping — mostly do not exist here. The trading suspension power that lets the SEC freeze a promoted microcap for ten business days has no equivalent for a token.
Three things do help.
The advertised tiering is a confession. A group selling early access is telling you that someone paid more to be ahead of you and that someone below you will buy at a worse price. That is the entire scheme, disclosed in the marketing.
Organisers repeat. The wallets and accounts running a campaign have almost always run others. Their history is public, and it is usually a list of collapses.
Volume without holders is not demand. A token whose turnover is large relative to its distinct holder count is describing concentration rather than interest.
The unchanged part
Strip away the channel and the mechanic is exactly what Exchange Act § 9(a)(2) was written to prohibit: a series of transactions creating apparent active trading, for the purpose of inducing others to buy.
The 1920s pools traded a stock among themselves to create the appearance of activity, sold into the public interest that generated, and dissolved. The 2020s groups do the same thing faster, cheaper, and in writing.
That continuity is the reason a site organised around techniques rather than around markets is useful. The channel changes every decade. The technique does not.
What a modern campaign costs
The economics are worth setting out, because they explain why the practice persists and why enforcement targets the supply chain rather than individual campaigns.
A traditional newsletter campaign required paying publishers — six outlets, an investor-relations intermediary, video production. A quarter of a million dollars was a realistic budget, and it bought access to other people’s audiences.
A messaging campaign requires an audience you already have, or access to one. The marginal cost of telling 40,000 people about a token is close to zero. What is bought instead, where anything is, is credibility: influencer placements, or the appearance of independent enthusiasm across accounts that are not independent.
That cost reduction has two consequences.
Smaller targets become viable. A campaign that costs $5,000 can profitably pump an asset that a $250,000 campaign could not. The floor on what is worth manipulating dropped substantially.
More campaigns run in parallel. Organisers running one scheme at a time became organisers running many, which is why repeat-participant analysis is so productive an investigative route.
The organic wave, and who it makes victims of
The structural feature that distinguishes modern ramps from older schemes is that most participants are sincere.
A boiler room had salespeople who knew what they were selling. A newsletter had a publisher who knew they had been paid. A messaging ramp has organisers who know, and a much larger group of people who saw a rising price and an enthusiastic community and joined in.
Those people are not co-conspirators. They are the mechanism by which the scheme achieves scale, and they are the largest group of victims. Enforcement has concentrated on organisers, which is correct and is also the only workable approach — there is no version of this where thousands of people who bought a token because they saw a chart are usefully prosecuted.
It does mean, though, that the harm is spread across a population that mostly does not know a scheme existed. A pump and dump in a microcap leaves a few hundred identifiable victims. A token ramp can leave tens of thousands who simply believe they made a bad trade.
What would actually help
Three things, in descending order of plausibility.
Exchange-level intervention. Venues can see coordinated inflows into thin assets in real time, and several now delist or restrict assets showing the pattern. This is the fastest available lever and it does not require anyone to be prosecuted.
Repeat-organiser identification. On-chain, organisers are traceable across campaigns. Publishing that history — which several analytics services now do — degrades the credibility that recruitment depends on.
Disclosure enforcement against promoters. Section 17(b) requires disclosure of payment and amount, and it does not require proving that anything said was false. Actions on that basis have been brought against promoters of both equities and digital assets, and they are considerably easier to bring than fraud cases.
None of these reaches organisers operating anonymously from non-cooperating jurisdictions, which is a large share of them. That gap is real and this post does not have an answer to it.
For the mechanics: chat group pumps, pump and dump, and the information-based family.
Techniques referenced
- Chat group pumps
- Pump and dump
- Social media ramps
- Paid stock promotion
- Boiler rooms
- Exchange wash trading
Cases referenced
| Action | Agency | Filed | Technique | Penalty | Status |
|---|---|---|---|---|---|
| SEC v. NanoBit Limited, et al. (chat group pumps, 2026) | SEC | 2026-06-29 | Chat Group Pumps | $1.2m | judgment |
| SEC v. Jammin' Java Corp. et al. (pump and dump, 2017) | SEC | 2017-10-03 | Pump And Dump | $26.4m | appealed |
| SEC v. Notis Global, Inc. (f/k/a Medbox, Inc.), et al. (paid stock promotion, 2017) | SEC | 2017-03-09 | Paid Stock Promotion | $6m | judgment |
| SEC v. Empires Consulting Corp., et al. (paid stock promotion, 2023) | SEC | 2023-07-06 | Paid Stock Promotion | $6m | judgment |
| CFTC v. Coinbase Inc. (exchange wash trading, 2021) | CFTC | 2021-03-19 | Exchange Wash Trading , Matched Orders +1 | $6.5m | judgment |