$SUBINDEX SAME BASKET · SUPPORTS INDEX EXPERIMENT 01 · 17-ASSET STRESS TEST

Basket breadth, not fee volume, is what stalls these payouts — and we are testing where the ceiling is.

Distributions on this launchpad are not autonomous. A single keeper wallet executes every round for every treasury, and a round is not one transfer — it is a Merkle commit followed by one transaction per holder. We measured 43 payout transactions in four minutes for one coin with 32 holders, zero failures. Keeper throughput is ~11 tx/min against ~900 holder-leaves across the live set, so a full sweep of the ecosystem takes roughly 84 minutes at best. A treasury advertising a 15-minute epoch is describing an intention the architecture cannot currently honour; a 15-minute epoch is a floor, not a schedule.

Basket breadth multiplies that cost. Every additional reward asset adds a conversion step before a round can even be computed, and then adds transfers per holder. The pattern in the data is suggestive rather than proven — the two widest baskets on the launchpad (17 and 8 assets) have executed zero rounds, while single-asset baskets fill the top of the rounds table — but it is confounded: those wide-basket coins also have no trading volume yet, and therefore no fees to distribute at all. Untangling breadth from fee starvation needs a case where one is held fixed.

So that is what SUBDEX is for. We operate SUBINDEX — deliberately the most demanding configuration available on the launchpad: 17 reward assets, a 15-minute epoch, 100% of distributions to holders, nothing to the creator. It is a test instrument, not a pitch. Every round it completes, fails, or delays is published here alongside every other protocol on the same terms, and we will report what breaks even when what breaks is ours. Disclosure: SUBDEX operates the SUBINDEX treasury. Nothing on this site is investment advice.

What SUBINDEX has actually shown, including where it failed.

The honest scoreboard: SUBINDEX has executed zero rounds and paid nobody anything. It holds 17 reward assets on a 15-minute epoch with 100% of distributions to holders — the most demanding configuration on the launchpad — and none of that is why it has paid nothing.

It is not that there were no fees. The fees exist and have not moved. The coin's own page reports $155.56 of fees collected against $0.00 distributed to holders and $0.00 creator earnings, across zero rounds. So the failure is not an empty treasury — it is a full one that has never executed a distribution. SUBINDEX also never graduated its bonding curve, and the most likely explanation is that the two are the same fact: the payout machinery does not begin running rounds until the coin migrates. We are stating the measurement and naming that as a hypothesis, because we have not yet watched a coin cross that boundary and start paying.

A correction, in public, on our own page. This section previously said there were no fees to distribute at all. That was wrong — $155.56 had been collected. We had been reading a treasury field named harvested as "distributed to holders"; for SUBINDEX it reads $0.08 while the coin page reads $155.56 collected, so the two are plainly different quantities and our label was the mistaken one. The fee columns on this site are being re-sourced against the figures the coin page itself publishes, and until that reconciles we would rather show the discrepancy than a number we cannot defend.

What we cannot yet claim. The tempting story is that 17 assets is too many, and the surface data looks supportive — every coin with eight or more reward assets has zero rounds. But that is two coins, SUBINDEX and one other, and neither has graduated either, so breadth is not being tested at the high end at all; the sample cannot separate it from the same fee starvation. Against it: BRRR runs 7 assets and 49 rounds, TEST 7 assets and 35. The keeper arithmetic is real and measured — ~11 tx/min, one transaction per holder, ~84 minutes for a full sweep, so a 15-minute epoch remains a floor rather than a schedule — but the claim that breadth is what stalls a payout is unproven, and we are not going to assert it because it flatters our own next launch. Nor does the distribution split explain anything: JACKET routes 100% to holders and has executed 184 rounds.

So the next instrument fixes the variable that actually failed. The lesson SUBINDEX taught is not "use fewer assets" — it is that a treasury is worth nothing until its coin graduates and trades. The next instrument is built to clear that bar first and to keep every round cheap enough that breadth never becomes the excuse: SUBDEX ($YLD), deliberately the cheapest round we can construct rather than the most demanding: a single reward asset, an epoch longer than the measured sweep, and USDG as the pair. Same operator, same publishing terms, opposite end of the cost curve. SUBINDEX stays live and stays reported — a control is only useful if you keep running it. $YLD is not deployed yet; nothing on this page is an offer or investment advice.