How To Improve Solana's Market Structure | Eugene Chen
🎯 Summary
Podcast Episode Summary: How To Improve Solana’s Market Structure | Eugene Chen
This 55-minute episode of Light Speed features an in-depth discussion with Eugene Chen, co-founder of Ellipsis Labs, focusing on the technical challenges of on-chain market microstructure, particularly on Solana, and the strategic shift toward building specialized Layer 2 solutions.
1. Focus Area: The primary focus is Crypto/Web3, specifically Solana’s blockchain architecture, decentralized exchange (DEX) market structure, Central Limit Order Books (CLOBs) vs. AMMs, and the development of specialized Layer 2 solutions optimized for high-frequency trading (HFT) applications.
2. Key Technical Insights:
- Proposal Monopoly & Sequencing Control: The core limitation of Solana for efficient exchanges is the “proposal monopoly” held by the leader validator every 400ms, allowing for manipulation of transaction ordering. While multiple concurrent leaders aim to mitigate this by distributing ordering authority to the protocol, a centralized sequencer (as planned for Atlas L2) offers superior, continuous block building necessary for HFT efficiency.
- CLOB Superiority for Quality Liquidity: For highly liquid and competitively traded pairs, CLOBs provide higher-quality, tighter, and more profitable liquidity compared to AMMs, where LPs can be tricked into negative EV trades masked by yield.
- Atlas L2 Specialization: Atlas is being built as an opinionated Layer 2, specifically designed to be the best possible home for high-frequency trading applications like Phoenix, prioritizing features like reliable oracle updates and prioritizing market maker welfare over toxic taker flow.
3. Market/Investment Angle:
- Market Preference for Efficiency over Max Decentralization: The market currently favors centralized exchanges (CEXs) due to superior volume and flow, suggesting that maximum decentralization is not the current “North Star” for most users; product quality and efficiency are prioritized.
- Perpetuals as Bootstrapping Tool: Perpetual futures are strategically advantageous for new chains (L1s or L2s) because they abstract away the complex asset bridging problem, requiring only a single collateral asset (like USD).
- Nuance in CLOB Discourse: The current discourse often oversimplifies CLOB implementation, focusing too much on data availability (like Hyperliquid) rather than the critical, nuanced details of market microstructure required for true efficiency.
4. Notable Companies/People:
- Eugene Chen (Ellipsis Labs): Co-founder, expert in market structure, formerly involved in math competitions (quant background).
- Ellipsis Labs Products:
- Phoenix: The initial fully on-chain Solana CLOB (proof of concept).
- Gavill: A platform for on-chain token launches designed to minimize MEV extraction.
- Atlas: The new, specialized L2 blockchain designed to host the next generation of Phoenix.
- Solphai: The largest on-chain market maker on Solana, developed by Ellipsis.
- Katana: Sponsor, a DeFi-first chain incubated by Polygon Labs, focused on deep liquidity and high yield.
5. Regulatory/Policy Discussion: The discussion touched upon the spectrum of trust models. Chen explicitly avoids labeling Atlas as “decentralized” because it has a permissioned, centralized sequencer operator (Ellipsis Labs). He argues that systems should be honest about their trust assumptions (e.g., custodial vs. non-custodial) rather than mislabeling themselves, contrasting this with CEXs where users trust the operator entirely.
6. Future Implications: The industry is moving toward specialized execution environments (application-specific L2s) where the sequencing and ordering layers are tailored precisely to the needs of the application (e.g., HFT). The importance of the settlement layer (like Ethereum) is viewed as secondary to the quality of the execution layer for most applications today. There is an expectation for more experimentation in L2 design, particularly around transaction ordering mechanisms.
7. Target Audience: Crypto/Web3 professionals, DeFi developers, quantitative traders, market makers, and blockchain architects interested in the technical limitations of current L1s for high-performance applications and the strategic rationale behind building specialized L2s.
🏢 Companies Mentioned
💬 Key Insights
"Whereas on Solana today, the incentive is just to maximally screw the user whenever you can if you are some provider somewhere in the infrastructure. And that is just not conducive to internet capital markets. It is not conducive to bringing good assets on-chain and having very liquid markets on-chain."
"...the incentive is set up such that the validator is incentivized to be maximally greedy on behalf of their stakers. If they're not, their stakers have economic pressure to move elsewhere. And that means the validator needs to be building the most valuable block or fucking the transactions as much as possible."
"Current crypto market structure incentivizes people to create bad fills instead of good fills, and if regulators catch on to this or something like this could be negative for crypto."
"This market structure is induced by Jupiter caring about the quality of their quotes because I think it's quite noble from them, to be honest, where clearly their end users on the website just do not care if they get half a basis point of price improvement. No way they care; there's no way they know. But Jupiter cares enough to make giving the best price the most important thing for getting the order flow."
"I think it's pretty embarrassing that crypto likes to talk about being so transparent and everything goes on the blockchain, and then you have this whole centralized exchange business where it seems like everyone who's launching a token is signing 12 of these side deals with some really shady actors, and it's totally accepted. It's almost like required."
"the validator does have full agency over what they put into their own blocks, whether they're actively front-running or delegating the front-running out to somebody else, which, by the way, is illegal in trad-fi, and it's unclear the legality in crypto, but certainly the protocol itself allows it; the protocol cannot get rid of it."