Blog · 2026-11-19
What 2,400 new BNB Chain tokens look like in their first minutes: honeypots are rare, rug levers are not
For a little over a day (17–18 November 2026) we picked up every new PancakeSwap V2 pair against WBNB on BNB Chain within about a minute of its creation and ran our full scan on each token: a simulated buy, transfer and sell against live pool state, a bytecode check for owner powers, and a liquidity and LP-lock check.
2,377 new pairs. Here is what they looked like.
60% never got liquidity
Of the 1,801 pairs we had finished scanning, 1,088 (60%) had no liquidity when we scanned them, even after two rescans (at about 30 minutes and 2 hours). Creating a pair costs only gas; many are test deployments, abandoned launches or pairs created ahead of a launch that never happened.
Everything below is about the 713 tokens that did have liquidity.
Classic honeypots: about 1 in 23
- 31 honeypots (4.3%). The simulated buy worked and the sell failed.
- 9 more (1.3%) couldn't even be bought: trading switched off, or the token blocks contract buyers.
- Taxes were rarely the trap. Only 13 tokens (1.8%) charged any sell tax above 0.5%, and just one charged 10% or more. The days of the 25% "reflection" tax seem to be over on new launches.
So a token that passes the sell simulation is very likely sellable right now. That's the good news.
The levers that remain
- Unlocked liquidity: 72%. 512 of 713 tokens had less than half of their LP tokens burned or sitting in a known locker (PinkLock, UNCX, Team.Finance, Mudra). Whoever holds that LP can remove the liquidity in one transaction.
- Tiny pools: 62% had under $5,000 of liquidity. Only 47 tokens (6.6%) had $50,000 or more. Average: about $10,600. In a $3,000 pool, a single $500 sell moves the price by double digits.
- Owner powers were uncommon: 5 tokens could blacklist wallets, 5 could change fees, 4 had a trading switch, 4 were upgradeable proxies, 2 had an owner mint function. Each one is a lever that can turn a sellable token into a trap later.
A day later: almost all of them were gone
On 19 November we rescanned a random sample of 80 of these tokens that had launched with at least $1,000 of liquidity, 12 to 36 hours after launch:
| Group | Tokens | Pool value down more than 90% |
|---|---|---|
| LP not burned or locked | 51 | 51 (100%) |
| LP burned or locked | 29 | 25 (86%) |
Locked liquidity stops one specific exit, the pull. It doesn't stop holders and insiders from selling, which drains the pool just as surely. Only one token in the sample had grown its pool by more than 20%.
So the honeypot check answers "can I sell this?", not "will this still be worth anything tomorrow?". For new launches the second answer, statistically, is almost always no.
What this means before you buy
- Run a sell simulation. It catches the 4–6% that are honeypots or untradeable today.
- Then look at the LP, not just the contract. In this sample the most common way to lose money wasn't contract trickery but a pool that can be emptied, or that is too small to exit.
- Recheck while you hold. Fees, blacklists and liquidity can change after you buy.
How we got these numbers
Every token above has a public report at novaquill.dev/token/bsc/<address>; the live list of new launches is at /launches. Counts come from our scan database for pairs detected between 21:15 UTC on 17 November and 23:00 UTC on 18 November 2026. Some flags were measured minutes after launch, when many tokens were still being set up.
Check any token yourself at novaquill.dev, in Telegram with @NovaquillScanBot (send /watch <address> to get a message if its liquidity is pulled or its taxes change), or from code with the API.