Legitimate Market Making vs Wash Trading: What's the Difference?

A market maker takes genuine risk. It posts real bids and real asks with its own or borrowed inventory, and anyone can hit them. If the market moves against the position, the firm loses money. That risk is precisely what earns the spread, and it is why regulated venues in every asset class have formal market maker programmes with obligations attached.

The activity is legal in the overwhelming majority of jurisdictions, and on many venues it is contractually encouraged through maker rebates and designated liquidity provider schemes. Nothing about crypto changes that analysis.

Wash trading is executing trades where beneficial ownership does not change — buying from yourself through two accounts, or coordinating with a counterparty so the position nets to zero. The purpose is to print volume that misleads other participants and ranking sites. There is no risk transfer, so there is no market function.

Regulators treat this as manipulation. Enforcement across major jurisdictions has covered exactly this pattern in digital asset markets, and exchanges run surveillance specifically to detect self-matching. For an issuer, the exposure is not only regulatory: a venue that detects wash trading on your pair can delist the token and blacklist the provider.

Exchange surveillance teams and third-party monitoring vendors run a consistent set of checks against every actively traded pair. They look at self-matching across accounts that share KYC details, IP ranges, or funding sources; abnormally high fill ratios on both sides of the book from the same or linked entities; and volume that spikes without any corresponding change in holder count, social activity, or news flow.

They also compare reported volume against order book depth: a pair showing millions in daily volume on a book that never shows more than a few thousand dollars of resting depth is a textbook mismatch, because genuine volume of that size would need genuine depth to trade against. None of these checks require access to your internal records — they are run entirely from public trade and order data, which is why the pattern is so hard to disguise over any meaningful period.

Spoofing is posting orders you do not intend to fill in order to create an impression of demand or supply, then cancelling once the market reacts. Layering is the same idea across multiple price levels. Both are distinguishable from legitimate quoting by intent and by pattern: a market maker cancels and reposts constantly because prices move, but it fills when hit.

Surveillance systems look at fill ratios, cancellation latency, and whether order placement correlates with the firm's own aggressive trades on the other side. A provider running these patterns on your pair creates a record that will surface in any future exchange or investor diligence.

Not everything is black and white. Quoting with a deliberate skew — showing more size on the bid than the ask — is normal inventory management, not manipulation, unless the intent is to create a false impression. Widening spreads during volatility is prudent risk control. Providing liquidity across venues that are then arbitraged is standard practice, and the resulting volume is real.

What resolves the ambiguity is documentation: a written mandate describing quoting obligations, records showing genuine two-sided risk, and reporting that measures depth and spread rather than notional traded. Intent is what regulators examine, and your paperwork is the best evidence of it.

Write the prohibitions into the agreement explicitly: no self-matching, no wash trading, no coordinated price movement, no orders placed without intent to trade. Require depth, spread, and uptime reporting, and specifically refuse volume-based compensation. Insist on trading through dedicated exchange sub-accounts so activity is separable and auditable.

Then keep records. Board minutes approving the mandate, the KPI schedule, monthly reports, and any incident correspondence. If a venue or a regulator ever asks how liquidity on your token was provided, a clean file answers the question in an afternoon rather than a quarter.

Frameworks differ in scope but converge on conduct. The EU's markets regime treats market manipulation in crypto-assets on broadly the same principles as other instruments. Gulf and Asian regimes including VARA in Dubai, the MAS in Singapore, and the SFC in Hong Kong publish market conduct rules that address false or misleading appearances of trading activity. US authorities have pursued wash trading in digital asset markets under existing fraud and manipulation powers.

The practical implication for a founder is simple: assume the conduct standard applies wherever your token trades, and build the mandate to satisfy the strictest venue you might one day want to list on.

Before any capital moves, a compliant engagement should produce a written specification covering quoting obligations, prohibited conduct, reporting cadence, and termination rights, reviewed by counsel rather than negotiated informally over calls. The provider should be willing to disclose which entity is trading, under what registrations, and through which sub-accounts, without treating the question as intrusive.

A reference check is worth the effort it takes: ask two or three existing or former issuer clients whether reporting was delivered on schedule, whether depth matched what was reported, and how the provider behaved when markets were volatile. Providers unwilling to offer any verifiable references are telling you something about how the rest of the engagement is likely to go.

The most useful clause in a market making agreement is the one that names the conduct you never want to see, rather than one that generally promises "best practice." Specify explicitly that the provider will not trade against its own resting orders, will not route flow through affiliated or commonly controlled accounts to create the appearance of two-sided interest, and will not coordinate execution with any counterparty to leave beneficial ownership unchanged. Vague language about acting "in good faith" gives a compliance team nothing to point to later.

Pair the prohibition with teeth: an audit right to pull raw execution data from the exchange directly rather than relying on the provider's own summary, a warranty that all trading runs through disclosed sub-accounts, and a termination-for-cause clause that triggers on a single confirmed instance rather than a pattern. A provider confident in its own conduct will not resist any of this; resistance to specific language is itself useful diligence information.

Exchange listing teams and regulators approach the same question from different angles but land on similar evidence. Listing teams want to see a written market making mandate, confirmation of which entity is trading, sub-account structure, and a KPI schedule expressed in depth and spread rather than volume — because a volume target is the one number most closely associated with abuse. They will also cross-check reported activity against their own surveillance data before, not after, approving a listing.

Regulators reviewing a token retrospectively, whether prompted by a complaint or a routine sweep, tend to ask for the same paper trail: the mandate itself, board approval of the arrangement, monthly reports, and any correspondence about irregular activity. Where that file exists and is coherent, reviews close quickly. Where it does not, the absence of documentation is treated as evidence in itself, regardless of whether the underlying trading was actually legitimate.

Good record-keeping is not a compliance afterthought bolted on for auditors; it is the mechanism by which an issuer can answer a hard question quickly instead of reconstructing events under pressure. At minimum, retain the signed mandate and any amendments, board or management approval of the engagement, monthly depth and spread reports as delivered, sub-account confirmations from each venue, and a log of any incident or irregularity raised with the provider along with how it was resolved.

Set a retention period that outlasts the mandate itself — several years is typical practice — because questions about a listing's early liquidity can surface long after the original provider relationship has ended. Store the file somewhere a second person can access it without depending on whoever originally negotiated the arrangement, since that person may well have moved on by the time the file is needed.

A retainer paid regardless of performance gives a provider no reason to quote tightly once the contract is signed, which is why retainer-only deals should always carry a KPI schedule with real teeth, including a right to reduce fees or terminate on sustained underperformance. A pure loan-and-option structure, where the provider borrows tokens and is compensated through deeply out-of-the-money options, aligns incentives only while the option has some chance of finishing in the money.

The healthiest structures blend a modest retainer covering operational cost with performance-linked elements tied to depth, spread, and uptime rather than to volume or price. Whatever the structure, write down explicitly what behaviour it is designed to produce, and revisit that assumption if the desk's actual quoting does not match it after the first full reporting cycle.