Graduation is the finish line a pump.fun coin crosses when it fills its bonding curve. The coin leaves the curve and starts trading on PumpSwap, an open exchange. The word sounds like an achievement.
We split our 93 pick record by it. The picks entered after graduation lost deeper.
The ledger holds 93 scored picks from the board. 52 of them were entered with the coin already on PumpSwap. 41 were entered while the coin still had curve left to fill.
The median graduated pick sits at 0.17x from entry.
The median curve pick sits at 0.34x.
The whole record, as we printed it two days ago, sits at 0.24x. The graduated half is what pulls it down.
Deeper than the medians, the shape of the losses:
| Multiple from entry | Graduated (52 picks) | Still on curve (41 picks) |
|---|---|---|
| Under 0.2x | 29 (56%) | 15 (37%) |
| 0.2x to 0.5x | 6 (12%) | 15 (37%) |
| 0.5x to 1x | 8 (15%) | 7 (17%) |
| 1x to 2x | 5 (10%) | 3 (7%) |
| 2x or better | 4 (8%) | 1 (2%) |
Compare the two loss bands. In the curve group, 15 picks sit under 0.2x and another 15 sit between 0.2x and 0.5x. Even split. In the graduated group, 29 sit under 0.2x and only 6 between 0.2x and 0.5x. The losses pile at the bottom.
Under 0.2x means down more than 80 percent from the price the board printed at entry. And that is where those coins stand today, not where they dipped on the way.
The board paid more for those tickets, too. The median graduated entry was $55,792 of market cap. The median curve entry was $15,704. 33 of the 52 graduated picks were entered above $40,000 of cap. Higher entry, deeper median loss.
And here is the part that keeps this from being a simple story. Six of the eight picks standing at 1.3x or better are graduated coins. AgenC sits at 3.19x from a $400,325 entry. Leverhedge sits at 1.37x from $321,896. The biggest multiple in the whole record, 4.71x, belongs to a curve coin, Goyim, picked at 72.7 percent of curve.
Both things are true. The graduated group holds the floor and most of the winners. The curve group bunches in the middle. Graduation does not protect a coin and it does not ruin it. It widens the spread.
What changes at the crossing is mechanical, not moral. Before it, every trade hits the bonding curve, a fixed price ladder with the coin’s pool behind it. After it, the coin trades against an open pool that anyone can drain by selling. The checker prints exit cost for any coin, what a $500 sell actually takes out of the pool. The word graduated tells you where a coin has been. Exit cost tells you what leaving it costs right now.
For a reader the finding is narrower than the drama. A graduation badge is history. It says a coin once had enough demand to fill a curve. It does not say the demand stayed. In this record the difference between the two groups was not the badge. It was the entry price. The picks that entered near $56,000 of cap after the crossing bought the most expensive tickets in the ledger, and they took the deepest median loss of any group we have split so far.
Method. The sample is the 93 scored picks in the board’s grading ledger. Each pick’s entry market cap was recorded when the pick was scored. The standing market cap is the live tracker read captured in this run, and the multiple is standing cap divided by entry cap. A pick counts as graduated when the coin carried no bonding curve percentage at entry, meaning it already traded on PumpSwap. The 41 curve picks carried a median 78.2 percent of curve at entry, max 96.8. Limits: this is a standing record, not a trade record. Nothing was bought or sold. A coin that ran 3x and fell back counts at its fallen price, so the winners named above are coins standing there today, not coins that peaked there. Both samples are small. A handful of coins moving between groups would shift these percentages by five points or more. It is a description of one board’s record, nothing else.
