According to Arthur Hayes, co-founder of the BitMEX exchange and CIO of the Maelstrom fund, Hyperliquid’s rise in prediction markets depends on who captures the benefits, not just cheaper trades.
CoinDesk earlier reported that Hyperliquid was preparing a zero-fee opening model for event trading under HIP-4. The Hyperliquid Improvement Proposal (HIP)-4 is a proposal that introduces event-driven trading on Hyperliquid.
Hayes said structure is just the first layer. In a note to CoinDesk, he argued that the real differentiator is HYPE, Hyperliquide’s exchange token, which he says allows users to benefit from the platform’s activity in a way that Polymarket and Kalshi currently do not.
“HIP-4 will quickly become a dominant prediction market due to Hyperliquid’s large user base, much cheaper trading fees, and very robust technology infrastructure,” Hayes told CoinDesk. “Users who own the $HYPE token can directly profit from their use of HIP-4.”
Polymarket is expected to launch a token, often called $POLY.
On Gate, pre-market perpetual contracts tied to a potential $POLY token are trading around $14, implying a fully diluted valuation of around $14 billion. HYPE, by comparison, has an FDV of around $38 billion, according to CoinGecko data.
Pre-listing markets are often highly speculative and may be thinly traded, meaning that any implied valuations should be treated with caution and may not reliably reflect actual market demand.
The argument also comes down to geography. Polymarket registered with the CFTC last July and is rebuilding its US operations, placing compliance at the center of its strategy.
However, in Asia, the company still questions how regulators classify its products. It is geoblocked in Singapore, Thailand and Taiwan, partially restricted in Japan. Meanwhile, in Hong Kong, prediction markets are more widely on the radar of gambling regulators.
Hyperliquid faces no equivalent constraints and its user base is concentrated towards Asia, where crypto-native trading is already significant.
The contrast is clearest with Kalshi.
As a CFTC-regulated exchange, Kalshi’s model is built around compliance and licensing, not token incentives, which likely excludes the type of value accumulation layer Hayes discusses.
This makes it the most direct test of his thesis. Users can trade event results on Kalshi, but they have no path to rising on the platform itself. In traditional markets, this type of upside is usually achieved through shares, such as an IPO, although for now, Kalshi user participation is limited to trading on the platform.
Between the three platforms, the division is structural: Hyperliquid already ties usage to a token, Polymarket seems to be moving in that direction, and Kalshi’s model probably prevents it altogether.
