How to spot fake volume and wash trading

Read volume against three things it cannot fake for long: liquidity depth, holder growth and trade-size distribution. A token showing $2M of 24-hour volume against an $8,000 pool is not describing a market you could exit — the same capital is being cycled. Real buying leaves holders behind it. Wash trading does not.

Almost everything published on spotting fake volume is about centralised exchanges, where the manipulation happens inside a private matching engine and outsiders infer it from statistical fingerprints. On a decentralised exchange the problem is the opposite shape: every trade is public, permanently, with wallet addresses attached. You are not inferring hidden activity. You are looking at visible activity and asking whether it represents anyone actually taking a position.

What wash trading looks like on-chain

Wash trading is buying and selling the same asset against yourself to create the appearance of activity. On a DEX it requires no exchange cooperation, no privileged access and no sophistication — a handful of funded wallets and a script are enough. It costs the trading fee and gas on every leg, which is the only real constraint on it, and on cheap chains that constraint is weak.

Four patterns show up repeatedly, and none of them requires special tooling to see.

Round trips between a small set of funded wallets. Open the pair's transaction list and read the addresses rather than the amounts. If the same ten or twenty wallets appear on both sides repeatedly, and those wallets were funded from a common source shortly before the activity began, you are looking at a closed loop. The funding step is often the most legible part: a single wallet sends similar amounts to a cluster of fresh addresses within a few minutes, and the cluster then starts trading with itself.

Volume with no holder growth. This is the single most useful check and the easiest to run. Genuine buying pressure creates new holders, because the buyers are different people. If 24-hour volume is large while the holder count is flat or drifting down, the volume is being produced by wallets that already hold the token trading with each other. Volume goes up, holder count does not, and those two facts are hard to reconcile innocently. Reading a holder list properly is covered in how to read a token's holder distribution.

Uniform trade sizes and mechanical timing. Humans trade in messy amounts at irregular intervals. Scripts do not. A transaction list where trade sizes cluster tightly around one or two values, or where trades arrive at suspiciously regular intervals, is describing an automated process. The tell is not that bots are present — bots are everywhere in this market — but that the bot activity is not responding to anything. Real bot flow reacts to price and liquidity changes. Wash flow just runs.

Volume spikes against flat liquidity. In a pool with a fixed amount of liquidity, large genuine flow moves the price and changes the pool composition. If reported volume spikes by an order of magnitude while pool depth, reserves and price sit still, the trades are either tiny relative to what the volume figure implies, or they are round trips that return the pool to where it started. The mechanics of why this is true are in how memecoin liquidity pools actually work.

The arithmetic that settles it: volume against depth

The most reliable single check needs no tools beyond a screener page, and it is arithmetic rather than judgment.

A constant-product pool cannot be drained without price impact rising steeply. If a pool holds $8,000 of liquidity, a $2,000 sell is a quarter of the pool and moves the price violently; a $2M sell into that pool is not a trade, it is a theoretical exercise. So when a token reports $2M of 24-hour volume against an $8,000 pool, the volume figure is not describing capital that entered and could leave. It is describing the same small amount of capital going round the loop hundreds of times, with each lap counted again.

Turn it into a ratio. Divide 24-hour volume by total liquidity. A ratio in the low single digits is ordinary for an active memecoin — pools genuinely turn over several times a day. A ratio in the hundreds is not a busy market; it is a small amount of money on a treadmill. There is no universal threshold, because a genuinely viral token on a thin pool can print a high ratio for a few hours, which is why the ratio is a prompt to look further rather than a verdict on its own.

Then ask the exit question, which is the one that actually affects you. Whatever the volume says, the amount you can sell without destroying the price is set by pool depth, not by the volume figure. A position larger than a few percent of the pool cannot be exited at anything near the quoted price. If the depth cannot support your intended exit, the volume number is irrelevant to you regardless of whether it is real.

Why anyone bothers

Manufactured volume is not vanity. It is an input into systems that allocate attention.

Screener trending lists, ranking feeds and sorting defaults are computed from volume, transaction counts and short-window price change. Feeding those inputs is the cheapest way to appear in front of people who are scrolling for something new, and the cost of the fake volume is small next to what a filled exit is worth. That whole economy — including the paid placements that sit alongside it — is the subject of how tokens buy their way onto trending.

There is a second motive worth naming. Where a token's supply is concentrated in wallets connected to the deployer, activity has to be manufactured before it can be sold into, because there is nobody else in the pool. Fake volume and bundled supply frequently travel together, and checking one should prompt checking the other — see bundled launches and sniper wallets.

How to actually run the check

The whole sequence takes a couple of minutes and the order matters, because each step is cheaper than the one after it.

Start on a screener and read four fields together rather than one: 24-hour volume, total liquidity, holder count and the number of transactions. Volume alone is the field most easily manufactured, and reading it beside the other three is most of the defence. If you are unfamiliar with which fields on a pair page mean what, how to use DexScreener properly covers the layout.

Compute the volume-to-liquidity ratio and note it. Then look at transaction count against volume: a very high volume figure produced by a small number of transactions means large individual trades, which a thin pool cannot support without visible price impact — check whether that impact is in the chart. Conversely a huge transaction count with modest volume and a flat price is the signature of a script cycling small amounts.

Check holder growth over the same window as the volume. Flat holders against heavy volume is the strongest single signal available to a non-technical reader.

Open the transaction list and read twenty or thirty rows of actual addresses. You are looking for repetition of the same wallets on both sides, similar trade sizes, and regular intervals. This is where a suspicion becomes something you can see.

If it still matters to you, follow the funding. On a block explorer, take three or four of the repeating wallets and look at where they were funded from and when. A common funding source shortly before the activity started is the closest thing to a smoking gun that public data offers. Supply-clustering visualisation tools do this graphically and are worth using when the wallet set is large enough that reading it by hand is impractical.

What this does not tell you

You cannot prove intent from on-chain data. This is the most important limitation on the page and the reason we do not label individual tokens. Everything above establishes a pattern — capital cycling among related wallets — and a pattern is not a state of mind. Market makers legitimately trade both sides. Arbitrage bots round-trip constantly and are a normal, useful part of how prices stay consistent across venues. A team seeding early liquidity in its own token is doing something that looks similar on a chart and is not necessarily deceptive. Public data shows you what happened, never why, and anyone claiming otherwise from transaction history alone is overstating what the evidence supports.

Wallet clustering is inference, not identity. Common funding is strong circumstantial evidence and it is still circumstantial. Exchange withdrawals, shared infrastructure and ordinary coincidence all produce clusters that are not one person.

Absence of these signals is not safety. A token with clean-looking volume, growing holders and a healthy depth ratio can still be a bundled launch, a honeypot, or simply one of the large majority of tokens that stops trading within days. These checks screen out one specific failure mode. As the survival data in why most memecoins go to zero shows, that failure mode is nowhere near the most common one.

Numbers move. Volume, liquidity and holder counts are live values, and every threshold above is a prompt to look harder rather than a rule. Figures Meme Central publishes from its own index describe tokens we have indexed, not the whole market, and are labelled that way on our analytics pages.

Frequently asked questions

How can you tell if crypto volume is fake?

Compare it against things it cannot fake cheaply. Divide 24-hour volume by total liquidity — a ratio in the hundreds means capital is being recycled, not deployed. Check whether holder count grew alongside the volume; real buying creates new holders. Then read the transaction list for the same wallets appearing repeatedly on both sides.

Is wash trading illegal on a DEX?

Jurisdiction-dependent and unsettled, and this is not legal advice. Market manipulation rules exist in most major jurisdictions, but their application to permissionless on-chain markets and unregistered tokens is contested and being litigated. Practically, no enforcement body is policing individual memecoin pairs, so the defence available to you is the check, not the rule.

What volume-to-liquidity ratio is normal?

There is no fixed threshold, and anyone quoting one precisely is overreaching. Active memecoin pools genuinely turn over several times their liquidity in a day, so low single-digit ratios are ordinary. Ratios in the hundreds are not, and warrant reading the transaction list before doing anything else. Treat the ratio as a trigger, not a verdict.

Can a token have real volume and still be unsafe?

Yes, and this is the more common case. Genuine buying tells you people are trading, not that supply is well distributed, that liquidity is locked, that the contract lets you sell, or that the token survives the week. Volume checks address one failure mode among several, and the others need their own checks.

Do screeners filter out wash trading?

Not reliably. Screeners report what the chain reports, and a wash trade is a real on-chain transaction that pays real fees. Some platforms apply filtering, none publishes a method detailed enough to verify, and paid placement products mean prominence and legitimacy are separate things on every major screener.

---

Lock the liquidity, because volume is the one number anyone can manufacture

Everything on this page exists because the headline metrics are cheap to fake and the structural facts are not. Team Finance — built by TrustSwap, which also builds Meme Central — locks LP tokens for a fixed term on Ethereum, Robinhood Chain, Polygon, Base and BNB, and the lock shows as a verified badge on the token's page in the Meme Central feed, where a buyer can check it in a second. A lock does not stop a creator selling their own allocation, and it does nothing at all about manufactured volume — it fixes one thing, verifiably, which is more than a volume figure does.

---

Nothing here is financial, legal or tax advice. Memecoins are extremely high-risk: most lose most of their value, and the majority of tokens launched never reach a decentralised exchange at all. Never spend money you cannot afford to lose entirely. Meme Central does not recommend any specific token. Data described as Meme Central's own reflects tokens indexed by Meme Central and is not whole-market data.