Methodology

1. The daily scan

Once every 24 hours, at midnight New York time, an automated scanner runs server-side. It works in three stages:

• Universe. It pulls the list of every cryptocurrency above $50 million market capitalisation from public market data. We use $50M as a floor because below it, concentration is so common and liquidity so thin that the signal becomes noise.
• Exclusions. It removes entire categories of asset that are concentrated by design and would otherwise produce false positives: stablecoins, tokenized real-world assets and treasuries, tokenized gold and equities, liquid-staking tokens, wrapped tokens, and centralized-exchange tokens. These are supposed to have large custodial holders; flagging them would be meaningless.
• Holder analysis. For every remaining coin that exists as a token on one of seven supported chains (Ethereum, BNB Smart Chain, Base, Arbitrum, Polygon, Optimism, Avalanche), it fetches the live on-chain holder list and computes the concentration metric described below.

2. The top-10 concentration metric

This is our signature measurement, and the one most worth understanding. For each token we take its holder list and remove the wallets that are not genuine private holders:

• Centralized exchange wallets (Binance, Coinbase, OKX, Kraken, Bybit and dozens more), identified by on-chain labels.
• Staking, vesting, treasury, reserve and foundation contracts.
• Bridges and liquidity pools (Uniswap, PancakeSwap, Curve and similar).
• Burn and null addresses.

From what remains, we sum the share of total supply held by the top ten wallets. Crucially, we keep unlabelled smart contracts in this count, because a team or insider’s multisig is usually a contract — excluding all contracts would hide the very concentration we are trying to surface. If that top-ten figure exceeds 50%, the coin is flagged for supply concentration.

Why 50%? Because once a small, coordinated group controls more than half the float, they can move the price more or less at will. The classic abuse is to hold the price artificially high while perpetual-futures funding stays negative — bleeding leveraged short sellers day after day — and then dump the entire position in minutes, collapsing the price 90% or more with no buyers left to catch it. You can read the full mechanics in our guide on the funding-drain scam.

Data hygiene. If a token returns broken supply figures (individual holders summing to more than 100%, which happens with some wrapped or migrated tokens), we discard the result rather than publish a nonsensical number.

3. The curated allow-list

Concentration alone cannot distinguish a scam from a legitimate but young, low-float project — a major layer-2 token a few months after launch can show the same top-ten percentage as an outright manipulation coin. To avoid falsely labelling established projects, we maintain a manually-reviewed allow-list of well-known protocols whose concentration reflects DAO, foundation and vesting contracts rather than manipulation. Coins on that list still have their concentration percentage displayed as a factual data point, but they are not given the scam flag. The list is reviewed regularly.

4. The composite risk score (0–100)

The headline risk score on each coin page blends seven independently-weighted factors. Each is scored from public data and contributes to the total, which is capped at 100:

• Price collapse — how far the coin has fallen from its all-time high. A 95%+ drawdown is a hallmark of an exit scam or abandoned project.
• Volume anomaly — the ratio of 24-hour volume to market cap. Volume far above market cap signals wash trading; near-zero volume signals an illiquid trap.
• Market-cap size — micro-caps can be moved double-digit percentages with a few thousand dollars.
• Supply inflation — a low circulating-to-total ratio means large future unlocks will dilute holders.
• On-chain holder concentration — the verified top-10 metric above, which carries the heaviest weight when a coin is flagged.
• Price volatility — extreme 24-hour swings consistent with coordinated pump-and-dumps.
• Transparency gaps — missing website, whitepaper, or public code repository.

A coin verified to have more than half its supply in ten private wallets is floored to a high-risk score regardless of how calm its price looks, because that single fact is enough to make it dangerous.

5. Live market data

Prices, market caps and 24-hour changes shown on the site are refreshed live in your browser from public market data, so they stay current between daily scans. The holder concentration figure updates once per day with the scan.

6. Limitations — read this

We are deliberately upfront about what this method cannot do:

• Concentration is a signal, not proof. A young legitimate project and a scam can look identical on this metric alone; that is exactly why the allow-list and the composite score exist.
• We only see EVM-compatible tokens on seven chains. Native assets (like Bitcoin) and tokens on unsupported chains are not holder-scanned.
• On-chain labels are imperfect. An unlabelled exchange wallet could be counted as a private holder, or vice versa.
• Market data can be delayed or wrong at the source, and our analysis can contain errors.
• Absence of a flag is not a safety guarantee. A coin can be uncatalogued, brand new, or dangerous for reasons our model does not measure.

7. The bottom line

Everything on FraudCoins.com is an automated, opinion-based indicatorderived from public data — a starting point for your own due diligence, not a verdict and not financial advice. If you believe a specific flag relies on inaccurate data, contact legal@fraudcoins.com. See also our full disclaimer.