Manipulation, insider trading and fraud: the distinctions that matter
Market manipulation falsifies the price signal. Insider trading exploits true information obtained in breach of a duty. Investment fraud takes money on a false promise without touching a market at all. All three are securities fraud in the broad sense; only manipulation corrupts the price, and the distinction determines which statute applies and who was harmed.
“Market manipulation” is used in ordinary speech to mean roughly something unfair happened in a market. The legal term is far narrower, and the gap between the two produces most of the confusion in this area — including in coverage written by people who ought to know.
This post sets out three categories that are constantly conflated, and why the difference is not pedantry.
The test
Market manipulation falsifies the price signal.
That is the whole distinction. Manipulation causes a market to produce a price that misdescribes supply, demand, or what participants actually believe. A spoofed order says there is demand that does not exist. A wash trade says a transaction occurred between parties who never took risk. A rigged benchmark reports a rate nobody would pay.
Conduct that does something else wrong — however seriously — belongs in a different category.
Insider trading: true information, breached duty
Insider trading puts nothing false into the price. The insider trades on information that is entirely accurate, and their trading moves the price toward what it will be once the information is public.
The wrong is not falsification. It is that the information was obtained and used in breach of a duty — to an employer, a client, a source, or a family member.
That has a consequence people find uncomfortable: some economists have argued that insider trading improves price efficiency, and in a narrow sense they are right. The legal system’s answer has been consistent — efficient prices obtained by breach of duty are not a defence, and a market whose participants are advantaged by position rather than by analysis is one people reasonably decline to enter.
The enforcement record shows the scale. This library records the Trafigura matter at $55 million and the Yin action at $39.5 million. These are serious cases. They are not manipulation cases.
Why it gets conflated: both are charged under Rule 10b-5, and a single enforcement release frequently covers both. This library’s records carry separate tags for exactly that reason.
Breach of duty without any price effect
Three techniques sit here, and none of them touches a market price.
Front running takes the price improvement a customer’s order would have produced. The customer is worse off by a measurable amount; the market price is exactly where genuine supply and demand put it, because the large order really was coming.
The crypto case makes this vivid. Sandwich attacks are economically identical to front running — someone sees a pending trade and profits by transacting around it — but there is no broker, no customer, no duty, and no confidential information. The consequence is that the legal position is genuinely unsettled. If the conduct were manipulation, the absence of a duty would not matter. That it does matter tells you which category it belongs in.
Churning is excessive trading in a customer’s account for the broker’s commissions. Every trade executes at the prevailing market. The victim is one named person.
Ponzi schemes pay earlier investors from later investors’ money. There is usually no trading at all, and frequently no real asset. It is investment fraud of a very pure kind and has nothing to do with price formation.
Why the distinction has consequences
Not taxonomy for its own sake. Four practical differences follow.
Different provisions apply. The express anti-spoofing statute reaches manipulation in futures and nothing else. Section 9(a)(2) reaches transactions inducing others to trade. Rule 10b5-1 and 10b5-2 address insider trading specifically. FINRA Rules 5270 and 5320 address front running. Regulation Best Interest addresses churning. Getting the category wrong means citing the wrong rule.
Different elements must be proved. Manipulation frequently requires establishing intent to affect a price — and in the traditional commodities claim, that the price was artificial, which means saying what it should have been. Insider trading requires materiality, non-public status and breach of duty. Churning requires excessive trading judged against the customer’s objectives.
Different victims. Manipulation harms everyone reading the price. Insider trading harms the counterparty and confidence generally. Churning and front running harm one identifiable customer, who can sue.
Different remedies. Where the victim is identifiable, restitution is possible. Where the harm is a distorted market-wide price, identifying who was harmed and by how much is frequently impossible, which is why Fair Funds are used less in manipulation cases than people expect.
The hard cases
Three places where the line is genuinely difficult, and worth naming rather than glossing.
Open-market manipulation. Marking the close and momentum ignition involve trades that are individually lawful, executed at real risk, at real prices. The manipulation lies entirely in purpose. Courts have divided on how far intent alone can convert lawful trades into unlawful manipulation, and the doctrine is not settled.
Insider trading that is not securities fraud. Trading ahead of a token listing fits the misappropriation theory precisely — except that if the token is not a security, Rule 10b-5 does not reach it. The successful charge has been wire fraud, which does not care what the asset is.
Naked short selling. The debate confuses a settlement failure with a manipulation scheme almost universally. Most failures to deliver are operational. Deliberate naked shorting to depress a price is manipulation and is charged as such. Treating failure-to-deliver data as a manipulation count is the single most common error in the area.
The one thing to remember
Manipulation lies to the market about what is happening. Insider trading keeps a truth from the market that it is entitled to have. Investment fraud takes money on a promise about something that is not happening at all.
All three are serious. All three are prosecuted. Only one corrupts the price, and that is what the word means.
A worked test: five scenarios
The categories are easier to hold onto applied than described. Here are five, with the answer and the reason.
A trader posts 6,000 contracts they intend to cancel, sells 200 on the other side, and cancels. Manipulation. The order communicated demand that did not exist, and everyone reading the book acted on it. This is spoofing, and in futures it is prohibited by name.
A lawyer learns of a takeover from a client file and buys call options. Insider trading. Nothing false entered the market; the option price moved toward what it will be once the deal is announced. The wrong is the breach of a duty owed to the client, reached through the misappropriation theory.
A broker buys 90,000 shares ahead of a client’s 400,000-share order and sells into the fill. Neither, in the manipulation sense. The client’s order genuinely moved the price. The broker took the improvement that belonged to the client, which is front running — a breach of duty to one person.
A promoter accumulates 6 million shares, funds a $250,000 campaign, and sells into the buying. Manipulation, and several other things. The campaign’s claims are false statements; the supporting trades create apparent active trading; and the distribution of a control block is an unregistered offering. A single pump and dump generates three or four theories at once, which is why enforcement releases in this area name so many provisions.
An operator takes $10 million promising a trading strategy, trades nothing, and pays returns from new deposits. Investment fraud, and nothing to do with any market. No price was affected because no security was bought. Ponzi schemes are prosecuted under the antifraud provisions and the wire fraud statute, not under the manipulation provisions.
Why “securities fraud” is not the answer
A tempting response to all of this is that the categories collapse anyway, because Rule 10b-5 covers everything.
It does cover a great deal — manipulation, insider trading, misstatements and schemes all fit within it. But the elements differ inside that single rule, and so does everything procedural.
A manipulation claim has to establish that the conduct affected or was intended to affect a price. An insider trading claim has to establish materiality, non-public status and breach of duty, and none of those is about price at all. A misstatement claim has to establish falsity and reliance.
Beyond Rule 10b-5, the specific provisions diverge sharply — the express anti-spoofing statute, the touting provision, the tender offer rules, the adviser fraud provisions. Reaching for the general rule when a specific one exists is usually a sign that the specific elements could not be made out.
So “it is all securities fraud” is true at a level of generality that is not useful for anything: not for deciding what a regulator must prove, not for identifying who was harmed, and not for understanding what actually happened.
Where this taxonomy came from, and its limits
A word about the classification this site uses, since the whole argument above rests on it.
Regulators do not classify by technique. They charge statutory provisions — Section 10(b), Rule 180.1, Section 9(a)(2) — and a single release will frequently describe conduct that spans several of the categories in this post without ever naming any of them.
The technique taxonomy is therefore ours, and it is an editorial judgement rather than a legal fact. We apply it because it makes comparable conduct comparable across agencies, years and asset classes in a way the charging language does not. A spoofing case brought by the CFTC under the express provision and an equity layering case brought by FINRA under its own rules describe the same behaviour, and only a technique tag connects them.
Two limits follow, and both are stated on every relevant page.
Tags are applied by keyword rules over the regulator’s text, then reviewed. Conduct described in unusual language will be missed; conduct mentioned in passing may be over-tagged.
A case tagged with several techniques appears in each. Summing across facet pages double-counts, which is why this site never presents such a total.
The editorial policy sets out the rules in full, and the keyword file itself is in the repository. If a tag looks wrong to you, it may well be — the corrections process is the fastest route to fixing it, and corrections are logged in public.
This site’s technique taxonomy organises everything around that test, and the related but distinct section exists specifically to hold the things that fail it — because a reference work on manipulation that says nothing about the conduct constantly mistaken for it is less useful, not more focused.
Techniques referenced
Cases referenced
| Action | Agency | Filed | Technique | Penalty | Status |
|---|---|---|---|---|---|
| CFTC v. Trafigura (insider trading, 2024) | CFTC | 2024-06-17 | Insider Trading , Price Manipulation | $55m | judgment |
| SEC v. Shaohua (Michael) Yin, et al. (insider trading, 2024) | SEC | 2024-08-30 | Insider Trading | $39.5m | judgment |
| CFTC v. HSBC Bank USA (spoofing, 2023) | CFTC | 2023-05-12 | Spoofing | $45m | judgment |
| SEC v. Jammin' Java Corp. et al. (pump and dump, 2017) | SEC | 2017-10-03 | Pump And Dump | $26.4m | appealed |
| CFTC v. Five Banks (benchmark submission rigging, 2014) | CFTC | 2014-11-13 | Benchmark Submission Rigging , FX Fixing +1 | $1.4bn | judgment |