HIP3 silver markets have grown to 2-3% of CME activity and have begun to show signs of weekend price discovery. And combined with unified accounts and portfolio margin, the UX inefficiencies of converting collateral to trade different markets are getting resolved. Now HIP4 expands the second dimension of trading: the payoff function for markets.

Previously, we only had linear perps. HIP4 introduces the generalized outcome market, any market where users pay some contract cost up front and receive a payout at maturity, without liquidation risk. These outcome markets introduce convex payouts, which means that we can now construct any payoff shape on Hyperliquid.

Outcome market mechanics

Perps are linear. If you’re long and price up, you profit. Price down, you lose. If you’re short, the opposite. Simple arithmetic, which is the problem—the arithmetic includes every tick between now and when you're eventually right. Take the example of Alphabet earnings on Feb 4. GOOGL tanked 7% during the call and immediately rallied back. This caused $7.5m in liquidations on GOOGL. The market remembered every step of the path even though traders were directionally betting on earnings performance.

[GOOGL earnings on Feb 4] Sudden retracements are not very kind to highly levered perps positions. But such dislocations also provide very good fills.

Outcome markets won’t work this way. Like perps, they are also derivatives on a reference market. But they aren’t path dependent. For outcome markets, you pay a premium and you hold to expiry. Even if price moves super unfavorably immediately after the event, you can’t be liquidated. This time has value, and as a result the payoff curve is curved precisely because of that time protection. People pay for this protection and often they tend to overpay because for most people getting liquidated feels really bad. But maybe in perps land where liquidation is more normalized, people will underpay for this protection (there’s probably a trade here against tradfi options!)

If outcome contracts are curved, how do you price them? One component of pricing outcome contracts is the volatility of the underlying. Which makes the trade-off pretty straightforward: if volatility turns out lower than priced, you gave away upside for protection you didn't need. The perp would've been the better trade. If you traded the perp, you kept the whole move instead of paying for convexity. In that sense perps give you more direct exposure to the movement of the underlying.

We simulate the pnl comparison over time between perp and option held to expiry (t+10) with implied volatility 100% using BS pricing on 10,000 GBM price trajectories. At the beginning payouts are equal but as time passes, the pathwise dependency of the perp changes the pnl distribution.

If you actually simulate out the payouts comparison between an outcome/option and a perp you get some nice intuitive conclusions.

  1. In lower volatility the options expire worthless but your deltas didn’t change much (left peak).
  2. Other times you got liquidated on the perp side and the option expired worthless (the middle).
  3. And then sometimes you get liquidated pathwise like the BTC move on Feb 5 but then price completely retraces.

So they're complements. Together they can generalize any payout curve for an underlying, perps for the linear exposure, outcome contracts to provide the curvature.

And if they operate under the same margin account, as suggested by portfolio margin and unified accounts, the capital efficiency compounds. You're not posting collateral twice. If you're scalping premiums or fitting a portfolio across onchain and offchain positions, that could be the difference between viable and not.

Looking forward

Binary markets on BTC and other assets will become a lot more efficient through HIP4 due to unified margin.

None of this works without liquidity. An HLP4 for outcome contracts seems like a necessity, or some other vehicle to bootstrap taker volume. These types of liquidity vaults can probably serve as a backstop source for permissionless deployments under a “request for liquidity” type of model. The deployer creates a market, request HLP4 allocation, the vault quotes it if it thinks it can make money.

Part of the challenges for smaller markets on HIP3 is that it’s very difficult to put on larger size without eating big slippage. Such a communal liquidity mechanism makes it more viable for deployers to build novel markets (where HIP4 might make more sense than HIP3) rather than spending cycles to bootstrap liquidity.

Few words from @GuthixHL:

"The time has come to replace HIP-2 with a dynamic liquidity provision that allows for user deposits how do we beat AMMs on CLOB for exotic illiquid pairs? the fate of the Hyperliquid spot markets depend on it."

Of course, liquidity is just one dimension. The other is what types of outcome markets to create. One of the visions for the Hyperliquid ecosystem is to enable true offline price discovery. We saw the silver market last weekend somewhat align with Sunday open. For more weekend liquidity and takers to come in, the markets need different ways to enter and exit risk when offchain markets are not active.

This means strictly on-chain methods to hedge out idiosyncratic risk (i.e. internal vs external pricing risk for HIP3). Otherwise, OI will remain thresholded. This is exactly where HIP4 can close this gap and exponentially increase the capital onboarded.

  1. Weekend gap options. These would pay out based on Friday 17:00 ET close versus Sunday 18:00 ET open.There are a lot of diverse portfolios that you can construct from these types of options and underlying linear deltas.
  2. Internal pricing <> external pricing deviations. These would pay out based on the max dislocation for internal pricing for an HIP3 deployer and can hedge liquidation risk.HIP3 deployers have different logic regarding mark and oracle over the weekends. Some threshold liquidation risk because they bound mark price. Others let mark price float.
  3. Funding rate options. These would let you hedge out negative funding risk when positions are hard to exit.Weekend funding has a lot of iterative room for improvement because of the unique combination of internal pricing <> excess OI relative to liquidity. As a result noisy activity causes a lot of funding thrash over the weekend.

The design space for potential markets is massive. We’ve seen throughout the space both breadth first and depth first product exploration. The unique moat of Hyperliquid is that these outcome markets can be embedded into an all encompassing exchange, and users can trade any asset with any payoff function in the same margin account.