Hyperliquid (HYPE)

TLDR: - Toly is describing a different design space (bounded manipulation window + conservative risk so insurance can cover it). - The ADL trilemma is about the structural deficit regime where you choose to run with more tail risk than your fee-funded insurance can support. In that high-leverage, heavy-tail regime, no choice of queue vs pro-rata vs fancy ADL rule fixes the basic budget: something among solvency, fairness, or revenue has to give. Please allow me to expand my thought. I think the key mismatch is about what regime the theorem is talking about (w.r.t. Toly's comment below). The ADL trilemma result is proved under a structural deficit assumption (J.3): the expected tail-loss rate mu_- is larger than the max fee diversion mu_Phi you can send to insurance without killing LTV. In that regime, no static ADL+insurance policy can give you all 3: (S) tight solvency, (F) bounded moral hazard, and (R) unchanged revenue. IMHO, @aeyakovenko's construction here (bound oracle attacks to a window T, de

HIP-3 volumes have just exceeded $8 billion. https://t.co/BKOi2sAink

~$1,4M on a slow Saturday Hyperliquid https://t.co/QMREMP7YU7

When I pointed out problems with Tarun's understanding of Hyperliquid, he challenged me to read the rest of his paper, suggesting I was too dumb to understand it. I don't like obscurantism or gatekeeping. So I did what he asked. Here's what I think: * The paper is wrong about what ADL is, even more fundamentally than what I pointed out yesterday. * The paper proposes that exchanges should sometimes take on platform-level insolvency risk, which I think is not a good idea. * And the core “ADL trilemma” is basically a restatement of its own assumptions, including an explicit assumption that insurance funds won't work. 🧵