Why
Two stories, one finding: a well-formed question routed through an instrument that answers a different question.
The first is the rally, and the rally is not the interesting part — the routing is. Does the Treasury deploy the General Account before a given date is about as clean a binary as exists: one decision-maker, a published daily balance, an unambiguous resolution source. It is precisely the shape a prediction market settles well. Instead the belief was expressed through Bitcoin, where the signal arrives mixed with dollar strength, ETF flows, leverage positioning and everything else that moves a global asset in a week. A 22% move tells you a belief changed. It does not tell you which belief, or by how much.
And the trigger makes it sharper than an ordinary macro trade. No commitment was announced. What moved was a quote — I have asymmetric information; what do I know that the market doesn't know — from the person who decides the outcome. decision-market-uncontrollability argues a market prices honestly only when participants cannot change what they are pricing; this is the same defect from the other side. The subject of the bet has a microphone, and one sentence repriced the asset by a fifth. A market that cannot separate the decision-maker hinted from the decision was made has no way to express the difference, because the only instrument available is a price that means many things at once.
The second story is the same defect with the easy half removed. One line in an exchange's promotional report is a different kind of fact from the rest of it. Volume totals and growth percentages are the exchange's own numbers and the honest thing is to note them and move on. But the largest single instrument by weekly volume is SPCX, and SpaceX is a private company. There is no continuous public price for the thing that instrument tracks, which means $378M of weekly volume settles against something — and the report does not say what.
Every hard part of tokenised equity lives in that gap. tokenized-equity-claim-rail worked out that a token which is a claim moves the problem from financial engineering to bookkeeping, and that the custodian's ledger is authoritative. A private-company instrument has no custodian's ledger to be authoritative about in the same way: shares change hands rarely, by negotiated transfer, often with transfer restrictions and a right of first refusal. So the price is not discovered continuously by a market — it is published, by someone, at intervals — and every property that matters follows from which someone and which interval. the-settlement-instant asks what settling at an instant requires; this is the version where the instant may be weeks old.
Both stories also come with a discipline about claim status, and it is the same discipline. fork-date-provenance sorts claims into confirmed, deferred, and circulating without a source. On the macro side: the Treasury has the option is confirmed; the balance is under consideration is a report of deliberation; the balance will be deployed has no source at all — and the rally priced the third. On the exchange side, 7-day above 30-day above 90-day is presented as proof of compounding and is equally consistent with one strong recent week; Korea's +689.9% sits in the same window as a second liquidity trading session being added, so a demand number overlaps a supply change — the precise defect monad-last-general-purpose-l1 was built around.
And one ambiguity is load-bearing enough to name. US: $25.6B, 66.1% does not say whether that means US-listed instruments or US-resident users. Those are completely different facts — the first is product mix, the second would be a regulatory position that sits oddly beside Coinbase restricting its tokenised equities to non-US users. The word that decides which is missing.
How it works
Case one — three states, one price
| Claim | Status in the source | What the market did with it |
|---|---|---|
| The Treasury has the option to deploy the account | Confirmed — roughly $936B on 08-19 | Correctly priced as a standing possibility |
| The balance is under consideration | Reported deliberation — CNBC, 08-24 | Read as intent |
| The balance will be deployed | No source | This is what the 22% priced |
| Buyback capacity raised | Confirmed — $2B to at least $4B per operation from 08-19 | Treated as evidence for the row above |
The operations number worth keeping: in the most recent buyback, $20B was offered and $2B was taken. Dealers wanted to sell ten times what the Treasury bought. That ratio is the closest thing to a fact about appetite versus deployment in the whole story, and it points opposite to the rally. Whether the gap narrows across the 09-09 to 11-04 window is the observable that settles the thesis, and it needs no model — the operations are published.
Case two — the one instrument that is not like the others
| Instrument | Underlying | Continuous public price? | Settles against |
|---|---|---|---|
| GOOGLX, AMZNX, TSLAX, MSTRX, NVDAX, INTC | Listed equities | Yes — an exchange quote | Licensed market data |
| SPCX | SpaceX — private | No | Unstated. This is the card. |
The three questions that follow, whichever source SPCX settles against:
- How often does it update? A price that moves quarterly cannot support intraday trading without something filling the gap — and whatever fills it is the real price source.
- Who can move it? A mark set by a small number of transactions is one a small number of participants can influence —
decision-market-uncontrollabilityarriving through valuation rather than governance. - What happens between updates? Positions opened against a stale mark and closed against a new one are exposed to a step change nobody trades through — a different risk from volatility, and usually not disclosed as one.
The growth claims, marked
| Claim | What is stated | What is not |
|---|---|---|
| Compounding growth | 7-day $436.8M > 30-day $386.4M > 90-day $320.4M | Whether the series is monotone or one strong recent week |
| Korea +689.9% | Growth against 30 days | That a second liquidity session was added in the same window |
| US 66.1% | A regional split | Whether US means listed instruments or resident users |
| Rotation, not exit | Inflows to large caps, outflows from high-beta | Net flow — rotation and withdrawal look identical per-ticker |
What the card would actually build
Write the binary, then price it twice. Does the TGA balance fall by more than X between date A and date B, per the published daily statement. Compare the probability implied by the move in an asset that responds to everything else against a direct binary quote. The gap is the cost of routing a specific question through a general instrument — knowable, publishable, and currently published by nobody.
The honest caution, and it applies to both
A clean question is not a liquid market. The same funnel that makes the macro case a good demonstration — one decision-maker, public resolution — also makes it a market where the decision-maker's own statements move the price. Building the instrument does not remove that; it makes it visible and prices it, which is the most that can honestly be claimed. And on the tokenised side, the reason this is worth a card is not the volumes. It is that the highest-volume instrument on a tokenised equity desk is the one with no public price — the exact case where the settlement question cannot be avoided by pointing at an exchange feed. Every card here about tokenisation ends in the same place: the ledger is the easy half, and the authoritative record lives somewhere else. A private-company instrument is that argument with the easy half removed.