Davide

Good read on @PhoenixTrade's perps approach, and strong validation of a thesis shared by @bulktrade @buildonparasol too. The article confirms what we've been building around: general-purpose blockspace isn't optimized for MMs, non-deterministic ordering kills tight quoting, and adverse selection stays high. The winning approach should solve execution without sacrificing composability, the one edge Solana L1 has over HL and EVM L2s. Kudos to them for innovating on perps DEXs, after bringing the top solution on propAMMs 👏

Exchange operators are always looking for liquidity, usually subsidizing 2-3 market makers to come on board and run hybrid maker/taker strategies against each other to create the appearance of volume and post some amount of visible liquidity on the book in exchange for a stipend or rebates. As the market matures those rebates decrease, but can't really ever go away as market making without rebates on leveraged instruments is not generally sustainable under all market conditions, yet markets need liquidity under all market conditions. What's often overlooked is the need to onboard a diverse set of takers, largely retail. A market with just 3-5 bots trading with each other around Binance mid has a verisimilitude of existing, but in fact does not. It's an echo chamber. This is the case for most perps DEXs out there, largely informed flow around the same reference price playing book spoofing and skew sniffing games, trying to catch mean reversion off of each other. It's a negative sum game either for the participants or the exchange depending on incentive structure. In 2010 Easley, Lopez de Prado, and O'Hara built a way to measure this. VPIN, volume-synchronized probability of informed trading, tracks buy/sell imbalance across volume buckets in real time. If you run VPIN on a tail perps DEX book it would read persistently elevated, because the flow is almost entirely informed. You can put a dollar figure on what that uninformed flow is worth. Citadel, Virtu, and Jane Street pay billions for the right to stand on the other side of retail order flow. Hyperliquid took a novel approach to the bootstrap problem via having in-house liquidity with HLP. Lighter took a different path, charging time instead of fees. Zero transaction fees but a 200-300ms delay on standard orders, the delay being the fee, with the vault trading against the delayed flow. This means that even with "zero fees", actual execution costs run 6-12 basis points per trade for standard accounts. Structurally identical to payment for order flow. Both of these venues were able to capture enough retail flow to get the flywheel spinning. So, as an exchange, what are the dimensions of differentiation? Instrument structure, leverage, underlying selection, latency, fee structure, UX. However, if all of the DEXs are trading the same basket of 30 cryptos copying the BitMEX margin system, running perps with the same kappa and iota, using NextJS and Tailwind with vanilla design and copycat product experience, websocket API hosted in Tokyo, then what really is there to differentiate on beyond incentive and fee structure? As mercenary capital rotates to new points programs and incentives, exchanges will live and die by, and only get traction, with proper incentive structures. Like anything there will be a handful that have staying power, and the meta overall is already pretty played out. That handful of venues is where the uninformed flow which makes markets sustainable will trade. For those high frequency trading on these venues there is obviously the latency and modeling race to be played, propagating information between venues and running volume to earn rebates. However, none of these games are really worth playing without uninformed flow present. It may well be, because of all of this, that significant value accrues at the perps frontend layer, enabling cross-DEX basis capture, providing superior execution, curating incentive structures. The exchange becomes the pipe. The frontend owns the user. And the user is the product that makes the market real. In fact, a large share of Hyperliquid's daily active users already trade through third-party frontends rather than the native UI. The top three builders have captured over tens of million in fees via builder codes. On Solana, trading terminals captured the majority of DEX revenue over the past year. This is unsurprising, and mirrors how tradfi played out with platforms like Robinhood. What it means is that owning the frontend experience is hugely valuable. The frontend owns the retail relationship, and the retail relationship is the uninformed flow, and the uninformed flow is what keeps markets healthy.

Under the Parasol ⛱️ - Deterministic txn ordering to protect makers & ensure fairness - Low latency for real-time execution - Dedicated blockspace to avoid contention The result: tighter spreads, deeper liquidity, and a new standard for Solana perps. @OpenBitLab explains:

Perpetual markets don’t just need throughput. They need high-performance execution that isn’t probabilistic. We solve this with an execution environment purpose-built for perps: matching and ordering run inside the engine; settlement, liquidity, and assets stay on Solana. This type of control enables tighter spreads, deeper books, reduced MEV exposure, without breaking composability. But once execution moves outside the base layer, how do you guarantee correctness? That’s where ZK comes in. Fast. Deterministic. Trustless. ➡️ ZK PERPS

See you there! DMs are open

What actually makes a ZK proof work? Last week, we broke down how computation can be encoded as polynomials and constraints. But none of that matters unless you can verify it efficiently. 📝 In Part II, we get to the core of how proofs are actually constructed. The key idea: why check everything when you can just check a random point? From this, we build the full proof flow: - committing to hidden data - compressing many constraints into one - using randomness to make cheating infeasible - turning a huge computation into a tiny check The result is verified computation without re-running it: the point where ZK starts to feel less like a hyped term and more like real tech that can actually power more efficient onchain apps. Full article here: buildwithparasol.io/blog/why-zk-is…

What if ZK wasn’t really about ZK? In many real-world applications of this hyped tech, what matters most is not the zero-knowledge part, but the ability to verify outsourced computation without running it again. 📝 That’s the starting point of our new article. In Part I, we unpack the foundations behind ZK proofs from first principles: - delegated computation - polynomials over finite fields - commitment schemes - circuits and algebraic constraints By the end, we land in a weird but important place: a program encoded as a spreadsheet, where computation becomes a set of constraints that must hold. Satisfy them, and you’ve proven correctness without re-executing the program. These are the building blocks that turn into an actual proof (but we’ll get to that in Part II). If you want a clearer intuition for what’s really going on under the hood of ZK proofs, read this: buildwithparasol.io/blog/why-zk-is…

@therealchaseeb Looking forward to reading it! If you'd like to speak with one perps team more, here I am 🫡

Hey grok. What is my 5,500 word perps article about?

While I understand @EthCC is an Ethereum-centric conference, are there any Solana folks attending it? Or is everyone busy at following Foundation drama?

@Allenbaez126 Building @buildonparasol

If you participate on Solana in anyway I will follow you.

@therealchaseeb @buildonparasol 👀

If you're seriously building/thinking about building perps on solana and don't have a token, i want to talk. if you spam me with a ticker, i'll block you. I'm not toly.

@francescoweb3 @EthCC See you there!

Who should we meet at @EthCC?

Hyperliquid is already trading more oil, gold, and silver than crypto, and now the same trend is emerging on Solana. A new era?

See you there! DM me if you will be around, specifically if you are a trader, prop shop, market maker and looking for the latest innovations about perp DEXs 🔍

Bringing some Sol to the French Riviera ☀️🇫🇷 In 2 weeks, our team lands at @EthCC, Europe’s biggest Ethereum gathering. @instantfinality @inks & @openbitlab will be around talking perps, ZK, and internet capital markets. Shoot us a message if you want to meet IRL!

Heading to @EthCC in Cannes in a few weeks. Any perps traders around? Would love to share what we’re working on and hear your thoughts on the best CLOB-based DEX

At Parasol, we’re building the trading engine to unlock institutional-grade perps for internet capital markets - powered by ZK and native to @solana. But ZK systems today come with inherent trade-offs between fast proving and compact proofs: - STARKs are fast to generate and require no trusted setup, but their proofs are large - SNARKs produce small proofs that verify quickly, but rely on heavier cryptography and trusted setups These design choices become especially relevant on Solana, where fast execution and compact transactions mean proofs must stay extremely small to be verified onchain. One promising direction we’re exploring is combining STARK and SNARK proving methods to balance scalable proving with lightweight verification. In this article, we break down the foundations behind this approach - FRI & KZG polynomial commitment schemes - and how they can support hybrid proving architectures on Solana. 📝 Check it out here: buildwithparasol.io/blog/polynomia…

@KentonPrescott Lot of respect for you 🫡

Rumpel’s time has come to an end. While this isn’t the ending we envisioned, we ran it hard, and I’m proud of what we accomplished. Over 4.5 years, the Sense/Rumpel team pursued two distinct visions in DeFi — fixed income (Sense) and liquid point incentives (Rumpel). We raised during the 2021 highs, weathered the brutal 3ac + luna + ftx winter, ran numerous MVPs, and pivoted to a product users genuinely wanted. Both protocols introduced new markets for exchanging risk, representing our bet on what DeFi does best. I’m equally grateful to have built alongside a genuinely exceptional team. Founded on strong ownership, high autonomy, and low ego — we stayed lean, ambitious, and honest with each other. @jparklev & @0xstvn brought the engineering expertise that made both protocols possible, and @gophillionaire brought the GTM expertise to make our products shine. 0x Labs is lucky to have a couple of them. To our investors and users, thank you for your trust, your feedback, and your willingness to use (& back!) something bold and new. It has been a distinct pleasure building for this community. As for me, I’m not going anywhere. I still believe in the promise of DeFi, a new technology paradigm that brings greater access, expressivity, and capital efficiency to global markets. Taking some time to decompress and think clearly. The next bet is already forming. See you onchain, anon 🫡

It seems the entire Solana ecosystem has an opinion on perps >Anza devs and Foundation folks will tell you its a product skill issue >Current live and in beta perp exchanges will tell you they've solved them (they have not) The reality is that Solana as an L1 is lightyears ahead of everyone else in terms of being able to offer a competitive venue, but is still considerably behind when compared to purpose-built environments There is an insanely high density of perp venues now and building a vanilla isolated margin perp dex is simply not good enough You need a venue of being fully capable of supporting risk engines that can facilitate cross and portfolio margins while doing pre and post matching flight checks, all while being incredibly fast (sub 100ms) You also need to be able to make execution guaranteed, at a fixed and predictable cost, which cannot happen with the current PF model... and even if you solve this, it still doesn't beat the gasless environments you are competing with You need to be able to give these venues and makers a guarantee that they can land every slot, without fail. You can't just ignore the wider traffic on Solana... Priority Fees help, but someone can always pay more. These things are computationally incredibly expensive, and cannot be solved with something as simple as giving makers the rule of the roost. This punishes all takers, not just toxic flow It becomes a dangerous precedent when you virtue signal to an entire ecosystem that there isn't a problem and everything is fine, because then things really don't get addressed and the people actually working on solving the problems have to navigate truth from reality