Recurring Prediction Markets on HIP-4

A market that never stops

Most prediction markets are one-shots. “Will BTC hit $100k by December?” You buy, you wait days or weeks, the market resolves, it’s over. That’s fine for elections and major events, but it’s not very interesting for traders who want to be active.

Recurring prediction markets solve this. A market like “BTC > $95k” can be configured to expire and auto-restart every 15 minutes, every hour, or every day. The moment one cycle resolves, a new one begins. It’s an infinite series of clean, independent trades.

This market type is unique to HIP-4 on Hyperliquid, and it’s the feature that turns prediction markets from a niche product into something that competes with perpetual futures for trader attention.

How it works

Let’s take a real example from testnet: “BTC > $95,000” with a 1-hour cycle.

At the top of each hour, a new market spawns. YES shares start trading near the current implied probability based on where BTC is relative to $95k. If BTC is at $94,800, YES might open around $0.45. If BTC is at $96,500, YES opens around $0.80.

Throughout the hour, the price moves as BTC moves. As the clock winds down, the price converges toward either $1 (if BTC is above $95k) or $0 (if it’s below). At expiry, the market settles automatically based on the price feed. Winners get $1 per share. Losers get $0.

Ten seconds later, a new hour-long market starts. Fresh order book, fresh prices, same question.

Why this is better than binary options

“Wait,” you’re thinking. “This sounds like binary options.” It is — in the same way that a Porsche 911 is a Volkswagen Beetle. Same basic concept, completely different execution.

Traditional binary options (the kind that got banned in most jurisdictions) had these problems:

The broker was your counterparty. They made money when you lost. Massive conflict of interest.
Off-chain settlement. The broker decided whether you won or lost. Disputes? Good luck.
Manipulated pricing. The broker set the odds, not the market.

HIP-4 recurring markets fix all of these:

You trade against other traders, not against the platform. The order book is two-sided.
Settlement is on-chain and automatic. The price feed determines the outcome. No human involvement.
Prices are set by supply and demand on the same matching engine that processes billions in perp volume.

This isn’t a semantic distinction. It’s the difference between a rigged casino game and a transparent market.

Trading strategies for recurring markets

Time decay is your friend (or enemy)

As a recurring market approaches expiry, the price becomes increasingly binary — it’s either going to $1 or $0, and the remaining uncertainty shrinks fast. This is mathematically identical to how options theta decay works.

If you’re a seller (shorting YES when it’s overpriced), time decay works in your favor. If the underlying isn’t moving much, the market will slowly converge to its “correct” price, and the premium you sold evaporates.

If you’re a buyer near expiry, you’re paying for certainty. The trade needs to move in your favor quickly or you’ll lose to time.

The “late break” pattern

Watch what happens in the last 5 minutes of an hourly market when BTC is hovering near the $95k strike. YES might be at $0.52 — the market is a coin flip. Then BTC ticks up to $95,200 and YES jumps to $0.75. Or BTC drops to $94,800 and YES crashes to $0.25.

These late breaks create massive volatility in the prediction market price, even though the underlying moved less than 0.5%. For traders who can read short-term momentum, these final minutes are where the action is.

Cycle-to-cycle momentum

Recurring markets reset, but trader behavior doesn’t. If BTC has been above $95k for the last six hourly cycles, the opening price of the next cycle’s YES tends to be elevated — traders who’ve been winning are overconfident, and sellers are hesitant.

But mean reversion is real. At some point, BTC dips below $95k for a cycle, and all those confident YES buyers get wiped. Fading a streak that’s gone on too long is a contrarian strategy that works here.

Cross-timeframe arbitrage

If the daily “BTC > $95k” market is pricing YES at $0.70, but the hourly market (with 45 minutes left) has YES at $0.60 while BTC is at $95,500, something is off. The hourly market is underpricing the probability given that BTC is already above the strike.

Exploiting these cross-timeframe discrepancies is a real edge, especially because many traders only watch one timeframe.

Why recurring markets matter for Hyperliquid

Hyperliquid already dominates on-chain perpetual futures. Recurring prediction markets complement perps in interesting ways:

Hedging: If you’re long BTC perps, buying NO shares on “BTC > $95k” is a defined-risk downside hedge. You know exactly what you’ll receive if BTC drops.
Leverage alternative: Prediction markets near expiry offer convex payoffs without liquidation risk. Buying YES at $0.10 with 5 minutes left is a 10x-if-correct bet with no funding rate and no liquidation.
Shared margin: These positions share collateral with your perp positions. No need to move capital between accounts.

This is the kind of product integration that standalone prediction market platforms can’t offer. It only works when prediction markets are native to a derivatives exchange.

Try it yourself

Recurring markets are live on Purrdict’s testnet. Trade a few hourly cycles. Notice how the price behaves differently at the start vs. end of each cycle. Watch what happens when the underlying crosses the strike with 2 minutes left.

The price dynamics are genuinely different from anything you’ve traded before. It takes a few cycles to build intuition, and testnet is the right place to develop it.

New to HIP-4? Start with What is HIP-4? for the technical overview.