A memecoin is a crypto token built around a joke, character or trend, usually with no product or cash flow behind it, so its price depends on attention and trading demand. A rug pull is when the people behind a token take buyers' money and leave. You can catch many of them by checking who can mint or freeze the token, whether liquidity can be withdrawn and how much supply a few wallets hold.
What makes a coin a memecoin?
A memecoin is defined mostly by what it lacks: usually no business, no revenue and no legal claim on anything. Dogecoin, launched on December 6, 2013, began as a joke about crypto speculation and set the template.
The US Securities and Exchange Commission (SEC) describes a meme coin as an asset "inspired by internet memes, characters, current events, or trends" whose value is "driven by supply and demand." Its 2026 interpretation groups meme coins with digital collectibles, which it says are not themselves securities. That does not make fraud legal, and rules vary by country.
A joke coin is not automatically a scam. Many simply fade when attention moves on. The scam is when creators lie, hide their control or rig the market against buyers. Our tokens hub collects related guides.
How do launchpads like pump.fun work?
A launchpad lets anyone create a token and start trading it without first supplying their own liquidity, the pool of funds that lets people buy and sell. Pump.fun, on Solana, is the best known. Its program documentation describes this flow:
- Creation. A new coin starts with a fixed supply of 1 billion tokens.
- Bonding curve. A bonding curve is a formula that sets the price from how many tokens have been bought. Pump.fun's follows the same math as an automated market maker, using "virtual" reserves so that it can start with no real money in it.
- Completion. When a buy takes the last tokens for sale, trading on the curve stops.
- Migration. The coin moves to a pool on PumpSwap, the platform's own exchange, and the pool's liquidity provider (LP) tokens are burned.
The documented default settings put 793.1 million tokens on sale, with virtual reserves starting at 30 SOL (Solana's native coin) and 1.073 billion tokens. By our calculation from those numbers, filling the curve raises about 85 SOL, and the last buyer pays about 15 times what the first buyer paid. Settings can change, so treat that as an illustration.
The takeaway is that the earliest wallets get a large built-in price advantage. Nothing on the curve tells you whether one of those wallets belongs to the creator.
How often do memecoins fail?
Most memecoins end up close to worthless, but the widely quoted figures measure different things and cannot be compared directly.
| Study | What it measured | Result | Limit |
|---|---|---|---|
| Solidus Labs, 2025 | Over 7 million pump.fun tokens with at least five trades, created before April 2025 | 98.6% classed as rug pulls or pump-and-dumps; about 97,000 kept $1,000 of liquidity | The test is liquidity below $1,000, which also catches coins that simply died |
| Chainalysis, 2025 | 2,063,519 tokens launched in 2024 on any chain it tracked, flagged if one address pulled at least 65% of a pool's liquidity (worth $1,000 or more), the pool had over 100 transactions, and it then went dead for 30 days | 74,037 tokens (3.59%) flagged; 94% of the pools were emptied by the address that created them | Detects a pattern, not intent; Chainalysis calls it a starting point for investigation |
Pump.fun disputed the Solidus report. Its spokesperson argued that most memecoins, like most social media posts, have little lasting value, and that this is expected.
So neither number is the odds that a given coin is fraudulent. Together they say something simpler: a narrow, suspicious pattern is common, and survival is rare. Chainalysis also found that only 1.7% of 2024's tokens had been actively traded in the 30 days before its report.
Hard rugs, soft rugs and honeypots
A rug pull has no single official definition. A 2026 federal indictment describes it as a scheme where developers "abruptly or gradually withdraw from the project, sell or trade some or most of their tokens," and keep buyers' funds.
"Hard" and "soft" are used loosely, and sources disagree at the edges.
| Term | Usual meaning | What you see |
|---|---|---|
| Hard rug | The token or pool is built to take money, through hidden code or by draining liquidity | Sells fail, or the pool empties in one transaction |
| Soft rug | The team hypes the token, sells its holdings or walks away | Insider wallets sell down while the roadmap goes quiet |
| Honeypot | A hard rug variant: "a token you can buy but cannot sell" because the contract blocks or heavily taxes the sale | Buys succeed, but sells fail for most wallets |
Solidus's 2022 report called tokens programmed to steal hard rugs. 1inch also files liquidity theft under hard rugs, while Solidus's 2025 report calls abrupt liquidity withdrawals soft rugs. When you read a headline, ask what actually happened rather than trusting the label.
Two real cases
Squid Game token, 2021. SQUID borrowed the name of the Netflix show with no official connection. A so-called anti-dumping mechanism stopped most buyers from selling. On November 1, 2021, the price hit about $2,861, then fell to $0.0008 five minutes later. Euronews, citing blockchain data, put the amount cashed out by a wallet linked to the creators at almost €3 million.
Kishu Inu, charged but not proven. The indictment linked above, filed in Chicago on October 6, 2026, accuses a co-founder of marketing the Ethereum token as "fair-launched" with no team tokens. Prosecutors allege about 12% of the supply had gone to eight wallets controlled by two founders before the public sale, and that those two made more than $9 million selling. These are allegations, and a defendant is presumed innocent.
The Kishu allegations show why a promise about team holdings is worth little by itself. Where you can, test claims about who owns what against on-chain data.
Why market cap misleads: a worked example
Market cap multiplies the price by the supply, but only the pool's liquidity can actually pay sellers. Take a coin with a 1 billion supply and a pool holding 100 SOL and 100 million tokens. Assume the pool follows the constant-product formula and ignore fees.
| Before the sale | After an insider sells 200M tokens | |
|---|---|---|
| SOL in pool | 100 | 33.3 |
| Tokens in pool | 100M | 300M |
| Price per token | 0.000001 SOL | 0.00000011 SOL |
| Market cap | 1,000 SOL | 111 SOL |
The insider's stake looked like 200 SOL but sold for about 66.7 SOL. The price fell 89% in one trade. The other holders' 700 million tokens looked like 700 SOL before the sale, and now only 33.3 SOL is left in the pool to pay all of them.
Notice that nobody touched the liquidity here, so burned LP tokens would not have helped. Burned liquidity stops the pool being withdrawn. It does not stop holders selling into it.
A red-flag checklist before you buy
Run these checks in order. Most can be done free on a block explorer or token scanner. New coins usually trade only on a decentralized exchange, so no listing team has reviewed them for you.
| Check | What you want to see | What it does not prove |
|---|---|---|
| Mint authority | On Solana, set to None, which fixes the supply permanently. On Ethereum-style (EVM) chains, no owner-controlled mint function in the smart contract | Nothing about who already holds the supply |
| Freeze authority | Set to None, so nobody can make your token account unusable. Accounts frozen earlier stay frozen | That the token is sellable in other ways |
| Solana token extensions | No permanent delegate, which can move or burn anyone's tokens, and no unexplained pause, transfer fee or transfer hook | Some have legitimate uses, so read why they exist |
| Contract ownership (EVM) | Verified source code and no upgradeable proxy, a setup that lets someone swap the contract's logic. Renouncing ownership disables owner-only functions | Other admin roles can survive renouncing |
| LP tokens | LP tokens redeem a share of the pool, so you want most of them burned or locked for a stated period | A lock ends on a date, and holders can still sell |
| Holder concentration | No large wallets once you exclude the pool, curve, burn and exchange addresses | One person can split tokens across many wallets |
| Sell test | A scanner's simulated sale works, and many unrelated wallets have really sold | A clean check is not a guarantee: taxes and upgrades can change later |
| Liquidity depth | A pool large enough that your exit does not move the price much | Liquidity can leave if it is not burned or locked |
| Promotion | No guaranteed returns, countdowns or free-token links, which often overlap with airdrop scams | Honest-looking marketing proves nothing |
There is no official concentration limit. As a rough rule of our own, any single wallet above 5%, or a top ten above 30%, deserves an explanation.
One more distinction helps on launchpads. We checked four large coins launched through pump.fun (Fartcoin, GOAT, PNUT and CHILLGUY) directly on Solana, and all four had both authorities revoked. On coins like these the authority check will often pass, so the risk moves to who bought early and how much they still hold. A pass there is the start of your research, not the end.
If you buy anyway
Treat a memecoin as money you could lose completely. On-chain trades can't be reversed, and any loss falls on the buyer. Using a separate wallet that holds only what you can afford to lose limits the damage from wallet drainers. Size your trade against the pool, not the market cap. This is education rather than financial or legal advice. If you have lost a meaningful sum to suspected fraud, report it to your local authorities and consider asking a lawyer. Our security hub covers related risks, and Solana's hub has more on the chain where many of these coins launch.
Key takeaways
- A memecoin usually has no product or cash flow, so its price rests on attention and trading demand.
- Launchpads such as pump.fun make launches cheap and give the earliest wallets a large price advantage.
- Loss-rate figures such as 98.6% and 3.59% measure different things, and neither gives the odds for one coin.
- Check mint and freeze authority, LP status, holder concentration and whether real wallets can sell.
- Burned liquidity stops pool withdrawals, but insiders can still sell into the pool and crash the price.
- Passing every check does not make a coin safe. It only removes some of the known ways to lose everything.
Sources
- Solidus Labs: Solana rug pulls and pump-and-dumps report
- Chainalysis: Crypto market manipulation 2025
- Pump.fun program documentation
- Solana Program Library: Token documentation
- Kishu Inu indictment, US District Court, N.D. Illinois
- SEC and CFTC: Application of the federal securities laws to certain crypto assets
- Euronews: Squid Game token collapse
Get the weekly commit
New blockchain deep dives every week.
