Outcome Markets

The linearity gap

Spot and perpetual swaps differ in important ways, since perpetuals add leverage and capital efficiency while spot is a tokenized version of an actual asset, but they share one property: both only express linear beliefs. Holding a perpetual position, at any leverage, produces a proportional profit or loss as the price moves. For a large share of use cases that is exactly the exposure a trader wants, but some beliefs are nonlinear or convex, and no combination of spot and perpetual positions can express them: a payoff that is binary, worth a fixed amount or nothing depending on whether an event happens; a payoff that is bounded, with capped upside or downside; or downside protection, wanting to hold an asset while being insured if it drops below some level.¹

The primitive

Jeff Yan's design, Hyperliquid's HIP-4, describes outcome markets as fully collateralized contracts where both sides put up capital and the contract settles somewhere in between, with the payoff able to be either binary or continuous. Because both sides are fully collateralized, there is no liquidation risk, since users effectively post capital and check back later to see what they have, without the price-path complications that can trigger liquidation in a leveraged position, and no funding rate is needed, since funding is a mechanism specific to perpetual contracts.¹

The primitive unifies several categories that are usually built as separate products: options and prediction markets are the obvious cases, and Yan also points to opinion-aggregation markets with no objective outcome, where opinions are aggregated toward a goal such as surfacing the most important issues, functioning as a social-choice mechanism rather than a way of settling a fact. The design philosophy follows Hyperliquid's general rule that native primitives should be as small and self-contained as possible while remaining as broadly useful as possible: rather than building a dedicated options product and a separate prediction-market product, the protocol builds the one convex-payoff primitive both reduce to, and lets builders deploy specific markets on top of it.

Closing an earlier loop

Outcome markets close an eight-year arc for Yan personally. In 2018, in the same era as Kalshi's early prediction-market efforts, he launched a decentralized prediction-markets product that failed, not because the underlying idea or infrastructure was wrong, but because demand for on-chain products simply did not exist yet: "literally no one will try your product." His stated lesson from that failure is that ideas are a very small part of building something successful. HIP-4 applies the same underlying instinct, permissionless infrastructure for placing financial bets on outcomes, at a moment when the ecosystem, liquidity, and demand finally exist to support it.¹

Why it matters

Outcome markets complete an instrument taxonomy: spot for transferring an asset on the ledger, perpetuals for linear leveraged exposure, and outcome markets for nonlinear or convex exposure. Nearly any financial position decomposes into some combination of these three. Delivering convexity as a permissionless, fully collateralized, native primitive addresses the part traditional finance makes hardest, since options markets there are fragmented across strikes and expiries, gated, and operationally complex, and the part decentralized finance has historically done worst.

Tensions

Binary and continuous outcome markets both depend on a trusted resolution of what actually happened, and ceding that real-time settlement truth to permissionless deployers is the hardest, least proven part of the design; prediction markets in general carry a known "arson problem," where markets on influenceable events create perverse incentives, that a permissionless primitive could amplify. Full collateralization also removes liquidation risk at the cost of being more capital-heavy than leveraged perpetuals, and whether convex demand is large enough to justify locked-up collateral at scale remains unproven. And opinion-aggregation markets with no objective outcome are the most speculative use case, blurring into governance and social-choice mechanisms whose manipulation surface is not well understood.