ex-HRT quant, now building the venue I always wanted to trade on
singapore
Joined September 2026
- Tweets13
- Following59
- Followers37
- Likes7
I think the bigger question is how many different market structures a trading chain can support.
A retail perp venue, an institutional futures market and a spot exchange serve different users. Their requirements for access, privacy, matching and liquidation management will differ. Sharing infrastructure shouldn’t mean inheriting the entire rulebook of the first exchange built on it.
That’s what makes the extra effort to enter the US interesting. It’s one of the most complex market structures to accommodate: exchanges, clearinghouses and brokers have distinct responsibilities, and the technology has to work within those relationships. That tests much more than whether the chain can process orders quickly.
If Hyperliquid can accommodate those business and regulation requirements, it demonstrates something valuable: operators can build substantially different trading businesses on the same infrastructure. The commercial opportunity expands from attracting traders to attracting venues, brokers and distribution partners.
The benefit is that each operator can reuse the trading infrastructure and focus on the decisions that differentiate its market: attracting liquidity, using capital efficiently and managing liquidations. Improvements to the shared infrastructure can benefit multiple venues, while each retains the rules suited to its customers.
That’s the opportunity I see for us @upside_max too: use shared infrastructure to build the market structure we think traders need. A trading chain becomes much more useful when its growth can come from supporting different exchange designs.
Ok, following up on this in thread form this time...I really appreciate the thoughtful responses to my questions about Hyperliquid. For the record, I am long $HYPE and am a big fan of what the team has accomplished. Not doesn't mean it will translate smoothly onshore.
One thing I forgot to mention (which was brought up in the responses) - will the CFTC permit ADL to exist on a regulated exchange? I suspect the answer is no, but we will see. If not, what are the ramifications? I have briefly mentioned ADL in the past: teapaw.org/drwconvexity/status/19…
ADL inverts the clearing house liquidation waterfall in a default scenario and will be viewed as untenable by any institution.
If you restrict automated trading, you’re filtering the flow as well as choosing the execution model. Less latency-sensitive retail flow is valuable—market makers pay to access it elsewhere.
Being the sole dealer against that flow makes consistent profitability less surprising. It doesn’t tell us whether customers get cheaper execution.
That needs an actual price comparison. But with external makers excluded, nobody can step in and offer a better quote on the same venue
"We are profitable every day" but we don't rip off customers by being the sole MM on @variational_io
The fuck you don't
Good writeup. My concern is that adding more slots doesn’t remove the incentive to buy faster data and optimize the infrastructure on top.
At @upside_max we’ve gone with:
- A flat speed bump on taker orders
- No paid priority on market data
- No rate limits on cancels
The speed bump gives makers time to cancel stale quotes before they get hit. We’re leaving priority fee revenue on the table. The bet is that makers facing less adverse selection can quote tighter spreads, which brings more volume.
gossip (read) priority fees on hyperliquid are broken
currently the team that is benefitting the most is wintermute (insert fist shaking red emoji guy meme)
this is not publicly known but can be determined with certainty by mapping out transfers to labeled wallets from the top winner of the gossip priority fee auctions (they have won 79% of the slot 0 auctions in the past 24h)
tldr; wahhh wahhhh wahhh hl gossip priority, not fair. add more slots or scrap the gossip auction
my proposed solution:
1) increase gossip priority to 20 equal-priority slots
2) make it so winning an auction slot grants IP whitelisting to a foundation-run peering slot
i think this basically solves it and would likely *increase* total gossip priority fees
---
some context
the purported intention, as stated in the docs, is to "reduce the need for market makers to hyper-optimize their infrastructure to remain competitive"
if you win the gossip auction slot, you get order book / node data 50ms (!!) faster than anyone else.
yes. one IP, across the *entire network*, gets the feeds 50ms faster. it's the whole ballgame
so in theory you could just buy the faster data slot instead of spending 10s of thousands monthly on high octane ec2 instances and [insert hft magic]
or you can do BOTH, which is what wintermute is doing
so despite the stated intentions it does the exact opposite and gives 1 wallet a dominant advantage over every other participant
further reducing the democratization of this mechanism is the fact that, practically speaking, you can only use a won gossip slot if you have a direct peering relationship with the foundation node.
multiple providers we work with couldn’t pass gossip priority benefits through to us without slowing their other peers (all independently confirmed this behavior - strange design)
and to get a foundation peering you to do 0.5% of the 14d HL maker volume (we don't quite meet this threshold so our application was rejected) as well as stake 10,000 HYPE ($780,000 here), or be like, a node service provider
---
without revealing too much i can say with 100% certainty that wintermute has maxed out BOTH the infrastructure optimization + expenditure AND is monopolizing the gossip fee auction
they're spending on average $26,000 *per day* ($10M/yr) to win this auction
so rn they're just eating everyone else's lunch
and the only way that approach is economic is if you are MM at the scale that they are, which only a handful of wallets are doing
1/ standing in toledo cathedral. started 1226, finished 1493. 267 years to build. 800 years standing as of this year.
2/ nobody remembers who laid the foundations. everyone photographs the stained glass. the glass is only still up there because the load paths under it were right the first time.
3/ in 1732 they cut a hole in the vault to light the altar. two and a half centuries after finishing, you could still open the roof, because the structure was built to take it.
4/ in an exchange the foundations are the matching engine, the margin rules, the incident process. the ui is the stained glass.
build slow, run fast. we’re spending our time under the floor.
I was constantly impressed by the concentration of talent and brainpower during my last few years at HRT, and always inspired by my colleagues.
Sadly I'm not quite in the billionaire club yet 🤣, but hopefully @upside_max can take us all to the next level!
1/
trading against a pool isn’t new, gmx did it years ago. what’s new at papertrade is who owns the pool: it starts at zero, fills only from trader losses, and losers are paid in PAPER, a claim on its fees and surplus. a house that pays its losers in equity.
2/
we backtested one of our mid-frequency strategies with a flat 25% haircut on wins: sharpe 3 went to sharpe -2. the real haircut is worse on small moves. on btc a 1bp win keeps about 21%, a 10bp win about 72%. that’s where short-horizon strategies make their money.
3/
so honest question: can we just be the LP? per the docs there’s no way to deposit. the only way in is to lose money and earn PAPER first.
Almost 15 years in trading. Teams that ran real share in equities and equity derivatives around the world. I've spent most of my career taking apart the market structure of every exchange I could get my hands on, because I've always cared about how markets work, not just what they pay.
And for the first time in a long time, I feel like a beginner again.
Assets are moving onchain faster than I thought possible. The technology is finally good enough. And I get to build with a group of engineers who are as serious about this as I've ever seen anyone be about anything.
So I'm here to build a market for the new world of trading. Infrastructure that is truly flexible and truly fast. Structure that bends to the asset instead of forcing every asset into the same box.
My promise: I'll be in the trenches with market makers, traders and asset managers until this works for real. Millions of people. Billions of AI agents, in time. Trillions in assets that finally get a home built for how they trade.
This is the most excited I've been in 15 years. Let's take onchain trading to the next level, together.
