avery.apt 🇺🇸
9.6K posts

avery.apt 🇺🇸
@AveryChing
Co-founder & CEO @AptosLabs. Building the global trading engine & decentralized cloud. Ex @Meta Libra/Diem crypto techlead. Supercomputing PhD. @CFTC GMAC sub















no one is pricing in how fucking atrocious @Polymarket infra is the orderbook just does not work. It’s an offchain orderbook and it’s somehow slower than @DecibelTrade’s cranker (a wallet which pays gas to advance the orderbook each block) for its fully onchain orderbook the only reason polymarket is polymarket is that it was the first mover. Most of its revenue comes from up/down markets, which have terrible UX because of how slow the orderbook is and how unreliable execution is when you are betting where price goes in the next 5 minutes, you need execution to be near instant this isn’t some esoteric secret it’s not good enough for trades to take seconds to go through and then have to spend the next 5s figuring out if it actually executed. Just placing a trade has already taken up 1/30th of the market’s lifetime if you’re betting on a sports match, a 3s delay on the book is just too long minutes, hours, days, weeks is too long for a market to resolve the only reason polymarket has any liquidity is due to pure brute force distribution — there are just so many people playing on it that there is bound to be liquidity but this is hitting a ceiling — most people that will use polymarket, certainly as makers, are already using polymarket and it just does not make sense to be a market maker 1) you don’t have cancel priority —> you’re bending over, just waiting to be adversely selected 2) the very nature of prediction market resolutions mean that you already have a much higher risk of adverse selection 3) the LP rewards, despite pm routing pretty much all of its taker revenue to them, are nowhere near enough to cover the adverse selection 4) building market making infra on polymarket is a development nightmare and it’s almost impossible to get it to work reliably for these reasons, I think polymarket (and kalshi) is facing an extinction event with HIP-4, at least on objective markets because 1) HL has cancel priority 2) HIP-4 resolutions are instant and objective 3) HIP-4 has no opening fees and only 7bps/4bps fees on closing (pm charges 2% on winnings and has fees on every open/close, kalshi charges 7% on every open/close) 4) aside from the liquidity that will come from market makers actually being able to MM reliably on a performant orderbook, HIP-4 taps into hypercore’s already exceptional liquidity 5) the data feeds from HL, thanks to things like @hydromancerxyz, @hypedexer and @hyperpc_, are much better than pm tooling 6) MMs can hedge more easily because they can use spot, perps and defi (on hyperevm) without leaving HL. I think this might be the catalyst that corewriter needs 7) jeff is building it I’ve seen many people speak about the HL S3 and polymarket airdrops in the same breath, but even if POLY TGEs high, it will be down only imo all of the bullishness is just driven by speculation and the fact that it’s so widely adopted, but that adoption isn’t monetised well (even since the fees turned on, they’ve just been going to LPs in a desperate attempt to incentivise liquidity) and even if it is, most of it won’t go to the token and that isn’t even considering the very real possibility that they screw over users on the airdrop. They have VCs who they sold equity too and they run a gambling company — you’re not going to find the hyperliquid ethos here aside from tokenomics, HIP-4 will force them to completely abandon their current infrastructure. They need to leave polygon and rebuild their orderbooks from scratch and I can’t see them being afforded the time to do that once jeff turns his attention to your vertical, you’ve already lost hyperliquid








