What did Belphor's four signal families do this week?
In the seven days to 2026-09-20 Belphor recorded 1711 signals across its four families: no pump.fun community takeover, 304 momentum ignitions scoring 65 or above, no smart-money confluence signal and 1407 pump.fun migrations. A further 12 rows arrived under pump.fun's new takeover instruction (admin_cto and rare_admin, recorded since 12 September); they are not promoted to alerts and are not counted in the takeover figures above. 1609 of the 1711 have a price reading at least six hours after detection; every rate below is computed over those. Of the ignitions with a later reading, 38% reached +15% and 12% reached +50%, while 12% were at a fifth of their detection price or lower. 91% of the migrations with a later reading were at a fifth of their migration price or lower. Updated 2026-09-20.
Cite as: Belphor Research (2026). What did Belphor's four signal families do this week? Belphor Data, https://belphor.com/data/this-week. Updated 2026-09-20. Inputs d11033562988, formula 1.0, covered 1609, missing 102. CC BY 4.0.
The four families in the last 7 days
| Family | Signals | With a later reading | Reached +15% | Reached +50% | Went to zero | Median peak (x) |
|---|---|---|---|---|---|---|
| Takeovers | none this week | |||||
| Ignitions | 304 | 285 | 38% | 12% | 12% | 1.08x |
| Smart money | none this week | |||||
| Migrations | 1407 | 1324 | 27% | 16% | 91% | 1.02x |
Takeovers are pump.fun creator-change instructions of type cto with a price at the alert, demo rows excluded. Ignitions are momentum signals scoring 65 or above, the gate the ignition-survival page uses. Smart money is the Terminal's confluence rule, two or more tracked wallets on one coin. Migrations are pump.fun graduations with a price at the migration. Every rate is computed over the signals with a price reading at least six hours after detection, so it is a lower bound for the week as a whole.
Recorded but not promoted: pump.fun's new takeover instruction
| Instruction | Rows this week |
|---|---|
| admin_cto and rare_admin | 12 |
pump.fun changed its takeover instruction on 12 September 2026. Belphor records the new one (admin_cto, and the earlier rare_admin) but does not alert on it and does not count it as a takeover, because its outcomes have not been measured yet. The count is here so the change is visible rather than silent.
The week's biggest moves
| Coin | Family | Detected (UTC) | Score | Market cap at detection | Peak (x) | Now (x) | Outcome |
|---|---|---|---|---|---|---|---|
| PUMP pump | Ignition | 19 Sep 07:56 | 68 | $1934.9M | 4719.42x | 0.97x | reached +50% |
| PUMP pump | Ignition | 18 Sep 13:42 | 74 | $2002.5M | 4504.25x | 4504.25x | reached +50% |
| CRUSDATA | Migration | 18 Sep 17:02 | 88 | $59.2k | 78.25x | 0.00x | reached +50% |
| JEANPHIL | Migration | 19 Sep 19:12 | 78 | $64.4k | 74.06x | 67.89x | reached +50% |
| KIBA | Migration | 15 Sep 12:30 | 88 | $2.7M | 44.55x | 0.04x | reached +50% |
| USDF | Migration | 14 Sep 04:52 | 78 | $46.7k | 15.76x | 15.76x | reached +50% |
| MIZO | Migration | 13 Sep 08:54 | 78 | $46.5k | 14.22x | 4.55x | reached +50% |
| WOFI | Migration | 16 Sep 20:44 | 88 | $56.6M | 13.58x | 0.00x | reached +50% |
| BRAIN | Migration | 14 Sep 21:54 | 78 | $45.4k | 11.69x | 5.55x | reached +50% |
| PONYX | Migration | 16 Sep 10:28 | 88 | $2.8M | 11.25x | 11.25x | reached +50% |
| Paid | Ignition | 18 Sep 08:08 | 67 | $21.6k | 9.41x | 0.00x | reached +50% |
| Pigeon | Migration | 19 Sep 02:06 | 78 | $46.8k | 6.75x | 2.20x | reached +50% |
The twelve largest peaks of the week across all four families, among signals with a later reading. These are the extremes, not a typical result: the rates in the first table cover every signal of the week. Where two coins in this list share a ticker they are different mints, so each carries the first four characters of its own mint address. The Now column is what the same coin was worth at the last reading.
How this is measured
The window is the seven days ending at the moment this page was recomputed, and every figure is measured from that same moment. Demo and test rows are excluded everywhere.
A signal is "covered" when Belphor has a price reading for it at least six hours after detection. Rates are computed over covered signals only, and the covered count is printed beside every one of them, so a week whose newest signals have not been read yet reports a smaller base rather than a guess. "Reached +X%" uses the highest price observed after detection; "went to zero" means the most recent reading was 20% of the detection price or lower.
This page carries no claim the counts do not support. A family with nothing in the week prints "none this week" rather than a zero rate. Prices come from Belphor's own snapshot record (DexScreener pair data). Recomputed daily from the raw record, with the source row ids and the values read frozen into receipts_inputs and hashed into the Cite as line above, so any week can be reproduced after the fact.
Frequently asked
How many Solana memecoin signals did Belphor detect this week?
1711 in the seven days to 2026-09-20: 0 community takeovers, 304 ignitions scoring 65 or above, 0 smart-money confluence signals and 1407 pump.fun migrations. 1609 of them have a price reading at least six hours later.
Were there any pump.fun CTOs this week?
No. Belphor detected no takeover of type cto in the seven days to 2026-09-20. 12 rows did arrive under pump.fun's new instruction, which Belphor records but does not yet alert on. The longer-run takeover record is on the CTO outcomes page.
Where does this data come from?
From Belphor's own on-chain detection of pump.fun instructions and its pool scanner, plus its minute-level price snapshots, recomputed daily from the raw record. The JSON and CSV files on this page contain the same numbers.
Belphor detects pump.fun takeovers on-chain in real time, scores every signal, trades member strategies on paper or live, and publishes every outcome. No promises, just receipts.
Explore Belphor →