matteo

Matteo explains why Hyperliquid's priority-fee revenue hasn't ramped: validators must explicitly enable the gossip priority config and most haven't, so winning the auction today doesn't guarantee prioritized mempool access. Pre-upgrade, API traders paid validators tens of thousands/month for sentry peering — the new mechanism internalizes that, adding ~$500K–$1M/mo HYPE buying pressure immediately. BIS estimates $5B/yr global HFT extraction; HL growth-mode markets charge 0.45–0.9bps — capturing priority could roughly double protocol revenue on those. Bold take: priority fees become >50% of HL's revenue in a few years if TradFi flow grows.

Matteo analyzed Hyperliquid's weekend trading across 35 HIP-3 instruments and found 100% directional accuracy predicting Monday's opening gaps, with a regression slope of 1.06 and R² of 0.973—median prediction error just 14 basis points. The cleanest signal arrives around 20:00 UTC, three hours before CME reopens, when liquidity providers still maintain 66-84% of book depth; in the final hours, metals overshoot (Gold slope jumps to 1.61) as books thin and convergence trades distort prices. Alpha exists in knowing when the signal is purest and fading opening dislocations between perp mids and oracles, which mean-revert within minutes.