The Prediction Vertical · Event Markets

The first prediction market that's actually a market.

Today's prediction venues shipped the ticker and called it the market: the odds are whatever the last trade printed, nobody has to stand behind them, nothing has to catch a fall, and a position can't borrow a dollar. A market is trading plus structure — and that's what this is: a floor of committed capital under every answer, odds only money can move, shares that work as collateral, settlement that can't fail to find a buyer.

The market

Every answer is a share. Every share has a floor.

"Which faction wins the season?" — RED, WHITE, or someone not yet named. Not a yes/no; any set of answers works the same way, and the list can grow:

the outcomesEach answer gets its own share token. Holding RED is holding a claim that pays if RED resolves true. Multi-outcome from day one — binary YES/NO is just the two-answer special case.
the open rosterThe answer list is never final, and markets never restart: a market launches in an open phase where a new contender enters by simply being minted — deposit USDC on the newcomer and they exist, with their own floor, no restart, no permission. Open "President, 2050" today with two names; the list grows as history happens. At a disclosed freeze the roster closes and the market graduates (dial one, below).
mint = fund the floorThe primary way in: minting RED deposits USDC that stands under RED — new conviction is new committed capital, and the floor only rises by funding, never by decree. The same ratchet as the NFT believer stepping the bond from $15k to $18k.
buy the floatThe secondary way in: buy existing shares on the open market at the floating price, which trades above floor by the option value of being right. Prints move freely; floors move only when money does.
the discoveryFloor ratios across outcomes are the market's funded odds — a probability distribution weighted by what each side actually committed. Permissionless: any conviction, any size, no market maker, no license.
the resolverAt the end, someone must say which answer was true. Every market names its resolver at creation — credit-tiered, disputable, disclosed (section 07). Note the word: curators allocate capital and are permissionless; the resolver declares truth and is a chosen, trusted party. The two never blur.
Try it: fund a floor, trade the print around, let the arbs in, resolve it.
Phase
OPEN
Prints sum to
74%
Committed capital
$60k
Funded odds
67 / 33
REDfloor $40k · print 48%
WHITEfloor $20k · print 26%
"President, 2050" — launched today with two names. Bars are floors (committed USDC); percentages are prints vs funded odds. The market is in its open phase: parimutuel pot, anyone can be minted into the race, prints float unanchored. Watch what each button can and can't move.
Price discovery

Two numbers per answer — and only one of them can lie

the printThe floating price — last trade on the open market. Fast, expressive, and cheap to move: a determined whale can trade the print to any number for the cost of slippage — buying secondhand shares off other holders moves the screen, but not one new dollar enters the pot. Prints are for trading, never for truth.
the funded oddsThe floor ratios. Moving RED's funded odds means funding RED's floor — lying costs the price of the lie, held at risk until resolution. Everything that matters — borrowing power, stops, settlements, the published feed — reads floors, never prints.
the honesty gapPrints float above floors; the spread between what the market says and what it funds is a live uncertainty measure no poll or order book publishes. Tight gap: a confident market. Wide gap: talk outrunning money — visible, quantified, tradeable.
The doctrine, one level deeper than tokens: neither trades nor idle cash vote — only money at work does. Same sentence on the Overview; new asset class.
03 · Two dials, dugopen · non-blocking

Settled mechanics, two honest forks

Both forks have a recommended answer and neither blocks the build — they're published here so the reasoning is on the record before parameters get chosen.

Dial one · Complete sets
the forkTwo settlement regimes exist. Pure parimutuel: winners split the whole pot pro-rata — open entry is native (a new contender is just a mint; the pot grows, odds recompute, nothing breaks), but a winning share's payout depends on bets that arrive after yours: poor collateral, and no arbitrage anchors the prints. Fixed-$1 sets: the winning share pays exactly $1 — clean collateral and mint-a-set arbitrage pins the prints, but the answer list must partition certainty, so free-minting a newcomer would retroactively break every existing set.
the lifecycleRecommended: both, sequenced. Open phase — pure parimutuel; contenders enter by being minted, roster churns for years, credit runs shallow (bullet-only, low LTV) because collateral floats. Roster freeze — the list closes at a boundary the market names at creation. Conversion — exact and lossless: a pot of $P becomes P complete sets; a holder of s of RED's S shares receives s·(P/S) fixed shares, each paying $1 if RED wins — arithmetic identical to the parimutuel payout, so no one's claim moves a cent and no consent is needed. Funded phase — arb-pinned prints, $1-cap collateral, full credit unlocks exactly when leverage demand peaks: near the event.
the launchpad, againLook at the shape: curve ↔ pot, graduation raise ↔ roster freeze, curve-price-becomes-reference ↔ parimutuel-odds-become-prints, raise-becomes-seed ↔ pot-becomes-sets, maturation ↔ LTV unlock. A market that launches loose and graduates into a financial system — the protocol's founding move, reappearing on its own.
field, demotedThe FIELD outcome ("someone unnamed wins") is a per-market option, never a requirement — offer it where the unnamed mass is itself a bet worth holding (the early decades of a 2050 race), skip it everywhere else. In the open phase it's unnecessary by construction.
the freeze ruleopen · creator's dialWhere the boundary sits — a calendar date, a resolver attestation ("ballot certified"), or entry-rate-zero — is the market creator's choice, disclosed at creation, exactly like the resolver. Whether LTVs step up at conversion or ramp with time-to-resolution rides along. Left open by design.
Dial two · Unscheduled events
the problem"Will X resign?" has no calendar — so no quiet period, and an outcome's floor becomes a standing bid that whoever hears the news first gets to hit. Every floor fill near an unscheduled resolution is toxic by selection: insiders dumping factually-dead shares onto bidders who can't know yet. Unmitigated, no floor capital survives this lane.
outer deadlineFirst fix: every question must carry one — "X by DATE" (unresolved by deadline → NO). It's also what makes the market resolvable at all, and it restores the calendar covenant: tenor ≤ deadline.
the lookbackSecond fix: floor fills stay pending for window W. A resolution reported within W settles those fills at resolved truth — 0 or 1 — not at floor price. The insider's edge evaporates: dump dead shares into the floor, the report lands two hours later, the fill settles at zero. Honest early exits clear untouched after W. "Internal fills are reversible" — the pending machinery, doing its fourth job.
the tierPlus: risk-engine-capped floor heights on unscheduled categories (deep-discount bids only), and a day-0 tier where exotic resolvers run market-only until floors are earned. Recommended: floors-with-lookback over market-only — market-only forfeits credit and funded odds on exactly the most viral markets.
Leverage on conviction

The trade that exists nowhere

No prediction venue offers native margin. Here the floors make it just another market on the factory:

the positionPost RED shares, borrow USDC against them up to RED's floor, mint more RED. Conviction becomes a sized position instead of a capped bet — bounded by floor depth and the market's caps, like every other market here.
known-fill stopsA stop on a prediction is a conversion into the floor — the buyer committed in advance and cannot leave. In event markets, where depth evaporates on news, "my stop filled 40 points lower" simply cannot happen: the fill prices were funded before the news existed.
self-repaying finishResolve the right way and winning shares redeem at $1.00 — redemption repays the debt with no market, no executor, no slippage. The exit is the resolution itself: the most extreme version of "the settlement counterparty is locked in by construction."
shorting, nativelyShorting RED is buying the field — the complements are the short. No securities lending, no borrow fees, no squeeze mechanics; complete-set arbitrage keeps the two expressions priced against each other.
Priced like the option it is

Probability doesn't diffuse. It gaps.

A frontrunner withdraws and 60% becomes 3% in one block. Bands can't trim ahead of news — so this vertical runs in bullet mode, the template built for vesting locks and NFTs:

bullet loansAll risk is priced at entry and settled at expiry — no mid-course skims or band tranches pretending news arrives smoothly. The entry fee is the premium on an option written on a binary option, priced by jump intensity per event category, not diffusive vol.
LTV by bandLending against a 90¢ favorite is a different animal than a 50¢ coin flip — bounded upside, catastrophic gap. Loan-to-value is a surface over probability band × days to resolution, conservative at the extremes where binaries are cruelest.
the quiet periodTenor ≤ resolution (or the outer deadline) — the collateral itself matures, so the calendar is the covenant. And origination closes N days before scheduled resolution: no fresh leverage into the verdict.
when the gap winsA move no floor could absorb writes down in the same transaction that realizes it — socialized within that event's market, never a neighbor's. The entry fee was the put premium floor capital charged for exactly this.
The floor bidder's seat

One event is a bet. A hundred events is an insurance book.

put-writerSame seat as every market here: paid fees to stand at odds you chose — repaid in full, or assigned the outcome shares at your quoted probability. Never forced to trade against informed flow; never quoting two sides into news. The adverse selection that ate every AMM prediction-market LP structurally cannot reach a one-sided resting bid — and the lookback covers the one place it could sneak in.
the actuarial turnOn one event, floor capital is an option writer. Across a curator's sleeve of a hundred events — sports, macro, governance — binary risk pools into actuarial risk: gaps become a loss rate instead of a coin flip. Event-basket curation is a genuinely new fund category, and the sleeve model was built for it.
uncorrelated carryElection and sports outcomes don't track crypto beta. Event-floor yield is one of the few stablecoin carries genuinely uncorrelated with everything else on the Yield page.
the USDC experienceCleanest of any vertical: winning shares redeem at $1.00 — cash by construction, no executor needed. Losing shares go to zero — the priced tail. Pre-resolution assignments can be held to redemption or auto-sold via the standard cash-preference flag.
The resolver

We killed the price oracle. Nobody can fire the referee.

Someone still has to say which answer was true. Resolution is this vertical's disclosed trusted party — the same honest slot NAV feeds occupy for RWAs:

pluggableResolvers are per-market and credit-tiered: optimistic challenge games, committees, regulated event feeds. The tier is disclosed at market creation and priced into fees — a shakier resolver means a more conservative book, mechanically.
disputes = pendingA proposed resolution doesn't finalize instantly — redemptions and assignments enter the pending window, disputes freeze the clock, finality waits out the challenge period. The reversible-settlement machinery from the Risk page, reused without modification.
the spiciest laneEvent contracts are CFTC territory, and leverage on them is the most regulatorily exposed thing on this site — said plainly. The permissioned-market lane (whitelisted hooks, jurisdiction gating) is the entry ticket, exactly as it is for securities-backed RWAs. This vertical ships where it's lawful and waits where it isn't.
The odds feed

An export the world can quote

the productFunded odds as public infrastructure: the floor ratios and their honesty gap, updated only when committed capital moves. Anyone quoting "the market says 65%" can quote a number that costs real dollars to bend — refundable only by being right.
why it mattersPrints can famously be traded to any number — a single determined whale has steered the most-watched market odds in the world for the cost of slippage. The manipulation budget on funded odds is explicit: move the mark and you've funded the odds you claimed to believe. Lying is a position, and positions settle.
the B2B laneA capital-backed probability feed is oracle-grade input for everyone else — protocols, desks, media. It joins RWA services on the revenue surfaces list: infrastructure others pay to read.
The one-liner: everyone else publishes the last trade. This book publishes what the money believes.
Versus

The seat nobody's filled

PolymarketKalshiAave / MorphoThis book
rolevenueregulated venuemoney marketfloored market + credit layer
oddslast trade — movable by trading alonelast tradefunded odds + honesty gap
outcomesbinary pairsbinary pairsN answers, each with a floor
leveragenone nativenonecan't list outcome shares — no oracle exists, binaries gapbullet loans, priced as options
passive LPbleeds to informed flown/an/aone-sided put-writer, sleeved across events
stopshope for depthknown-fill conversions into funded floors
the venuethe product itself — one brand, one bookone regulated branda commodity on the factory: anyone launches their own branded venue on this market structure
Two answers to "do you compete with Polymarket?" At the product level — yes, from structural advantage: funded odds vs prints, credit vs none, floors that must catch vs depth that evaporates. At the infrastructure level — we're not a venue brand at all: the factory is permissionless, so anyone — a media company, a sportsbook, a community — can launch their own venue on this book and brand the front end. Competitor to Polymarket the product; substrate for the hundred Polymarkets anyone can start. And the door stays open: existing venues' positions can become collateral here through the permissioned lane, whenever they'd rather plug in than fight.
Open parameters

The remaining dials

dial one & twoComplete-set settlement and unscheduled-event floors — dug in section 03, recommendations attached, final call pending.
LTV surfaceThe probability-band × days-to-resolution grid, and how hard it clamps at the extremes.
fee surfaceJump intensities by event category (elections vs. sports vs. protocol governance) — seeded from historical resolution archives, hardened by our own record.
resolver registryTier definitions, dispute-window lengths per tier, and what a failed resolution does to open loans and floors.
lookback windowW per event category, and the exit-fee schedule for pre-deadline floor exits.
cross-event marginPortfolio margin across correlated events — deferred; single-event bullet loans first.
roster & freezeOpen-phase entry rules (who may mint a newcomer, attested how), drop-out treatment (early-resolve-to-zero vs merge-back), and the freeze boundary — the market creator's dial, disclosed at creation (calendar date / resolver attestation / entry-rate-zero), plus LTV step-up-at-conversion vs ramp. Left open by design.
long-dated floorsMulti-decade markets (President 2050) want the idle portion of floor capital in a yield-bearing wrapper — the t-bill lane rhyme; solvency accounting must still see it as present cash.
external collateralAccepting existing venues' outcome shares (e.g. Polymarket positions) as collateral in permissioned lanes — later; native markets first.