Donovan

Donovan analyzes 224K wallets that traded TradeXYZ markets between Oct 2025 and Apr 2026. 47% had zero prior Hyperliquid activity — a sybil signal. But trade-size distribution is mixed, and the largest user spikes map onto the Strait of Hormuz crisis (93% of the March surge traded $CL crude oil) — organic geopolitical trading, not coordinated farming. The decisive signal is frequency: median xyz-only wallet made 2 trades on 1 day then went dormant; 78% inactive within a week vs. multi-market wallets' median 144 trades over 69 days. Read: meaningful sybil activity in the user count, but a real organic long tail underneath.

Donovan argues HYPE at a $9 billion valuation looks expensive. A reverse DCF assuming 30% returns over four years requires $11.5 billion in revenues by 2030—implying 110% CAGR from the current $601 million annualized run-rate, growth rates with no historical precedent in exchange history. His bottom-up analysis suggests base case revenues of $4.7 billion by 2030, creating a $6.8 billion shortfall; only the bull case of $14 billion in revenues justifies today's price, but that requires DEXs capturing 60% of a vastly expanded perps market while Hyperliquid holds 45% share—assumptions pricing in most of the upside already.

Donovan argues AI agents on blockchains remain mostly a meme: x402 onchain transactions peaked in November 2025 then collapsed, with merchants offering only speculation plays rather than useful services. Three binding constraints prevent growth: discovery (no registry of x402-enabled services), identity (no way to verify unknown wallets), and reputation (no chargeback mechanisms). The missing layer is an agentic PageRank combining onchain volume, attestation reviews, and completion rates—whoever builds it could own the agentic economy's monetization funnel, potentially larger than Google's AdWords.