We Flagged a Token at 79% Insider Supply. It Fell 69% in a Day
By the MemeAssist Research Desk · Published 2026-08-02 · Updated 2026-08-02 · 5 min read
On July 29, 2026 our screening engine rejected the token 'dog' because one wallet held 79% of the supply and the top 10 held 86% (excluding liquidity pools). Within 24 hours of that first logged flag the token had fallen 69.3%. Extreme holder concentration is the most reliable predictor of collapse we measure.
Key Learnings
- First rejected at 04:44 UTC on July 29, 2026: top 10 wallets held 86% of supply, a single wallet held 79% — logged automatically at a ~$500k market cap.
- 24 hours after the first flag the token was down 69.3%.
- The engine re-encountered and re-rejected the token six more times that day (high bundler count) as it bled from ~$950k to ~$650k market cap — the flags never cleared.
The on-chain receipt: 79% insider supply flagged before the drop
At 04:44 UTC on July 29, 2026, our screening engine evaluated a token called dog (mint 9gAV…pump) and rejected it. The logged reasons, verbatim from the database:
- "top 10 wallets hold 86% of supply (excl. pools)"
- "single wallet holds 79% of supply"
At the time it had about $64k of liquidity and a ~$500k market cap, with active volume and a live community. Twenty-four hours later it was down 69.3%.
Through the rest of that day the engine crossed paths with the token six more times — and rejected it every time, now also flagging a high bundler count — as the market cap slid from ~$950k at the local peak to ~$650k and kept bleeding. There was never a moment where the holder picture improved.
Why 79% in one wallet is game over
Holder concentration is arithmetic, not opinion. If one wallet controls 79% of the supply, then that wallet — not the market — sets the price. Every green candle is happening with the owner's permission. The float that outsiders trade among themselves is a rounding error; the moment the dominant wallet sells, there is no bid deep enough to absorb it.
Our data across hundreds of tracked rejections backs the arithmetic: concentrated-supply flags are among the most frequent reasons our engine rejects tokens, and the flagged cohort's forward returns are consistently negative.
A wallet holding 79% of supply isn't a red flag. It's the whole flag factory — the only question left is when, not if.
"But the pools hold a lot too, right?"
A common trap when checking holders manually: the top of the holder list often includes the liquidity pool itself, which makes concentration look scarier — or, if you mentally discount the top entries as "probably pools," less scary than it is. Our engine excludes pool and exchange accounts before computing concentration, so "79% in one wallet" means a real, sellable, private position. When you check by hand, always subtract the pools first — here's the full method.
The holder-concentration pattern to internalize
- Concentration is a ceiling on your outcome. You can be right about the narrative, the timing, and the momentum — and still lose 69% overnight because one wallet chose that night.
- Flags don't expire on their own. Six re-checks across one day never cleared this token. A bad holder structure at launch almost never heals; supply doesn't voluntarily de-concentrate.
- The check takes five seconds. Paste the mint into the free MemeAssist analyzer and the holder breakdown — pools excluded, clusters detected — is the first thing you see.
Verify the insider concentration yourself on-chain
The mint address is 9gAVAtdnsrniW3GCwsvDeRPf5eJDjtT9S5bu1j7tpump. The holder history and the July 29–30, 2026 price collapse are publicly visible on any Solana explorer.
Frequently asked questions
What percentage of holder concentration is dangerous on Solana?
We treat top-10 concentration above ~65% of supply (excluding pools) as a hard rejection, and a single non-pool wallet above ~30% as disqualifying on its own. 'dog' was far past both lines: 86% top-10 and 79% in one wallet.
Do liquidity pools count toward holder concentration?
They shouldn't — pool accounts hold tokens structurally, not as a sellable insider position. MemeAssist excludes pools and exchange accounts before computing concentration; manual checks that skip this step get misleading numbers in both directions.
Can a highly concentrated token still pump?
Yes, and that's the trap — the owner benefits from a pump before selling into it. 'dog' nearly doubled its market cap intraday while flagged. Concentration doesn't prevent pumps; it decides who gets paid at the end of one.
Where can I see how often these flags are right?
Every safety-gate rejection is tracked for a full 24 hours after the call, and outcomes are averaged with no cherry-picking — the engine is judged on everything it flags, not hand-picked wins.