Why
Put two documents side by side and they draw the same line for opposite reasons. The essay's own earlier critique of prediction markets is that a market only discovers an outcome if participants cannot change it — once they can, the market stops predicting and starts manufacturing. Korea's communications standards commission cited an event the user cannot control among its grounds for treating prediction markets as criminal gambling (jurisdiction-decides-the-category). So uncontrollability is simultaneously what makes the price mean anything and what makes the product illegal in one jurisdiction.
That is a trap, not a coincidence, and it is worth naming before designing around it. Move the product toward events the user influences — skill, own action, own performance — and the legal ground weakens while the epistemic one collapses, because participants now have a position in the outcome they are pricing. Move it toward pure exogenous events and the price is meaningful and the regulator's first ground is fully met. There is no configuration that is comfortable on both axes, which means the choice is a real one rather than a matter of finding the clever framing.
A decision market is what it looks like to cross that line deliberately. Participants are not predicting whether a treasury spend would be good; the price causes the spend. Umia does not pretend otherwise — it accepts that manipulation is possible and tries to make it expensive: settle on a TWAP rather than a closing price, run a No-Op market representing the status quo, and execute only when the proposal beats No-Op by a margin. Note the direction of that last rule: change is not the default, and silence means nothing happens. That is a governance stance, not a market mechanism.
Which turns the whole design into one number. If manipulation is merely expensive, the honest question is how expensive, at what depth — and the essay concedes the weak point itself: before the ecosystem reaches scale, the cost of moving prices enough to influence an admission decision may stay low. That is the same computation governance-capture-cost performs for token voting, transplanted into a market, and it decides whether decision markets are governance or theatre.
How it works
The same property, read by two authorities
| Criterion | Verdict when the user cannot influence the outcome | |
|---|---|---|
| Market designer (this essay) | Participants must predict, not create | Sound — the price means something |
| Korean regulator (방심위) | An event the user cannot control | Gambling — cited as grounds for the block |
| A decision market | The price executes the decision | Deliberately on the wrong side of both, and honest about it |
What Umia does about it, and what each defence costs
| Defence | What it stops | What it costs |
|---|---|---|
| TWAP settlement | A push in the last blocks before close | Responsiveness — the market reacts slower to real news |
| No-Op market | Comparing a proposal to nothing | A second market to seed and keep liquid |
| Execution threshold | Acting on noise | Genuine improvements below the margin never happen |
| Cayman SPC / BORG | A team ignoring the outcome | A shared legal wrapper, and a curated rather than permissionless track |
Read the first row against the-settlement-instant: that card settles at an instant, which is exactly what TWAP is defending against here. The two designs sit at opposite ends of the same trade-off, and seeing the cost written down on both sides is worth more than either design alone.
The number, and why it transfers
Capital required to hold a TWAP above threshold T for window W against arbitrage, at pool depth D. Compute it at today's depth and at the depth where it exceeds the value of the decision. That single curve answers whether a given decision is genuinely market-governed or merely priced, and it applies unchanged to any resolution mechanism that reads a price — including the settlement path in this project. Manipulation cost is not a property of the mechanism; it is a property of the mechanism at a given liquidity, and mechanisms are usually described without one.
What the August auction actually tests
Umia is selling $UMIA through its own platform, so the auction is the product's first public run rather than a fundraise: whether seven days of CCA price discovery behaves as described, whether liquidity exists immediately afterwards, and whether the promised high initial float materializes instead of the low-float pattern it criticizes. A capital-formation model that cannot form its own capital has answered the question early, which is a rare case of a project being genuinely falsifiable on day one.
Update (2026-09-07) — four verdicts, one price, one week
CPI (9/11), a Clarity-bill cloture (9/15), the FOMC (9/15–16) and an options expiry (9/18) land within days of each other. They reprice four different axes — inflation, jurisdiction, monetary, derivatives — but the market prints one number, so any single-cause story about that week's move is probably wrong. The measurement discipline is the transferable part: record a probability snapshot immediately before and after each event, separately. That is the only way to reconstruct afterward which verdict actually moved the price — and it is the same move as this card's core number, which insists a mechanism means nothing without the state (here, the event) attached to the reading. Max-pain ($78,000 on the 9/18 expiry) is a coordinate, not a signal; expiry-week volatility compression is a tendency, not a law, and this week stacks three verdicts on top of it.
Review clarification
What the mechanism is
A decision market is the conditional-market tool; futarchy is a governance system that uses it to choose and execute policy. Umia did not invent either concept. Its distinctive move is to package paired markets, an on-chain treasury and a legal wrapper so the result can bind a token-native organisation.
| Market | Counterfactual question |
|---|---|
| Proposal | What is the token worth if the proposal executes? |
| No-Op | What is it worth if this proposal does not execute? |
One price is insufficient: Proposal at 120 is favourable if No-Op is 100 and harmful if No-Op is 150. No-Op means this proposal is not executed, not that nothing else ever changes.
Participation and manipulation
The pair creates a comparison; it does not create informed traders. Thin participation can leave a large treasury dependent on one trader, and incentives can add volume without information. A credible rule therefore needs minimum liquidity and independent participation as well as a TWAP spread.
expected net manipulation cost < external value of the decision is a general attack test, not an official Umia formula. It matters because a trader may lose in the market yet profit from a grant, contract or strategic benefit produced by the decision. D, T and W alone are insufficient: the AMM curve, fees, external fair price, arbitrage capacity and speed, order flow and recoverable position value also matter. Likewise, uncontrollability may be one legal factor in gambling analysis; this card does not establish it as a sufficient condition by itself.