animol

x.com/i/article/2033…

Birth date: 11th December Find us at Breakpoint @SolanaConf for exclusive access to mainnet. The journey of it all 👇

The event to end off the year. December 11th. Trading Lounge at @SolanaConf, co-hosted with @AphelionTrading. Sign up now, link below 👇

x.com/i/article/1996…

With mainnet reaching feature completeness, here's the primer on why we started @bulletxyz and how perp trading is coming back in a big way to Solana 🚅

Catch up with BULLET Co-Founders @PDAnimol and @Tristan0x, November 25th at 9am EST- right here on X.🎙️ Coming to you live from BULLET city🏙️

After analyzing @bulletxyz, I’m convinced its architecture is superior for delivering a high-frequency trading (HFT) execution environment compared to Solana’s current offerings. Bullet aims to be the Nasdaq onchain. How Bullet Works Bullet runs a streaming execution model: orders are pre-executed and return a millisecond soft-confirmation; in the background the sequencer batches thousands of transactions and settles them. With a single sequencer, rollback risk is lower than L1’s skipped-slot profile, so cancels land and fills feel deterministic during volatile periods. There is no public mempool to fight. Ordering is deterministic (FIFO or a stated policy), so traders aren’t worrying about random block placement. On correctness, the entire workload is zk-provable. The engine runs at native Rust speed and then proves state transitions asynchronously on a cadence, posting succinct proofs to Solana. That takes L1 compute limits out of the critical path and enables feature-rich risk and unified margin while keeping the millisecond path at the point of trade. As a result, trades execute instantly, with proofs generated on a cadence, providing users with deterministic soft-confirmations and verifiable cryptographic proofs on Solana for all users to see on the public ledger. Succinct proofs let Bullet compress off-chain risk calculations into small attestations, cutting verification costs by an order of magnitude or more versus re-executing on L1. The ZK here is for verifiable computation: execute at native Rust speed, prove later on a cadence, and verify cheaply on Solana. This prioritizes millisecond trading while enabling real-time netting and robust risk models (e.g., portfolio VaR). At its core, Bullet's sequencer acts as the gatekeeper for transaction ordering and block production in its Solana extension (a rollup-like setup with the SVM). This gives the team full control over how orders are sequenced, batched, and committed, bypassing the consensus bottlenecks of a shared L1 like Solana. Why This Matters For the team, this unlocks rapid iteration on market structures, think custom order types (e.g., advanced limit orders or auctions), tailored risk engines for unified margin (leveraging ZK proofs to compute complex collateral netting off-chain which should be a step function better when managing options), and flow segmentation which can power a payment for order flow (PFOF) business model allowing them to bootstrap more liquidity by advertising “Zero Fee Perps.” I say all this to hit home the fact that they have a lot of flexibility in how they operate. Thinking in Bets: The Payoff-Matrix Method I want to lay out my framework for comparing two architectures: game theory and payoff matrices. Annie Duke’s book, Thinking in Bets, has had a huge impact on me in my investing career. How do you make good decisions without all the information? You must weigh probabilities and make the best possible decisions with the information that you do have. This does not mean you will always be right. And when you are right, that does not always mean you made the optimal decision. Just because you bought something and it went up does not mean it was a good decision. Just because you bought something and it went down does not mean it was a bad decision. So how do you make good decisions? Weighting the probabilities and making sound decisions based on the information you do have. This is why I tend to lean heavily on game theory(ish) types of graphs. The thesis is pretty straightforward: regardless of the tech and architecture, all these technical improvements mean nothing if the user experience/execution (UX) is not 10x or 100x better than what exists today. Therefore, as I weigh various outcomes, I continue to come back to execution (UX) being more important than the user interface (UI) and composability when it comes to adoption. Why? Because you could have the greatest distribution channels in the world, but if your product sucks, no one will use it. This is one of the few times I don’t overweight distribution, as distribution is essentially shots on goal. This is by no means perfect, but again, helps me get my head around the opportunity here. The figures below are revenue and multiple estimates based on hypothetical assumptions on the market share Bullet is able to capture from Jupiter and Drift. If we assume modest success, it is pretty clear that $ZEX’s upside is more than modest. Now long term liquidity will matter a lot, especially as they start to compete with those outside the Solana ecosystem like Hyperliquid. How well can Bullet market make in their vault? How quickly can they list new markets? Are spreads tight and book depth deep on the majors? Can they keep market makers and users after incentives turn off? What about quoting the long-tail? Can their risk engine survive an October 10th type of liquidation cascades? Tokenomics from Zeta Markets are still quite poor, will they change/fix? Lots of questions need to be answered. But zooming out, this all ladders up to the payoff matrix above. Bullet's tech, sequencer flex for zero-fees/RPI (if desired by the team), ZK for unified margin scaling, gives it a real shot at 10-100x better execution than the incumbents on Solana. Disclosure: I hold $ZEX.