HostOriginal data study · computed 2026-11-08 · reproducible from the open dataset · free to cite with attribution
State of token safety — 2026
NexFlow's flagship annual read of the on-chain token-safety landscape. We grade every token that clears our data gate — at least $25,000 of pool liquidity or 250 holders — on the structural signals that decide whether a holder can exit and whether the contract can be turned against them: mint and freeze authority, holder concentration, liquidity depth, and creator-wallet stake. The result is a cross-sectional snapshot of 2,970 tokens across 19 blockchains carrying $16.7 billion in tracked liquidity.
Key findings
- 60.4% of gated tokens grade A (structurally clean); 9.0% grade C, D or E — carrying at least one structural red flag.
- 60.5% of tokens have both mint and freeze authority confirmed renounced; 9.2% still carry at least one live dangerous authority.
- The median token's top 10 token accounts hold 82.8% of supply — and that includes liquidity pools, lockers and exchange wallets.
- 66% of all tracked DEX liquidity sits in just 10 tokens.
- 37 tokens collapsed below the data gate this period, draining $0.98 billion in peak liquidity.
What we measured
The universe is every token that clears the gate — deep enough or held widely enough that a safety read is meaningful. Each token is graded on a deterministic 100-point scale from on-chain facts: whether the contract can still mint new supply or freeze balances, how deep the pool is relative to the token's size, how concentrated the holder base is, how much the deployer wallet still holds, and (on EVM) the wider contract-control surface a token-security scanner reports. Grades are structural — they describe the contract and the market, never the price.
Grade distribution
| Grade | Tokens | Share | What it means |
|---|---|---|---|
| A | 1,794 | 60.4% | Both dangerous authorities renounced, workable depth, no structural red flags |
| B | 908 | 30.6% | Minor concerns — modest depth, small holder base, or an unmeasured contract field |
| C | 169 | 5.7% | At least one live authority or a material structural concern |
| D | 51 | 1.7% | Multiple structural red flags |
| E | 48 | 1.6% | Severe structural risk — live mint and/or freeze authority, thin exit |
The grading function is public and deterministic — the same inputs always produce the same grade, and the per-token reasons are shown on every token page. The full scoring method is documented in how to read a token risk scan.
Authority posture — who can still turn the contract against you
The two authorities that matter most are mint (can inflate supply and dilute every holder) and freeze (can lock balances so they cannot move). Across the gated set:
| Posture | Tokens | Share |
|---|---|---|
| Both mint and freeze renounced | 1,796 | 60.5% |
| At least one dangerous authority still live | 273 | 9.2% |
| Mint state unresolved in this snapshot | 899 | 30.3% |
Live-authority cases split by chain because the two ecosystems are measured differently. On Solana (2,046 tokens), mint authority is read from the mint account through the market data feed: 1,533 tokens (74.9%) are confirmed renounced and the remainder are unmeasured in this snapshot; no live-authority case resolved. On EVM chains (924 tokens), a token-security scanner reports the contract controls directly: 253 tokens (27.4%) carry a live mint authority, 52 (5.6%) a live freeze/blacklist, 263 (28.5%) have both renounced, and 388 (42.0%) went unmeasured. The EVM read also surfaces controls Solana does not have — upgradeable proxies, hidden owners, self-destruct, transfer taxes and whitelists — each folded into the grade.
Holder concentration — how much of supply a few accounts control
For 897 tokens we resolved holder concentration directly from chain data — the share of total supply held by the ten largest token accounts. The distribution is heavily skewed:
On Solana alone (199 resolved tokens) the median top-10 concentration is 77.8%. Read this with its caveat: the ten largest accounts include liquidity pools, lockers and exchange wallets, so this is concentration of token-account balances, not proof that one person holds them. Even so, the shape is the finding — token ownership is concentrated at the account level far more than the holder counts suggest, and a handful of accounts can decide whether a market functions.
Liquidity concentration
Liquidity is even more concentrated than holders. Of the $16.7 billion in tracked liquidity, the top 10 tokens hold 66.0%, the top 50 hold 86.6%, and the top 100 hold 92.5%. This is partly structural — major assets and wrapped tokens dominate DEX depth — but it means the "tracked universe" is, in liquidity terms, a small set of deep markets plus a long tail of thin ones. The long tail is where the safety grades do the most work.
Creator-wallet stake
The deployer wallet's share of supply is a single-wallet concentration read, resolved for 1,523 tokens. 3.9% of resolved tokens have a creator wallet holding 30% or more of supply, 6.5% hold 10% or more, and 9.2% hold 5% or more. A creator sitting on a third of supply can move the price or sell into holders at will — the grade treats that as a structural red flag, not a price prediction.
Collapses — the exit doors that closed
37 tokens that once cleared the data gate have since collapsed below it, draining $0.98 billion in combined peak liquidity. Each collapse is recorded with dated figures in the obituaries ledger. A collapse is an outcome, not an accusation — teams migrate, projects fail, and liquidity simply leaves — but the on-chain record is unambiguous about one thing: holders who stayed can no longer exit at meaningful size. The ledger notes the dates and the peak, and withdraws a notice automatically if a pool refills past the gate.
By chain
| Chain | Tokens | Liquidity | Grade A | Grade B | C / D / E |
|---|---|---|---|---|---|
| Ethereum | 129 | $6.23B | 19.4% | 61.2% | 19.4% |
| BNB Chain | 141 | $3.19B | 33.3% | 47.5% | 19.1% |
| Base | 110 | $2.53B | 30.9% | 51.8% | 17.3% |
| Solana | 2,046 | $2.24B | 77.1% | 20.5% | 2.3% |
| Polygon | 79 | $1.95B | 22.8% | 57.0% | 20.3% |
| Arbitrum | 68 | $0.25B | 13.2% | 66.2% | 20.6% |
| Monad | 64 | $0.04B | 34.4% | 56.2% | 9.4% |
| Avalanche | 58 | $0.05B | 20.7% | 48.3% | 31.0% |
| Optimism | 59 | $0.05B | 13.6% | 35.6% | 50.8% |
| Berachain | 25 | $0.03B | 40.0% | 28.0% | 32.0% |
| Gnosis | 26 | $0.02B | 11.5% | 61.5% | 26.9% |
| Blast | 24 | $0.01B | 37.5% | 33.3% | 29.2% |
| Scroll | 11 | <$0.01B | 0.0% | 63.6% | 36.4% |
| Linea | 16 | <$0.01B | 25.0% | 50.0% | 25.0% |
| zkSync | 9 | <$0.01B | 0.0% | 88.9% | 11.1% |
| Cronos | 35 | $0.08B | 8.6% | 48.6% | 42.9% |
| Sonic | 27 | <$0.01B | 18.5% | 48.1% | 33.3% |
| Unichain | 20 | <$0.01B | 20.0% | 65.0% | 15.0% |
| Mantle | 23 | <$0.01B | 13.0% | 56.5% | 30.4% |
Solana dominates the count (69% of the gated set) but not the liquidity — Ethereum, BNB Chain and Base carry most of the tracked depth, and they also carry the lowest grade-A shares, because major centralized assets (wrapped tokens and fiat-pegged stablecoins) intentionally retain mint and freeze authority and grade accordingly. The grade describes the contract, not the asset's legitimacy.
Method & reproduction
Every figure above is computed from NexFlow's committed weekly on-chain snapshots, published as an open dataset under CC BY 4.0. Solana authority and supply reads come from chain RPC; EVM contract controls come from a token-security scanner; liquidity and holder counts come from market and chain data feeds. The grading function is deterministic and published. Snapshots date the moment — the chain moves on — so any individual token should be re-read live with the free scanner before acting. Figures are dated 2026-11-08 and will be refreshed as the dataset updates.