Meme Coin Market Intelligence — 9.28–10.4, 2026

The weekly board: $23.91B → $23.66B (-1.1%), 4 up / 4 down, dispersion 15.3 points — all eight members still above the August 16 base, the quarter's leadership map unchanged, and the September monthly report published alongside this issue.

1. The eight closes

DOGE remains 69.7% of the board's combined cap, so the week's arithmetic is simple: the board fell 1.1% mostly because PENGU unwound 10.7% and DOGE drifted 1.2%, while the four small names (BONK, WIF, FLOKI, PEPE) absorbed the ETF pause and stayed green. A cap-weighted board that falls less than its worst member and less than half its equal-weighted downside is a board whose base is holding.

2. What the board's concentration means for reading the pause

The index is more concentrated than its eight rows suggest: DOGE is 69.7% of the combined cap, so the board's weekly print is set by the anchor before the other seven register at all. That single weight is why the board's -1.1% and DOGE's own -1.2% are the same number to the first decimal - the index is, functionally, a DOGE position with a small-cap tail, and its headline is best read as one.

Concentration has shaped the quarter in both directions. On the way up it made breadth look stronger than the average member earned: last week's +20.8% was carried by the same name that holds the weight, and the index outran its own equal-weighted mean. On the way down it makes a quiet anchor read as a quiet sector - a board whose weight sits in one coin moves as one whenever that coin is flat, and a pause is exactly what that looks like from the top. Neither tendency is hidden in the tables; both fall straight out of the weight column.

The practical consequence is sequencing. This board resumes or rolls on DOGE's levels before anything else does: the $0.0927 shelf, held on every test since September 23, and the $0.1004 ninety-day-high close are the two prints that settle direction, and the small names' relative strength is the confirmation rather than the driver. It is also why the rotation map below can show leadership changing while the index barely moves - the margins rotate, and the anchor holds the level.

3. Rotation map — everyone still above the August 16 base

The structural fact of the quarter is unchanged: every member sits above the mid-August base of our close series, and has since the September 6 week except for the September 7–13 washout dip. What changed this week is the ordering within it — WIF overtook PEPE at the top of the net-from-base table, BONK closed to within two points of PEPE, and PENGU fell from third to fourth. Leadership by net gain has now changed in each of the last three weeks, which is rotation behavior, not trend behavior — the melt-up moved the whole base, this week moved the margins.

4. Flow quality

The week's flow picture, in three numbers. One: +$241.1M of spot Bitcoin ETF net inflow for the Sep 28 – Oct 2 trading week (Farside, via YFarmX) — a tenth of the prior week's $2.39B, with the nine-session streak broken by September 30's -$148.7M. Two: -$26.15M across the all-crypto ETF complex for the week (CoinGlass, via The Crypto Times, October 3) — the first weekly aggregate slip of the run. Three: +$292.6M recovered across October 1–2, meaning the heal was faster than the break. Volume told the same story at half-throttle: Bitcoin's Sunday session ran $15.0B against a $30.5B ten-day average (CryptoOceans, 23:55 UTC).

5. Reading the flow pause against the quarter's base

The throttle map in this column frames the week as the third in a three-week flow arc. Read against the August 16 base, the pause is even quieter: the board sits 17.7% above the August 30 monthly close and 40.8% above the mid-August base, so a -1.1% week is a rounding error on the quarter, not a reversal of it. The flow that built the melt-up did not leave - it idled, and the board held 94% of the final week's gain. The heal was the tell: +$292.6M recovered on October 1-2 against the -$148.7M that broke the streak means the break was a one-session event, not a trend.

What would make the pause a fade is specific and already on the watchlist. A second negative ETF week with six or more members red, DOGE losing the $0.0927 shelf, or dispersion re-widening above 25 points would each convert 'pause' to 'fade'. None printed this week. The honest line is the one the intelligence column draws: one week cannot say, and the data says pause more than fade because breadth held, outflows healed fast, and Bitcoin bid into the Sunday close. The base case from the structure is consolidation above $84,000 Bitcoin with flow staying positive - which would confirm the arc rather than extend it.

6. The throttle map: flow to board, three weeks running

(Flows: Farside, via YFarmX; the September 14-20 figure is the weekly print the week-5 report carried. Closes: our Sunday-close series.) Three weeks is a small sample, and we label it as such - but the mapping is unusually clean. A near-zero flow week produced a +5.1% squeeze that the next week's data proved hollow; a record-flow week produced the quarter's best close; and a tenth-of-throttle week produced a board that kept 94% of the gain. The interim conclusion we are willing to print: on this board, in this regime, flow level sets the direction and dispersion sets the warning. The all-crypto ETF aggregate slipping -$26.15M for the week (CoinGlass, via The Crypto Times, October 3) is the one soft datum - the complex idled slightly harder than the Bitcoin funds alone, which is what late-cycle consolidation looks like from the demand side. If the throttle returns to even half of September's pace, the board's four green small names are positioned to lead the next leg; if it dies for a second week, the $22.19B September 6 close - the arc's last support above the washout zone - becomes the level that matters, and this table will have been the early warning.

7. The month this week closed out

September's report is live in this column as its own page — Meme Coin Monthly Report — September 2026. Headline: +18.4% cap-weighted ($20.11B → $23.80B), +19.0% equal-weighted, seven of eight green, dispersion 53.2 points between BONK's +36.7% and SPX's -16.5%. The month's shape: three weeks of chop and washout, and the largest weekly gain of the quarter in its final seven days, funded by $2,647.7M of September ETF inflow (Farside, via YFarmX). This week's -1.1% pause is the first read on whether the month's close holds — and it cost the board only a third of the final week's gain so far.

8. What the small names are telling the base

The week's most quietly bullish datum: BONK +4.6% on the quarter's flattest tape. BONK has now risen in four of the quarter's five weeks without once leading a headline, and its net-from-base figure (+68.7%) sits within two points of PEPE's despite none of PEPE's structural theater. Steady accumulation in a low-beta small name during a pause week is how bases extend. The mirror image is SPX — second-worst week (-2.8%), worst 30-day print on the leaderboard (-24.5%, tracker table via Toutiao, October 4), still +40.1% from base. Both can be true; the pair is the definition of a rotation that has not finished.

FAQ

Related reading

• Meme Coin Monthly Report — September 2026
• Weekly Meme Coin Data Report — 9.28–10.4, 2026
• Meme Coin Weekly News Review — 9.28–10.4, 2026
• Meme Coin Market Intelligence — 9.21–9.27, 2026