Memecoin rug checks: how our desk decides a coin is real before it measures it

Posted by the operator of RebateMax

I earn a referral commission from BingX and LBank when you register through my links, and both exchanges list meme coins. Nothing in this post depends on that: every check below works the same whether you trade on an exchange, on-chain, or not at all. I have no relationship with any coin, token creator or data provider named here. Every figure from our own record was read on 21 October 2026.

The short version

• A ticker is not an identity. The first version of our desk looked coins up by name, and all five coins it measured were the wrong coin. Only the contract address identifies a token.
• A pool that holds almost the whole market cap is a prop, not a market. Scam clones are built with one pool holding nearly all of the token. Our desk refuses to match a coin to any pool holding more than 40% of the token's market cap. (The test identifies the coin; once its contract is settled, a ratio that drifts over the line later does not unsettle it.)
• The price has to agree with an independent price. For a coin listed on an exchange, the on-chain price must be within 3% of the exchange's own price. In our record no namesake - a different real coin that happens to share a ticker - has ever passed it. Purpose-built clones, which quote the real coin's price on purpose, sometimes do, which is why it is one check of several.
• Read the chain off the ticker. Exchanges sometimes add the chain to a ticker (TROLLSOL is TROLL on Solana). Searching for the wrong chain finds the wrong coin.
• "No data" is not "rugged". A data provider not showing a coin proves nothing about the coin. Rugged means a measured pool with almost no money left in it.

Why this matters before any price does

Every number about a memecoin, its price, its market cap, how far it rose, is a number about one specific token. If you have the wrong token, every number after that is about something else, and it will look perfectly plausible.

This is not a hypothetical problem. When I built the Memecoin Desk, the page that tracks every meme coin listed on BingX and LBank, the first version found coins by searching their ticker and taking the most liquid match. It measured five coins. All five were namesakes: other tokens with the same ticker and more money in their pools. It showed WOJAK at a market cap of $403 million when the WOJAK our research had covered, a Solana token, was worth $2.1 million (measured 20 September 2026). There is more than one real WOJAK, too: the page now also tracks an Ethereum token of that name through LBank's published contract. Same ticker, different coins. Nothing about that number looked wrong. That is the danger.

So the desk now has one rule that outranks the rest: a coin is measured only when its contract is verified. Never by its name. The checks below are how it verifies, and each one is there because something got past the one before it.

Check 1: find the contract from someone who knows

A token's contract address is its only real identity. Anyone can create a token called PEPE in a few seconds; nobody can create a second token at the real PEPE's address.

The best source for the address is someone with a reason to get it right. Our desk accepts exactly three:

• The exchange's own published contract. LBank publishes, for each asset it lists, the contract address of the token it lists.
• Our own research. Coins our research lane has already written up carry the address that research verified.
• A price match, explained in check 3, when neither of the above exists.

On 21 October 2026 the desk held 75 verified listings: 40 confirmed through LBank's published contract, 29 through a price match against BingX's own price, and 6 from our research write-ups.

For you: take the address from the exchange's asset page, the project's own verified channels, or a block explorer you reached yourself. Never take it from a reply, a direct message, or a search result for the ticker.

Check 2: even a published contract has to agree with the market

The exchange's word is not the end of it. LBank publishes a contract for SHIBDOGE, but the token at that address traded about 100% away from LBank's own price for it. Something does not line up: a migrated contract, a different token, a dead market. We do not know which, so the desk does not guess. SHIBDOGE sits on the page with no numbers and its reason stated. Two other LBank listings are also held back, because their published addresses return no market data at all.

Our rule: a published contract still has to trade within 10% of the exchange's price before anything is measured.

For you: if the token at the address you were given trades at a very different price from the venue you are comparing it with, stop. One of the two is not the coin you think it is.

Check 3: the 3% price test

When nobody has published a contract, the desk collects the on-chain tokens with the same ticker and real money in their pools (at least $10,000 of liquidity, the largest eight), sets aside any that fail check 4 below, and compares each remaining one's price with the exchange's own price. A token counts as the coin only if exactly one candidate trades within 3% of the exchange. If none do, or if two do, the coin stays unverified.

This works because a namesake has no reason to trade at the real coin's price. The fake WOJAK and the real one were not close.

Exactly one matters. Two candidates inside 3% means at least one of them is copying the price on purpose, and the test cannot tell you which.

For you: before you buy a token you found by name, compare its price with a large exchange that lists the coin. If they disagree by more than a few percent, you have the wrong token.

Check 4: the pool that is too big

The price test has a weakness, and the desk ran into it the same day. On 20 September 2026 the desk found three tokens on Solana called PEPE, each priced within 3% of the real PEPE. They passed the check I had written.

What gave them away was the size of their pools. A real coin with a $100 million market cap might have a few million dollars of liquidity across its pools: a small fraction. A clone's creator owns the whole supply, so the clone's single pool can hold an amount close to its entire "market cap". A clone can quote any price it likes, because nobody else is trading it.

The desk's first version of this rule rejected pools holding more than half the market cap. A fake SPX slipped through with a ratio of exactly 0.5000, because its pool held the whole supply and the token side of a pool is half of it. The line is now 40%. Of the 75 verified listings the desk measured on 21 October 2026, the middle one had about 6% of its market cap in its pools, 51 had less than 10%, and the highest was 42.4%. That last one sat under the line when the coin was matched and has drifted over it since. The test identifies the coin; once its contract is settled, a ratio that drifts over the line later does not unsettle it - which is why this check never works alone.

For you: on any pool page, divide the pool's liquidity by the token's market cap. A big coin with almost all of its value in one pool is not a big coin.

Check 5: ignore the pool that quotes a crazy price

A genuine coin can still give you a fake number. On 21 September 2026 one pool for the real PUMP token quoted $21.47 while both exchanges traded it at $0.0044. Taken at face value, that one pool put the token's market cap at $21 trillion.

The desk now judges every pool against a reference price first: the exchange's own price when there is one, otherwise the middle of the pools weighted by their liquidity. Any pool more than 20% away from that reference is ignored for price, market cap and liquidity alike. If what is left still sits more than 10% from the exchange's price, or nothing is left, the desk prints no number that day.

For you: a coin's price is not whatever its most liquid-looking pool says. If one pool disagrees wildly with the rest, the pool is wrong, not the coin.

Check 6: read the chain off the ticker

BingX and LBank sometimes add the chain to a ticker: TROLLSOL is TROLL on Solana, TRUMPSOL is TRUMP on Solana. Search for "TROLLSOL" on-chain and you find nothing, or worse, an unrelated token that happens to use that name. The desk strips the suffix and accepts only candidates on that chain.

Some coins have no contract at all. DOGE runs on its own blockchain, so there is no token address to check against an exchange price. The desk lists DOGE as unverified with that reason rather than borrowing a number from somewhere it cannot check.

For you: know which chain a coin lives on before you look it up, and look it up on that chain only.

Check 7: "no data" is not "rugged"

A rug pull is when the people behind a coin pull the money out of its pools, leaving holders with something nobody can sell. The word gets used for any coin that disappears from a chart. That is a mistake.

SNEK lives on Cardano. LBank publishes SNEK's address in a short form the data source we use does not recognise, so that address returns nothing. A careless tracker would mark the coin dead. Ours holds that listing as pending, no number, reason stated, while the same coin, found at its full address through BingX, is measured and alive.

The desk calls a coin rugged only when it has measured the coin's verified pools and found less than $5,000 of liquidity left. It calls a coin faded when its market cap has fallen below 40% of where it was when we started measuring. On 21 October 2026, 46 of the 50 coins on the page measured alive and four had faded - each one's market cap is below 40% of where it began - and the page says so. None has rugged. Check the page for the current split rather than this sentence: a coin crosses the 40% line on an ordinary day, with nobody deciding anything.

For you: before you believe that a coin has rugged, find its pools yourself. Missing data is a question, not an answer.

The checklist

Run these in order. Stop at the first one that fails.

1. Get the contract address from someone with a reason to get it right: the exchange's asset page, the project's verified channels, or a block explorer. Never from a reply or a direct message.
2. Check the chain. Strip any chain suffix from the ticker and look only on that chain.
3. Compare the price with a large exchange that lists the coin. More than a few percent apart means the wrong token.
4. Divide the pool's liquidity by the market cap. Above about 40% is a prop, not a market.
5. Ignore any pool that disagrees wildly with the others and with the exchange.
6. If data is missing, find out why before you decide anything.

None of this tells you whether a coin will go up. It only tells you whether the coin you are looking at is the coin you think it is. As the memecoin guide sets out, almost every new coin dies, and these checks do nothing about that.

What our page shows today

On 21 October 2026 the Memecoin Desk showed 50 coins with verified contracts: 12 that our research lane covered and 38 tracked from their BingX and LBank listings. Seven exchange listings were shown with no numbers at all, each with its reason: DOGE twice (no contract to check), SHIBDOGE and two other LBank listings (published contract, but the market disagrees or cannot be read), and two BingX listings where no single on-chain token matched the exchange's price.

Those seven are the point. A page that showed a number for every coin would look more complete and be less true.

Questions people ask

Can these checks spot a coin that is about to rug? No. They identify the coin and catch fake numbers about it. A real coin with honest pools can still be abandoned tomorrow.

Why trust the exchange's price over the on-chain price? Because it comes from a different market, run by a company with its own money at stake in getting the listing right. We use it as the reference, not as the truth: when the two disagree, we print nothing.

Do you check coins that are not on BingX or LBank? Only the ones our research lane has covered. Without an exchange price there is no independent reference for the 3% test, so we rely on the address that research verified.

Is a coin that passes these checks a good buy? Passing means "this is the coin it claims to be". It says nothing about price, and nothing here is a recommendation.

Sources

• Our own record: the Memecoin Desk and its public data file at rebatemax.trading/data/v1/memecoins.json, read 21 October 2026. The rules quoted (3% price match, $10,000 minimum candidate liquidity, 40% pool ratio, 10% published-contract and measurement tolerance, 20% pool outlier, $5,000 rugged line, 40% faded line) are the desk's own settings on that date.
• The WOJAK, PEPE, SPX and PUMP cases were measured by the desk on 20 and 21 September 2026; WOJAK, PEPE and PUMP are also described in the memecoin guide.
• On-chain pool data via DexScreener. LBank published contracts via LBank's public asset configuration. Exchange prices from BingX and LBank's public market data.

Disclosure: RebateMax earns affiliate commission on trading fees generated by accounts registered through its BingX link and returns 50 of those 60 commission points as a lifetime discount. The full arithmetic is here. RebateMax also earns a referral commission from LBank. It has no commercial relationship with any coin, token creator or data provider named in this post. Memecoins are extremely high risk and most go to zero. This post is general information, not financial advice, and nothing here is a recommendation to trade. Think a figure is wrong? See the corrections policy.