Pump.fun parent programs: migrated vs not migrated
This study compares Pump.fun launches from the latest 24-hour data window.
For each token it counts distinct buys by parent_program during slots
launch_slot + 1 through launch_slot + 128, keeps tokens with at least 48
buys, and labels it migrated when the mint appears in pfamm_migrations by
the cohort cutoff. Launches without a complete 128-slot horizon are excluded.
Window end: 2026-08-23T11:20:03Z · Eligible tokens: 2,992
Cohorts
| Cohort | Tokens | Share of eligible launches | Mean buys in first 128 slots |
|---|---|---|---|
| Migrated | 259 | 8.66% | 141.5 |
| Not migrated | 2,733 | 91.34% | 116.4 |
Parent-program composition
Every token is normalized to a distribution before cohort averaging, so a high-volume token cannot dominate the result. The difference is migrated minus not migrated, in percentage points.
| Parent program | Migrated | Not migrated | Difference |
|---|---|---|---|
FLASHX8DrLbgeR8FcfNV1F5krxYcYMUdBkrP1EPBtxB9 | 35.880% | 37.188% | -1.308 pp |
<direct> | 25.854% | 30.782% | -4.928 pp |
MAyhSmzXzV1pTf7LsNkrNwkWKTo4ougAJ1PPg47MD4e | 17.995% | 9.370% | +8.626 pp |
6Vo3245eszAb5wuqEMw8mGdbfRUdKbHhDHP5LcaGuTAB | 4.984% | 3.340% | +1.643 pp |
term9YPb9mzAsABaqN71A4xdbxHmpBNZavpBiQKZzN3 | 3.991% | 4.050% | -0.059 pp |
GMgnVFR8Jb39LoXsEVzb3DvBy3ywCmdmJquHUy1Lrkqb | 4.035% | 4.021% | +0.014 pp |
9ddjzqYhSTMHaBrrKukRXRfy4WzHUPjdX88uPXZ7MXyn | 1.319% | 0.255% | +1.064 pp |
7bopoA2tnvDYEfp54FhgNDiLR9wSPcEs9J7FaUGURsXj | 0.912% | 0.293% | +0.619 pp |
J7pourVwqP1VtRNBiFNqbdGxwkx7ta2LHAURURRRJqmd | 0.676% | 0.576% | +0.099 pp |
JUP6LkbZbjS1jKKwapdHNy74zcZ3tLUZoi5QNyVTaV4 | 0.589% | 0.455% | +0.135 pp |
proVF4pMXVaYqmy4NjniPh4pqKNfMmsihgd4wdkCX3u | 0.061% | 0.556% | -0.496 pp |
FAdo9NCw1ssek6Z6yeWzWjhLVsr8uiCwcWNUnKgzTnHe | 0.313% | 0.524% | -0.211 pp |
| Other (175 programs) | 3.391% | 8.590% | -5.198 pp |
View the profile SQL on GitHub and the cohort calculation.
Migration is observed only through the displayed cutoff. Tokens launched near that cutoff have less time to migrate, so the not-migrated cohort is right-censored. This descriptive comparison does not establish causality or predict future migration.