Market manipulation is intentional conduct designed to interfere with the free and fair operation of a market and create an artificial or misleading appearance of price, volume, or supply/demand — so that the manipulator profits at the expense of participants who act on the false picture. This section is the umbrella for that domain: what the recognized types are, how each is mechanically formed, how surveillance teams detect them, and — for an analyst — how to avoid mistaking manufactured activity for genuine signal. The defining tension that runs through every subtype is intent: most of the observable behaviors (placing and cancelling orders, buying and selling the same asset, trading ahead of a move, promoting a stock) are, in isolation, perfectly legal and ubiquitous. What converts them into a federal offense is the deceptive purpose behind them — which is invisible in the data and notoriously hard to prove. That gap between observable form and hidden purpose is the central problem of the whole field, both for prosecutors and for anyone trying to read the tape.

What this section covers

The four child nodes document the most recognizable and most-prosecuted archetypes, organized by the mechanism of deception (a standard way the academic and regulatory literature carves up the field):

• [Pump & Dump] — an information-based / trade-based scheme: accumulate a thin asset, inflate its price with false hype and coordinated buying, then sell into the manufactured demand. The canonical retail-fraud archetype, now migrated heavily toward low-cap crypto. See the child for measured base rates (Frieder & Zittrain's −5.5% late-buyer / +4.3% spammer finding; crypto Telegram-pump studies).
• [Spoofing & Layering] — order-based manipulation: post large non-bona-fide orders to fake supply/demand and move price toward a genuine order on the other side, then cancel. Made an explicit standalone offense in U.S. derivatives markets by the 2010 Dodd-Frank Act (CEA §4c(a)(5) via §747); in equities it is reached through general fraud statutes. Landmark cases (Coscia, Sarao, JPMorgan's ~$920M resolution) are in the child.
• [Wash Trading] — trade-based manipulation with no real change in ownership or risk: buy and sell to yourself (or a confederate) to fake volume/liquidity. Rare in raw terms on regulated venues, but the dominant integrity problem on unregulated crypto/NFT venues (Bitwise's ~95%-fake-BTC-volume finding).
• [Front-Running] — information/duty-based abuse: trade ahead of a known, market-moving order, misusing non-public information about a pending transaction one has a duty to handle faithfully. The child carefully separates true fiduciary front-running from the contested "HFT front-running" / latency-arbitrage framing of Flash Boys.

The broader taxonomy (and what's out of scope here)

These four are not the whole landscape. The regulatory and academic literature also recognizes cornering/squeezing (amassing a dominant position to dictate price — a market-power technique), marking the close / painting the tape (trading at the close or in bursts to set a misleading print — a classic open-market manipulation), benchmark manipulation (e.g., LIBOR, FX-fix rigging), churning (excessive broker trading for commission), and insider trading (trading on material non-public information — usually treated as its own distinct category, not "manipulation" in the price-distortion sense). Where those touch the four covered types they are cross-linked, but their depth belongs to their own nodes. A useful organizing distinction from the case law: traditional manipulation rests on an objectively improper "bad act" enumerated in Securities Exchange Act §9(a) (fictitious trades, wash sales, matched orders), whereas open-market manipulation involves otherwise-legal trades that become unlawful only in context and intent — the harder, more contested category to prosecute.

The legal spine

In U.S. equities, manipulation is reached primarily through Securities Exchange Act §9(a) (the specific anti-manipulation provisions — wash sales, matched orders, price-rigging "for the purpose of inducing" others) and the broad catch-all §10(b) and Rule 10b-5. Section 9 is rarely charged directly; prosecutors lean on 10b-5's general fraud framework. Derivatives fall under the Commodity Exchange Act and the CFTC (which gained an explicit spoofing prohibition via Dodd-Frank §747). Enforcement is shared across the SEC, CFTC, DOJ, FINRA, and exchange-level surveillance, and has been a clear regulatory priority over the past decade. Each child carries its specific statutory hooks.

How detection works (the common thread)

Because intent is invisible, surveillance keys on behavioral fingerprints and economic substance, not on any single trade:

• Anomaly detection — abnormal deviations in price, return, volume, and trade count over short windows (the core signature of pump-and-dumps and momentum ignition).
• Order-flow asymmetry — very high cancellation rates and extreme order-to-trade ratios correlated with same-side genuine fills (the spoofing/layering tell).
• Linkage and circularity — account/address clustering, common funding sources, self-loops and circular flows that reveal a single controlling entity on both sides (the wash-trading tell).
• Trade-ahead sequencing — reconstructing whether a desk traded for its own account ahead of a held customer order (the front-running cousin).

Modern surveillance increasingly layers machine-learning classifiers on top of these rules. The unavoidable limitation, repeated across every subtype: these are flags, not proof. Legitimate market-making cancels most of its orders; news-driven momentum spikes look like pumps; offsetting trades happen innocently. The law requires deceptive intent, so surveillance must show the pattern correlates with the manipulator's own benefit — which is why on-chain estimates and cancellation-rate accusations are so often contested.

When it matters vs. when it doesn't

Manipulation risk is highly venue- and liquidity-dependent. It concentrates where the deceptive activity is cheap and hard to trace: thin float, low surveillance, pseudonymous accounts, and venues that treat displayed volume or depth as trusted signals. That means OTC/microcap equities and unregulated crypto/NFT markets are the high-risk corner; deep, surveilled, large-cap U.S. equity markets are comparatively (not perfectly) clean. For most regulated-market analysis the practical upshot is defensive and modest — not paranoia about every move, but skepticism toward two specific inputs: displayed order-book depth (can be spoofed) and reported volume on lightly regulated venues (can be wash-traded).

Sources

• Cornell LII — Securities Exchange Act of 1934 (§9, §10(b), Rule 10b-5 overview)
• Emory Law Journal — The New Market Manipulation (taxonomy of types, traditional vs. open-market)
• Mondaq — Open-Market Manipulation Under SEC Rule 10b-5 (traditional vs. open-market distinction; marking the close / painting the tape)
• Association of Corporate Treasurers — A short history of market misconduct techniques (cornering, squeezing, front running, pumping-and-dumping, benchmark distortion)
• SIX — 6 Types of Market Abuse
• Sullivan & Cromwell — CFTC Enforcement Priorities (spoofing, disruptive closing-period trading, wash trading as priorities)
• Child nodes (full sourcing within each): Pump & Dump; Spoofing & Layering; Wash Trading; Front-Running.

Notes: This is a section-overview node — measured base rates, case law, and statutory detail live in the four child docs and are not duplicated here. The "trade-based / order-based / information-based / action-based" mechanism taxonomy is one common scholarly framing among several; it is a useful organizer, not a settled legal classification. The line between manipulation and aggressive-but-legal trading is genuinely contested in every subtype (intent problem), and the boundaries of "front-running" (HFT/latency arbitrage) and "fake volume" percentages (crypto/NFT) are actively disputed — see the relevant children.