The @a1research__ CLOB campaign is still in full swing. But after many weeks of going deep into the rabbit hole, I think it's time to share some learnings and key predictions I have about HL and the broader CLOB / perp DEX wars on a more personal level. Let me kick off with the learnings: ➤ Hyperliquid is still king (most prominently in terms of OI dominance where it's at 57% market share, but also wrt 30d volume, which sits at almost $290B, equalling 33.54% of market share) ➤ Dominance has been decreasing amid rise of competitors, amid which specially @Lighter_xyz, @edgeX_exchange, @tradeparadex, @OfficialApeXdex and @Aster_DEX stand out with multi-billion dollar 30d volumes (yet OI remains comparably low) ➤ High volume coupled with low OI hints at mercenary capital and farming activities, often driven by incentives (more on this in my post here: ➤ As @Louround_ outlines in his post here: it remains to be seen how well those emerging players are able to retain users, liquidity and volumes post TGE/incentives ➤ What's clear though, is that the design space is broadening and we increasingly see a move towards modular L2 appchains, which is largely attributable to the reduced cost + time to market when leveraging pre-built infra components vs building an L1 entirely from scratch (like HL did), as well as the access to base layer liquidity and users that rollups have (less friction vs L1 appchain design) ➤ Even more clear is the fact that CLOBs are a true infra-level disruptor, already commanding >90% of perp DEX volumes in the DeFi space. However, for the long-tail of assets, esp as the onchain asset universe grows through asset tokenization oracle/RFQ-based models are likely to gain more relevance as well (see post by @kaledora here: My opinionated takes & predictions: ➤ The market opportunity for onchain CLOBs is absolutely gigantic and even bigger than the market for centralized CLOBs as it taps into new asset classes and can serve regions/demographics globally that don't have access to centralized venues. ➤ Perps will largely replace memecoins as the main gambling and "make money quickly" kind of play in crypto. Don't get me wrong, memes will still exist, but I expect a lot of activity to shift to perps where you have similar upside but lower risk of being rugged lol. ➤ Perp DEXes will become the onchain financial supercenters of the future, with many of them likely morphing into "all-in-one" financial services platforms that offer trading (perps and spot), borrowing/lending, their own stablecoins and vault product, etc. As composability among apps matters here, it's imo also not a coincidence that many players increasingly follow Hyperliquid's lead to opt for a universally programmable execution environment (in HL's case HyperEVM) around their CLOB infra (see Paradex's DimeVM or @bulletxyz_'s BulletSVM or Monaco building a permissionless CLOB infra layer in @SeiNetwork's universally programmable and composable EVM ecosystem). ➤ Latency and performance matters, yet it's not the ONLY success factor that will be relevant. Highly recommend listening into our recent @a1research__ space here: where we discussed this topic in quite some depth. In short tho: latency and speed matters to institutional traders and is highly significant for the onboarding of big TradFi entities, an angle that @MonacoOnSei with its micro-second execution latency and sub 400ms settlement latency specifically optimizes for. It also matters to market makers as confirmed by @carlosguzman___ from @GSR_io. ➤ Zero-Fee models can work, which @trevor_flipper explains very well in his post here: The prime example is Robinhood and PFOF (pay for orderflow), which in a nutshell means that platforms route user orders to market makers for a fee that the MMs pay for the right to fill, enabling no-fee trades for retail while still monetizing successfully. @tradeparadex and @Lighter_xyz are teams aiming to replicate this onchain. This will imo lead to a compression of fees across the perp DEX market over time and set an important precedent that a large part of the industry will likely follow, yet orderflow curation will remain relevant. ➤ So who will be the winners? The "cake" is imo big enough for more than one player and multiple winners. Imo the winners will be the ones that can successfully find differentiators and USPs to cater to specific use cases and audiences. Zero-fee models for example highly promising to capture the retail masses, whereas latency/performance optimization could serve as an advantage to onboard institutionals. Also verifiable privacy and ultimately the (de)centralization/underlying trust assumptions can be crucial factors. Last but not least, especially in crypto, incentives, but also community and culture, will always be relevant. I think those latter points are where HL truly excels. Ultimately liquidity and network effects arising from real user adoption will be what defines the winners here, so any contender that does well on that front in the mid-long term has good chances. ➤ However, I do also believe that @HyperliquidX will continue to lead the pack on most fronts at least in the mid-term. I think its fully integrated L1 architecture does have some benefits vs fully modular L2s that rely on various external infra components from third-party oracles to DA layers, proving services, etc., which might impact trust assumptions. Generally, the levels of decentralization amid CLOB architectures are often at the center of controversial discussions. Be it around onchain vs. offchain order matching or centralized sequencers on modular L2 appchains vs. the permissioned validator set of HL's L1. While there are good arguments on both sides, I personally believe that Hyperliquid has at least a slight advantage with regards to infrastructure-level maturity. I also think that most players (incl. HL) will progressively strengthen the architectural setup and increase the level of decentralization further as the platforms mature and that early compromises made on decentralization in favor of performance and product are legitimate. And wrt to the latter (product), I rly think HL just has a disproportionally strong position still. HyperCore remains dominant in the sector, HyperEVM ecosystem is growing rapidly, stablecoin recently went live, HIP-3 is coming to mainnet soon, and more exciting stuff is surely in the pipeline already as well. In combination with the unique culture, strong community and the network effects around its traction and liquidity, that just makes for a strong defensive moat. Diving into this rapidly growing sector has been absolutely amazing so far, and one of the most interesting research missions I've embarked on in recent years. Highly appreciative of king @chameleon_jeff and the HL team who pioneered this vertical, and equally excited about all the amazing innovation we're seeing these days with many emerging teams building out entirely new network architectures and/or pioneering new approaches to (micro-)market structure, thereby ushering in what could turn out to be the most significant revolution of financial markets at global scale in decades. Hyperliquid.