L1溢价终结？|炒作、Venice与加密货币的未来
- Jason的2026年“选股者市场”论点基本正在兑现：大多数代币表现不佳，受结构性资金流（ETF/DAT）、基本面收入或强叙事驱动的资产则表现良好。 HYPE、Venice和Zcash是最明显的例子。他预计这一格局将延续至今年下半年，并表示市场参与者要么接受市场奖励的方向，要么逆势做空、承受结果。
- Ceteris梳理了Venice（VVV）的投资逻辑：订阅和API收入目前贡献约6000万美元ARR，周度新增ARR已从约200万美元升至220万–230万美元，再升至250万美元以上。 他认为未来12个月新增至少2亿美元ARR存在可行路径。Eric Vorhees自行出资创办公司，没有股权投资者，也从未围绕代币融资，形成了Ceteris所称的异常干净的结构——但这并不保证价值一定会流向VVV。预计7月起年化代币发行量将降至300万美元，Ceteris预计VVV年末可能转为通缩。
- Panel对L1的清算相当严厉：过去看空价值捕获的观点是对的，只是时间早了。 SOL已从约250美元跌至约85美元；L2BEAT数据显示，追踪的118条L2中有63条低于0.1 TPS，仍有实质活动的只剩Base和Arbitrum；而成功的应用越来越有能力迁移至自己的执行环境。Panel还质疑，如果机构多年来已经可以获得ETH敞口，那么一枚2500亿美元的ETH还能从哪里获得边际买盘。
- 一位Panel成员正在考虑削减比特币敞口、首次增配HYPE：HYPE与其说是加密资产，不如说是一家运行在加密基础设施之上的企业，业务正扩展至股票、指数、大宗商品及其他非加密市场。 与HYPE在约50美元交易时相比，如今已经出现HIP-3、HIP-4、HYPE ETF和DAT策略，Circle向其返还80%–90% USDC收益的安排也已落地，而市场此前预期团队每月抛售量要大得多。一位Panel成员还表示，HYPE最终很可能会引入KYC，届时既有利多也有利空影响。
- Jason表示，Web3的大部分内容已经不再吸引他的注意力：稳定币和DEX已经跑通，但他更偏好AI、机器人、国防及相关领域中规模更大的增长故事。 Jose通常会建议拥有5万–10万美元资金的人配置Mag 7公司，而不是加密资产；Panel则指出，加密市场早期由空投驱动的机会已变得更少、更集中，也更像一场PvP博弈。
- 国防讨论聚焦于无人机和反无人机系统：低成本攻击可以摧毁价值高出2–3个数量级的设备，迫使资金流向无人机、反无人机、激光及相关系统。 据称ACS正在筹集约20亿美元，用于将现有火炮接入追踪系统；一位Panel成员还表示，该团队投资了一家尚未公布名称的激光防御公司。数据中心、机场、船舶及其他高价值基础设施都可能需要这类防护。
- Kevin将Kang从加密转向机器人的过程视为自我重塑的范本：Kang很早就投资了Figure和Skild，后来又参与了约26亿美元估值的Figure轮次，之后该公司估值一度接近400亿美元；他还将这笔仓位注入Robo Strategy，后者目前的交易价格约为NAV的2–4倍。 另一位投资者认为，随着试点项目推进，机器人估值和滞后的私募市场估值都可能被重新定价。Kevin最后将不断扩大的财政赤字和国防支出，与比特币的货币贬值论联系起来。
1. 选股者市场：资金流、基本面或叙事
Jason回顾了Delphi的年度前瞻报告，表示报告的核心判断在半年后基本兑现：2026年不会是大多数代币都能上涨的市场。表现良好的资产，通常具备ETF或DAT带来的结构性资金流、真实业务和收入，或强有力的叙事。Zcash属于隐私叙事标的，Venice则横跨基本面和叙事驱动两类。他预计下半年格局不会明显变化：“市场会告诉你它喜欢什么”，参与者可以接受，也可以逆势做空然后感到失望。
谈到HYPE，Jason重点关注稳定币业务。其收入高度稳定，应该获得较高估值倍数；稳定币收益则能在加密活动和交易量下滑时提供缓冲。若用户撤出稳定币，业务仍会呈周期性，但波动可能较小。他援引“收入上涨25%，代币却上涨48%”这一在市场传播甚广的说法，认为这说明重新定价是合理的。
SEC有关股票的消息似乎也让代币化股票的监管环境更加友好，但Jason表示，他不确定这一变化的边界有多清晰、能持续多久。对永续合约DEX而言，这一点很重要：流动性充足的现货市场有助于维持永续合约与现货价格一致，支撑流动性，并让资金费率保持在合理水平。此前被称为“垃圾股”的Lighter，在流动性热力图显示清算点近在咫尺、此前适合做空之后，似乎也开始重新定价。
2. Ceteris的Venice测算：约6000万美元ARR，以及再增2亿美元的路径
Ceteris表示，重新定价始于约4周前，当时Venice新增订阅层级，并开始披露更细颗粒度的订阅数据。他估算公司当前ARR约为6000万美元，来源包括订阅和API收入，其中API收入可能约为100万美元。更关键的是前瞻收入：3周前新增ARR约200万美元，2周前为220万–230万美元，上周则超过250万美元。
即便按较弱的新增速度年化，再加上API收入，Ceteris估计未来12个月仍有新增至少2亿美元ARR的路径。订阅者披露、代币销毁和代币使用量上升，正帮助投资者把前瞻盈利故事与VVV当前估值联系起来。
一位Panel成员提出价值转移问题：公司存在股权结构，而代币销毁额每月只有约20万美元。Ceteris回应称，销毁的主要作用是证明订阅者数量增长、发出价值转移信号，而不是代表当前已经完成全部价值转移。他承认其中存在典型的二元结构：业务顺利时，代币和股权利益看起来一致；业务恶化时，代币持有者可能被挤压。
Ceteris认为这一结构异常干净，因为Eric Vorhees承担了全部股权侧融资，没有股权投资者，公司也从未围绕代币融资。这排除了部分法律保护机制，也排除了那些只想最大化股权退出回报的利益相关方，但并不构成任何保证。Ceteris认为，Eric最终最好的退出机会可能来自代币而不是IPO，并提到他长期坚持的隐私理念，包括其在ShapeShift和Bitcoin时期的经历。
他预计VVV年化发行量将从7月开始降至每年300万美元，相当于约3%–4%的通胀，并表示年化发行量一直以每月100万美元的速度下降。他预计这一趋势将持续，同时销毁量逐步上升，VVV可能在年末转为通缩。
3. DM：上限明确、剩余投机价值有限的实用型机制
一位Panel成员用收益率和持续经营框架替代了DM的回本周期模型。DM每年提供约365美元的免费推理额度，即每天1美元。其价值取决于现金的机会成本，以及市场为Venice持续经营所给予的溢价。
在5%的机会成本和45%的持续经营溢价下，计算结果是365美元除以0.50，约为730美元；若持续经营溢价为15%，则为365美元除以0.20，约为1800美元。发言者认为后一个水平下DM的估值相对合理，剩余投机空间很小：持有者可以买入、使用推理额度，然后希望以接近原价的价格卖回。
供应量软上限约为38,000枚DM，当前供应量接近38.2K，且长期以来仅在约4%的区间内波动。该机制被描述为一场营销活动和游戏，既为Venice带来关注，也创造了供应沉淀池，但同时也是一项负债：若达到最大使用量，每天可能对应高达3.8万美元收入，扣除利润后或许约为1.9万美元。实际使用量远低于上限。
在约5000万美元市值下，这一机制不需要极大的需求就能支撑价格。结论是，Venice增长所捕获的价值并不在DM；与用户基数增长直接相关的主要代币是VVV。
4. L1价值捕获逻辑已经失效
一位Panel成员表示，自己持有一些SOL ETF敞口，但没有现货SOL或现货ETH，这是他整个加密从业经历中都不记得曾经出现过的组合。去年SOL接近250美元，对应约1250亿–1500亿美元估值，如今约为85美元。问题在于，SOL或ETH当初为什么值得那些价格。
Jason总结称：“所有看空Layer 1价值捕获的人都对了”，只是判断得太早。Tetris Capital是对的，那些在SOL为7美元时就质疑其估值的人也是对的。Panel还指出，应用链逻辑的方向其实正确，但通过Cosmos表达这一逻辑是个错误。
前一天发布的L2BEAT数据显示，追踪的118条L2中有63条每10秒连1笔交易都做不到，即低于0.1 TPS。Panel称，真正仍在做事的只剩Base和Arbitrum；一位发言者表示，这正是他2年前就得出的结论。该发言者认为，L2逻辑对Ethereum和Celestia都已经破产。
解释是，许多所谓的“DApp”本质上只是带代币的普通应用，并不需要自己的L1。真正跑通的应用往往有足够实力迁移至自己的平台或执行环境。一位Panel成员表示，唯一表现良好的区块链，是能够创造收入的区块链，并称Solana的收入已经被Hyperliquid夺走。
一位Panel成员表示，在此前看空TIA后转而做多，是一个重大错误：这笔交易建立在相对ETH的比较之上，而如果ETH本身就被高估，相对比较并不是可靠的投资策略。若被迫二选一，Panel仍然偏好SOL而非ETH，但也承认，过去1年最好的决定其实是两者都不持有。
5. ETH的买盘从哪里来？
Panel指出，机构多年来已经可以通过现货市场、ETF和DAT买入ETH，因此看不到一波等待到来的新增机构资金。有人表示Tom Lee正在放慢买入速度，这说明边际买家正变得不那么激进。
在约2500亿美元的估值下，一位Panel成员表示，至少有20种资产是他更愿意持有的标的。没有已知催化剂足以逆转数年来的一连串失误，于是问题再次出现：买盘从哪里来？同样的逻辑也适用于SOL，只是程度较轻。
竞争不再只发生在加密市场内部。2020年和2021年，加密曾是显而易见的高增长行业；如今资本和注意力也流向AI、存储、半导体、深科技、国防、机器人和生物科技。一位发言者开玩笑说，加密基金都成了存储专家，开始交易SanDisk。加密已经不再自动是市场上最耀眼的机会。
另一位Panel成员表示，他正在考虑削减比特币敞口、增持HYPE，这是他第1次考虑卖出部分比特币。HYPE被视为一家运行在加密基础设施之上的企业，而不是一项加密资产投资。与约1年前相比，如今已经有HIP-3和HIP-4，也有HYPE DAT策略和HYPE ETF；Circle当时还没有向Hyperliquid返还80%–90%的USDC收益，市场也曾预期团队每月抛售约1000万枚HYPE。
如今HYPE的大量活动来自股票、股票指数和大宗商品，同时也来自加密资产。一位Panel成员表示，近期表现最强的3种资产中，只有Zcash是真正的加密资产投资；HYPE是一家交易所业务，Venice则主要是一家借助加密基础设施运行的另类业务。
6. 资本形成，还是“没好到可以上市的东西”？
Panel对代币模式的本质存在分歧。一位发言者认为，代币持有人没有任何股权权利，只能依赖Eric Vorhees的善意。Venice可能无法在公开股票交易所上市，因此代币看起来或许只是让投资者获得一家无法使用传统公开市场的企业敞口。
另一位发言者表示，Venice代币目前相对OpenRouter仅以小幅折价交易，而OpenRouter的利润率更高。发言者认为这一点值得注意，因为这是他首次看到一家做着与非加密公司类似业务的加密公司，相对后者出现有意义的折价，而不是溢价。
反方观点是，代币让普通人也能获得投资机会。普通用户不可能在Hyperliquid还处于早期时买入或空投获得如此大的公司仓位，再利用这笔仓位参与建设HIP-3市场、借贷市场或稳定币基础设施。另一位发言者反驳称，Venice实际上并没有募集资本，而当前传统风险投资市场对做同类业务的公司反而更具流动性。
Jason表示，投资者应区分哲学偏好与市场机会。有些加密业务可以去中心化、原生于加密；另一些则可以把加密基础设施与传统业务结合起来。他认为Hyperliquid相较传统交易所结构透明得多，即便它并不高度去中心化；对于整个行业而言，出现一个挑战Binance主导地位的新竞争者是积极的。
一位Panel成员认为，去中心化应当让产品更好，或让监管机构更难将其关闭，而不应只是价值观打卡。Panel认为Hyperliquid最终引入KYC的可能性相当高。看空逻辑是，平台会失去匿名用户或“鱼”用户；看多逻辑是，合规友好的访问方式将带来目前因VPN或监管限制而无法进入的用户。
Jason更广泛的结论是，Web3的大部分内容已经不再吸引他的注意力。他仍认为稳定币和DEX是重大成功，也认为Zcash和MetaDAO是有意思的加密原生实验，但相比创造新的0到1原语，他更偏好扩张已经验证的产品，例如永续合约、稳定币以及面向现实世界资产的开放API。
7. 5万–10万美元投资组合的配置建议
Jose表示，除非投资者确实热爱投资，否则他会偏好类似Mag 7的组合，例如Amazon和Google。他认为这些公司相对能够抵御AI冲击：在乐观情形下，它们可以表现良好；在悲观情形下，其交易估值约为10多倍或20倍出头市盈率，未必会遭到重创。
Jose不确定，对于5万–10万美元的投资组合而言，加密市场是否还像过去那样提供有吸引力的机会。他承认加密市场仍会产生重大错价，例如早期的VVV机会，但公开市场也有普通投资者能够参与的交易。他认为，当AI用户已经能看到瓶颈出现时，存储交易相对容易理解，只是现在价格更贵、参与难度也更高。
Ceteris表示，当大量新项目和空投同时启动时，小资金组合在加密市场尤其有吸引力。富有创造力的收益耕作可以带来显著的非线性回报，但如今新项目更少。一位Panel成员补充称，留下来的参与者更加成熟，使市场更接近PvP博弈：上涨会被卖出，价格也会出现更多震荡。
Panel回顾了DeFi Summer：当时人们把资金投入智能合约、耕作Yams，却没有遭遇项目方跑路，因而觉得自己像天才一样。那个时期10–50倍回报和空投很常见。近期机会则更加集中：从12月低点计算，VVV上涨2–3倍已经是很大的结果；一位发言者认为，即使Zcash涨到2000美元也很困难，尽管市场上有人在讨论10000美元。
Ceteris表示，获得私募市场机会在一定程度上是能力问题，但Panel指出，大多数人并非全职投资者，也没有数百万美元资金，最终可能进入糟糕的SPV，甚至遭遇项目方跑路。Jason的回应是，这类人可能也不应该盲目追入代币。
对于一个有冲劲、没有资本的20多岁年轻人，Yan建议参考Delphi的方法：学习、写作、分享观点，建立声誉和人脉，逐步找到投资岗位、创始人岗位或早期公司。核心不是等待捷径，而是靠自己赚取进入机会。
8. 无人机、反无人机系统与新的国防经济学
一位Panel成员指出了国防支出的两大顺风。第1，世界正变得更加多极化，因此各个地区都希望增加自身防务投入、发展本土公司，而不是依赖潜在对手。第2，乌克兰战争表明，无人机可以摧毁价值高出攻击系统2–3个数量级的装备。
从社会秩序角度看，占主导地位的军事技术会塑造制度：骑士支撑了封建制度，火药降低了对数十年骑士训练的需求，而无人机再次降低了攻击成本，并扩大了潜在攻击者的范围。
一位Panel成员表示，国防经济学已经从昂贵武器拦截昂贵武器发生逆转。低成本无人机蜂群让人难以证明用价值数百万美元的导弹去击落1000美元目标是合理的；同时，威胁的近距离和便携性也让探测更加困难。因此，现有防御体系需要补充或重新设计。
ACS据称正在筹集约20亿美元，但发言者表示他们并不是该公司的投资者。其系统将现有的.50口径武器、30mm火炮和机枪接入追踪技术，从而可以利用现有弹药打击无人机。另一位Panel成员表示，该团队投资了一家尚未公布名称的激光防御公司。
更好的激光可能提供更干净的防御工具：如果光束经过调校，能够摧毁无人机并在穿透目标后迅速消散，那么在人口密集地区部署会更容易。发言者也指出，激光自身存在风险，例如失手后可能继续射入太空，但技术进步和单位经济性改善仍然重要。
需求场景不止军事基地。数据中心、机场、货船及其他高价值基础设施都可能需要反无人机系统；一位发言者表示，未来船舶可能会把无人机或防御系统作为标准化运营配置。由于支出被定义为保护而非攻击，政治上也更容易获得支持。
Panel将重新武装比作大规模停电后购买发电机：一旦发生代价高昂的故障，继续处于无防护状态就是不理性的。防御方还必须为3-sigma级别的蜂群攻击做准备，而不是只防御单架无人机，因为攻击方会围绕成功概率进行优化。
9. Kang的机器人转型与比特币收尾
Kevin将Kang从加密转向机器人视为自我重塑的案例。Kang在约2017年进入市场，THORChain和Synthetix是其早期成功投资之一，后来还通过3AC完成了一笔大额交易。他也通过Mechanism建立了出色的风险投资记录。
Kang很早就意识到软件成本会趋近于零，因此将机器人视为更具防御性的AI机会。他极早期投资了Figure和Skild；发言者估计，Figure早期估值约为2000万美元，随后Kang又参与了约26亿美元估值的轮次，而几个月后Figure的估值已接近400亿美元。这笔仓位后来注入Robo Strategy，在讨论时已经变得具备流动性，交易价格约为NAV的2–4倍。
这一投资被描述为一次从加密到机器人的跨界押注，也是发言者见过的最强投资和交易之一。Kang不仅是交易员，更是很早就识别出市场将离开加密、并持续押注机器人的投资者。
Robo Strategy的另一位投资者认为，对NAV折价的批评忽略了两点。随着新模型出现、机器人进入试点项目，机器人估值可能需要重新定价，而部分私募估值仍然滞后。此外，该载体并非只是持有现有资产：它可以发行股份、投资新交易，并利用自身品牌、渠道、筛选能力和行业知识，寻找升值速度可能超过NAV溢价的公司。
更广泛的启示是，优秀投资者可以学习一个全新的行业，并反复完成自我重塑。Panel引用了那句格言：“事实变了，我就改变看法”（“When the facts change, I change my mind”），并指出，当ETH叙事重新回归时，Kang没有被重新吸引回ETH。
Kevin最后将国防与更广泛的宏观论点联系起来。国防支出正在上升，其中很大一部分由政府融资，对于获得订单的公司而言实际上不会造成稀释；欧洲盟友正被推动重新武装；政府支出超过收入，财政赤字仍在扩大。
随后他回到比特币的长期走势图。尽管市场存在波动和回撤，他表示自己每周任何一天都会选择这张长期走势图，并称其BTC论点依然和过去一样强，因为货币贬值以及上述支出动力仍在。最后的玩笑是：BTC正在上涨——只是这轮上涨发生在私募市场。
完整逐字稿
How’s everybody feeling about this market?
I mean, Jason is the big HYPE bull.
1. ETH, SOL & the collapse of the L1 premium thesis
Yeah. I mean, this is something we wrote about in our year-ahead report, and we’re currently in the process of revisiting it, scoring our predictions, and revisiting what we thought would happen so far. One of the big themes was that we thought 2026 would be a stock-pickers’ market in crypto. It would not be a market in which most coins did well, and you’d have to anchor to 1 of 3 things, or a combination: structural flows, maybe coming from ETFs or DATs; fundamental businesses and revenues that drive value to the token; or a really strong narrative-type play.
A Zcash privacy narrative is an example of that, while Venice straddles the two between something fundamental and something narrative-driven. As I’ve been going back and reading that massive report, that was one of the things that stuck out to me. It’s 6 months into the year, and that’s kind of how it’s played out. I honestly don’t really see that changing very much as we move into the back half of the year.
I think the market tells you what it likes, and it’s your job as a participant to either accept that or fade it and probably be upset.
Yeah, no, the HYPE move was a big one.
Venice is up way more, right? Venice has been the main story these last couple of weeks.
Yeah, no, and just to go back to HYPE real quick, I think the stablecoin component was a big one. You want to think about what multiple to value that revenue at, and I don’t really have a great answer, but it’s highly durable. It should get a pretty hefty multiple.
In particular, you see a lot of cyclicality in crypto, and HYPE is no different. As activity wanes, volumes come down and it reprices lower. I think the stablecoin element adds a nice buffer. Not to say it won’t suffer from cyclicality, in the sense that people will withdraw stablecoins and there will be less USDC on the platform, so it’ll be generating less yield, but I think that happens to a smaller degree.
It makes sense for HYPE to reprice off that. You saw the viral tweet going around—the messaging was, “Revenues are up 25%, and the coin is up 48,” and so on and so forth. It certainly made sense to have that repricing.
Then you had the big stock news from the SEC. I don’t really know exactly how defined or permanent that is, but the idea was basically that the environment for tokenized equities would be much more favorable.
The reason that’s important for a perp DEX is that you need a liquid spot market to help keep basis relatively aligned. You don’t want to have a big divergence between perp and spot. The size of the perp market is going to be a function of the size of a healthy spot market, because that allows you to create liquidity and keep funding in line, which is needed for good UX on that front.
2. Venice AI (VVV), AI tokens & crypto business models
I think those pieces of news helped quite a bit for HYPE, and Lighter, I think, is one of those that’s repricing as well. It’s been a dog for a while, but it seems to be repricing now. I think you had a situation where it was just a very comfortable short for a lot of people.
If you look at the liquidity heat map, there are some big liquidations in striking distance, depending on whether or not they add liquidity to their positions. It’ll be interesting to see what happens there.
Yeah, no, Venice has been the big one. We’ve been chatting about it for some time. The recent information that’s come out—not news, but information—started about 4 weeks ago, when they released more subscription tiers and started providing granular subscription data.
That gives you a view of the revenue ramp going forward, and that’s kind of the bet you’re making. It’s useful to understand the trailing revenue, but there’s no great way to do it. I estimated it at around $60 million in ARR at the moment, as a combination of subscription and API revenue.
The forward earnings are what will really drive interest in the token. The reason people are getting excited is that they’ve been adding revenue consistently. Three weeks ago, they added about $2 million in ARR. Two weeks ago, it was $2.2 million or $2.3 million, and last week it was $2.5 million and change.
Even annualizing the weaker of the two and combining API revenue, which I think is probably around $1 million, with subscription revenue right now in terms of revenue added, I’m guessing there’s a path to adding at least $200 million in ARR over the next 12 months. If you think about where the token trades now relative to forward earnings, I think there’s a lot of room for upside there.
I’m happy to provide some bear cases, but I think the source of the repricing is that they’ve been publishing more information on subscriber numbers, and token usage has really gone up as well. People are trying to put all of that together.
Are you fully expecting the token to get all the value? I know that’s something a few people have brought up, and it’s a good point. There’s the equity structure, which is something we’ve talked about a lot. Why is this one different? They’re only burning, I think, around $200,000 a month. What do you think about that?
Yeah, for sure. The burn has been more of a way to show how many users they’re adding. They’re disclosing new subscription numbers through the burn. It’s not really meant to be the full value transfer; it’s more of an indication of subscriber growth and a signal to the market that they are sending value to the token.
As you mentioned, you always have this dichotomy. It’s all fine and dandy when things are going well, but when things hit the fan, you realize who’s actually at the top of the totem pole, and tokens get the squeeze.
I think what’s unique here is that you don’t have competing interests in the 2 vehicles. Eric funded the whole thing on the equity side. There are no equity investors, and they never raised into the token. The setup is there for it to be—ultimately, it’s up to him in that case—and it’s one of the cleaner setups for something like that.
You don’t have legal protections and self-interested parties that are just trying to maximize their exit value when things go south. I think that cushions the risk a bit.
If I’m thinking from his perspective, I think his exit opportunity is probably best through the token. Maybe this can get acquired. I don’t know about going public, but obviously, that’s a much tougher road. He can always just pocket all the cash flows and not really send any value to the token. That can be a sneaky way to do it.
But I do think he’s been in the space for a while and is principled. He’s building this because he believes in the ability to have private inference and privacy in general. That’s been part of his ethos for a while, going back to ShapeShift and part of the reason he got into Bitcoin all those years ago.
I think the setup is there for a clean flow of value to the token. Right now, there are no guarantees on that front. At this stage, no company should really be sending anything too meaningful back to holders. They should be reinvesting it, and I think that’s what they’re doing now.
I anticipate that the token itself will become deflationary, probably by the end of the year, from reductions in issuance, which he’s continually doing. They’re basically going to be down to $3 million a year of VVV issuance starting in July. Call it 3% to 4% inflation, but he’s been lowering it by $1 million a month.
They’re not issuing $1 million a month, but the annualized rate has been going down by $1 million a month. I anticipate that will continue, and then the burns will gradually creep higher.
Yeah, they seem to really be leaning into the token. Obviously, it has the utility of staking for access to the platform, minting DM, and all of that. Reading the tea leaves, it seems probable, but it would be great if at some point he came out and definitively said it. I imagine he hasn't for a reason.
You also have the DM component, which is a fun thought exercise in terms of what it should be worth. I used to think of it as a payback period, but I decided that's not really the correct framework. I broke it down into 3 elements. There's the yield you get, which is $365 a year, effectively $1 a day of free inference.
The other 2 inputs are the opportunity cost of that cash had you just put it in Treasuries, and the third element is the going concern of the business. It's not about repayment. It's about the chances that I can buy this, use the free inference, and then sell it back at a reasonably similar price. It's assessing what you think the market views as the going concern of Venice.
If the going concern is very high, the formula is basically 365 in the numerator divided by the opportunity cost plus the risk premium of the going concern. If you think there's a big going-concern premium, call it 45%, then 365 divided by 0.5, and you're saying the fair value is 730.
But if you're fairly confident in their ability to continue—and I think as the business grows, it's more and more likely—the going-concern premium comes down. For me, I think it's relatively fairly valued now. It's all the way at the bottom, basically, where the going-concern premium is 15%. So you say 365 divided by 0.05 plus 0.15—365 divided by 0.2—which means the fair price is 1,800.
I think there's no speculation left in the token. It's more about people buying it, using the free inference, and then selling it back for relatively the same price that they bought it for. If you scroll a bit lower, there's a chart.
But I mean, what's the reason to even have this token model? Why not just build for inference normally? Does all inference basically require DM?
No, no, no, no. There's basically a soft limit of around 38,000 DM. It's a formulaic thing where, after a certain supply, the amount you issue against your VVV becomes very small, to the point where it isn't even worth it. Supply now is around 38.2K, and it's really not going to change from here. Supply is relatively fixed; it's moved within a 4% band for an extended period of time.
It's basically a good marketing exercise and a fun game. It brought attention to the platform when it started early on, and it created a supply sink for Venice. It's ultimately a liability for them, because they have to offer up a maximum of $38K worth of revenue per day to these holders. Obviously, their costs are going to depend on their margins, so it's not a $38K liability. It's probably more like a $19K liability.
People can hold it and get $1 of inference a day, assuming it's maximally used, which it never is. You can see in the other charts where they show the amount that's staked and the amount that's used. It's more of a fun thing, but I don't really think there's much speculative juice left in this.
You don't need an insane amount of users to rationalize the value and keep the price where it is. It's not like the market cap of this thing is around $50M, so it's not a massive liability or market cap that needs to be supported with bids. I think it'll probably sit around this price. But it's not really what accrues value as Venice grows. It seems a bit gimmicky, but I guess some people find it fun.
So the real way to play it is just through VVV?
Yeah. VVV is what appreciates as the user base grows.
Are any of the Base AI things interesting, or do you think it's just people chasing after VVV?
There are some I'm still digging through and trying to understand. They're quite a bit more complex. This is one of them that I'm trying to understand a bit more deeply. There are a lot of moving parts to this one, and it isn't as straightforward because of what they're building versus what Venice is building. I just haven't spent as much time looking into it, and it takes a bit longer to fully grasp.
Just talking about the market in general, it seems that right now—obviously, who knows how things change in the future—people are very interested in Venice, Zcash, and HYPE. Those seem to be the ones that have been doing the best. Bitcoin, ETH, and SOL have been pretty bad, especially ETH and SOL.
I own a bit of a SOL ETF, but I don't own any spot SOL or spot ETH. I don't know if that's ever been true for me—not to own spot of either of those. Basically the entire time I've been in crypto, I've always had at least one. It's like, did we just get it wrong in the past about how we were valuing stuff? We put so much money into ETH and SOL.
I'm still quite bullish on Solana from a technical perspective and everything, but Hyperliquid really was bad for the L1 token premium. You're seeing stuff like Venice now, which makes money too. Zcash is a bit different; it's a privacy narrative, so that's a pure store-of-value thing. But you're not getting a lot of those types of tokens doing well. It's basically just Zcash.
Solana was close to $250 last year, and it's at around $85 now. When you look at that $250, it's like, yeah, that was $125B or $150B. That's a big number. Why should it actually be worth that? Why should ETH be worth $250B or whatever it is? We've had this discussion forever, and we don't need to go back into it again.
Alt seasons that we've had in the past were: you get the majors moving, and then you get the other things after. ETH and SOL have just been bad since 10/10. Same with Bitcoin. But now you're getting these other tokens. Obviously, HYPE has been strong for a while, but Zcash and Venice have been stronger more recently, along with a couple of other things. I don't know. It is a pretty interesting market.
All the Layer 1 value-accrual bears were right. Tetris Capital was right. They were just early.
Yeah. Too early.
Way too early.
Same with people who were grave-dancing SOL at $7 for the same reason. It was all right. Crypto's brutal, man, because a lot of those people were right. I think it went back to $8.
Yeah. It's just like ETH is still a $250B asset. There are so many things I would rather own at that market cap than ETH. It's actually crazy.
No, it's tough.
I don't know. I think the old L1 value-accrual bears were right. The app-chain thesis that I think we were all pretty bullish on at Delphi—unfortunately, we chose to try to express a lot of it on the building side through Cosmos, which was a mistake. But I think Hyperliquid was always the thesis: the killer apps would just be exported onto their own chains that were optimized for that use case. I think it's definitely played out with Hyperliquid so far.
Even if Solana gets, like, a Bullet [?] or one of these perp DEXs that does really well, is that really going to be reflected in SOL's price?
I mean, it's already $50B. It's already worth a lot.
Yeah.
So that's the thing.
Yeah. The bet was that there would be a wide surface area of a variety of apps that would launch. It wasn't going to be just this super-powerful, exchanges-only type of thing. At least that was the initial bet on why L1s would have a premium: the use and activity that would happen there, with all of that being in one place.
But it turns out a lot of DApps don't make sense, and some do. The ones that do end up having all the power to move off to their own platform or execution environment.
And a lot of them—DApps aren't DApps. They're just apps with a token, and those don't really need an L1 in any real way, like VVV, right?
The thing is, yeah. We don't even have that many L2s. L2BEAT released something yesterday, and it was 63 out of the 118 L2s that they track doing less than 1 transaction every 10 seconds. That's less than 0.1 TPS. The only 2 that are still really doing anything are Base and Arbitrum.
What's crazy is that I wrote about this 2 years ago now—how Base and Arbitrum were basically the only L2s doing anything. It's crazy how literally nothing has changed there.
And so the L2 thesis completely blew up, both for Ethereum and for Celestia. I think that's the main thing that has hurt so much: the L2 thesis just completely blowing up. They have a lot of assets on it now, but again, you go back to the revenue thing.
And then Solana seems like everyone hates the Solana Foundation these days, but it's still tough. They're in a tough spot. What's funny is that when Solana was SOL versus ETH two years ago, everyone started talking about rev, which I always thought was correct, and I still do. That's the correct way that you should be valuing this stuff.
3. Why Hyperliquid changed the crypto investment landscape
Solana lost all its rev now, and Hyperliquid is the one making money. I don't know—it feels like we've been debating this, but it's been enough time now that I don't think you can really debate that revenue for L1s is important. The only blockchain doing well is one that gets revenue, and the other ones just have not.
Yeah, I think the SOL people, including us, always focused on the relative comp versus ETH, which I think was the right way to play the narrative game. It still looks okay on a relative comp versus ETH. It looks much worse than it did at the peak because rev has been destroyed, but it still looks good.
The problem is, what if ETH itself is overvalued?
Exactly. By a factor of 10 or whatever.
Yeah, I don't know.
It was the same thing, though. My biggest mistake over the last couple of years was longing TIA after I had been so bearish on it. Honestly, I was just trying to be a hero and play both sides of it.
What are you doing? I forgot about TIA.
My bull case for TIA was honestly the relative comp versus ETH, right? That's not a good investment strategy. I still think that if you have to own either ETH or SOL, I'd rather own SOL here. But again, if you're deciding to own one of the two, obviously owning neither has been the right choice over the past year or so.
Historically, I've always felt like the L1 tokens, especially the big ones that actually have users, were kind of your safe hold, and then you could dabble in other stuff.
Yeah, I guess my question is, it's not like institutions haven't been able to buy ETH for years, right? It's not some question of where the bid is going to come from for ETH.
Well, even Tom Lee is slowing down purchases now, too. That's your marginal buyer.
You know, $10 billion. Yeah, of course.
For ETH at $250 billion, to Jose's point, there are probably at least 20 assets that I would rather own right now. Institutions have been able to buy it, so it's not like they're going to rush in to buy. There's no massive ETH catalyst that I'm aware of that's going to change the game and undo a couple of years of missteps. So where does the bid come from? To a lesser extent, I would also apply that same argument to SOL.
I think that's a big part of the story, too. SOL versus ETH—owning neither and owning other things has been a lot more profitable because if you think back to 2020 and 2021, even before the bear market in 2022, crypto was the thing from a macro perspective. It was the high-growth industry everyone was getting into, so the relative-value trades maybe made sense.
This was before the ChatGPT moment, before AI, and before the latest deep-tech and hardware renaissance. If you're coming into crypto, where is that marginal or incremental bid coming from? If you're coming in, you're getting more concentrated in things that have either real fundamentals or a real thesis.
If you're not, you're trading. How many crypto funds or crypto investors do we know that overnight became memory experts and now trade companies like SanDisk? AI stocks have become the bellwether, where everyone's attention is. You're not just competing for capital within crypto anymore. You're competing for attention across a lot of exciting, high-growth, potentially massive industries that have pretty attractive valuations, depending on your thesis about what the future looks like in 5 years.
Crypto isn't the only game in town, or the most exciting or shiniest thing out there, like it was in prior cycles.
No, I absolutely agree. One thing I wanted to mention is that if you're coming into crypto as an institutional investor at this point, you could have bought Bitcoin or ETH for years, either spot on Coinbase, through ETFs, through DATs—however you wanted to get exposure, you could have. The same is true for Solana over the last year and a half or so.
But if you're coming in now, what is there? I've actually thought for the first time that I might want to decrease my Bitcoin exposure and increase my HYPE exposure. This is the only time I've ever even considered selling some Bitcoin.
I don't view HYPE as a crypto play. It just happens to be a business that's in crypto and has a token, but clearly what they're going for is not a crypto business. A year ago, when HYPE was trading at around $50, HIP-3 didn't exist, HIP-4 didn't exist, and there wasn't a Hyperliquid DAT strategies company bidding HYPE.
HYPE ETFs didn't exist. Circle wasn't giving them 80% or 90% of USDC yield on Hyperliquid. People assumed there would be 10 million HYPE tokens sold by the team for unlocks every month. All of those things were part of the situation last year when HYPE was at this value.
Today, it's fundamentally a business deriving a lot of its activity from things that are not crypto: trading equities, trading equity indices, trading equity perpetuals, and trading commodities. A lot of institutions coming in and a lot of people looking at the market will also come to that conclusion. Thinking about it as a crypto company is missing a lot of the upside.
I think you're completely right. Of the 3 assets that are doing well right now, only Zcash is a true crypto play. HYPE is essentially the exchange play that you just talked about. Venice isn't really crypto. They have the token, but it's really just an alternative.
Yeah, they have crypto rails, which are useful for agentic payments because you assume the bulk of inference will be from agents going forward. Having that setup helps, but I agree: their primary user base is going to come from outside of crypto. It's just a good bootstrapping mechanism, which I think is—
Which is kind of like—I don't know, that's fine. Building good businesses on crypto rails is fine, right? You don't have to be a fully crypto-native company.
I mean, it should be, right?
Is it? I don't know. The whole point was that otherwise it's just a way to go public without going public.
Well, think about Venice. You actually have no rights. If you owned equity in one of the neoclouds or whatever, you would have some rights. Here, you're just relying on the good faith of Eric Vorhees, which is worth a lot, clearly.
I mean, I know it's at a discount.
It's a bit depressing to me.
It trades at a very small discount. At least based on your analysis, OpenRouter has better margins and it trades at—Venice still trades at a pretty small discount to it. I think it's pretty—
Would Venice actually be able to get listed on a public stock exchange?
Probably not.
And so that's actually where the crypto angle is useful, right? The product is mostly just putting out the token.
So it's just things that aren't good enough to go public.
No, it's not just because it's a different type of instrument. If you're going public, you're IPOing. That's an equity issuance.
4. Crypto capital formation vs traditional equity markets
I hear what you're saying, but it goes back to one of the original killer crypto use cases—and one you've always been bullish on—which is capital formation, right? Is this not a good example of that? I know it's still early, and there are things that can be improved. The fact that token-holder rights aren't equity rights is something we can table, because I think we all agree that eventually needs to be solved. But from a capital-formation standpoint, this seems exciting.
What capital formation is happening here, in your view? They didn't raise money, right? And if they raise money, I think the traditional equity markets—not even the IPO ones, just traditional venture—are way more liquid than crypto. Right now, it's the first time since I've been in crypto that there's a meaningful discount for a company in crypto that's doing the same thing as a company outside of crypto. You actually have a discount with a token, whereas historically it was always the opposite. You have liquidity.
Yeah, you have liquidity. But that's the whole securities law issue, right? It's meant to stop this, because these things are actually just securities, which I totally think those laws shouldn't be there either. I think you should be able to list more easily and transfer things much more easily. But it's not really what got me excited about the space. I don't like any of this stuff, really, personally.
Yeah. That's different from seeing the opportunities in front of you.
Yeah. I don't know. I look at things like this: there are certain aspects of crypto that I philosophically align with, like Bitcoin, clearly, and even things about Ethereum with decentralization. Unfortunately, being in the space for years, you realize how little people actually care about it, so you get jaded.
But there are certain businesses that could just be crypto-native and decentralized, and have all the crypto-ethos aspects. There can also be businesses that aren't, and that mix aspects of the two. As a market participant in crypto, I feel like if you let what you wish was the case stop you from investing in a business that's—and I'm pretty sure you own HYPE, so this doesn't apply to you—but if—
I do indeed.
If you let your philosophical ideal stop you from buying clearly the best business that's come out of crypto in a while, for example, that's just a weird thing. I do think HYPE does a lot of good for the space. I don't think Binance having a stranglehold over crypto trading and exchanges is necessarily a good thing.
I think having a new player come in and challenge that, especially from a more open perspective, is positive. It's clearly not decentralized like a lot of other perps have been in the past, but it's also significantly more transparent than the incumbent. It's clearly moving the industry in a better direction in that sense.
I mean, decentralization is—just to cut you off—but it's supposed to make your product better, right?
No, you're fine.
It's supposed to make your product better, right? Decentralization makes your product worse from a UX perspective, but the benefit is that you're harder for regulators to take down and all this stuff. So, with Hyperliquid not being decentralized, they can get hit from these other angles.
If you think not being decentralized is what's going to cause regulators to completely shut them down, and that's why you're not buying, then that's fair. You've been wrong so far, but that's fair. But if you're thinking, “The ethos is that they're not decentralized, so I'm not going to buy it,” that's where you go wrong, because you don't just want to make everything decentralized for the sake of it. What's the benefit you're getting from this decentralization? That's what you need to answer.
People have tested this in multiple directions. Solana tested it more than Ethereum, and then Hyperliquid tested it more than Solana. You saw CME and NYSE come out the other day saying they want some action to take place on Hyperliquid.
I do think it's pretty likely that Hyperliquid eventually has KYC, but I also don't think that's necessarily bearish for them. You can look at it both ways. The bear case is that more institutions will feel comfortable trading there, but you lose all the fish. The bull case is that there are a lot of people who don't trade on Hyperliquid because VPNs block them, because of compliance reasons, and all this stuff. If Hyperliquid were regulated, maybe all those people would start trading there.
Yeah, the other element we're ignoring is the democratization of access to these platforms at a really early stage, right? There's no world where everyone could have bought an asset like Hyperliquid, or even been airdropped for using it, and then bought this asset at such a low price and decided, “Okay, I actually have a huge stack. I'm very invested. I'm going to go build on top.”
You create all these incentives for people to partner and continue to build, whether it's HIP-3, lending markets, or stablecoin setups. I think that stuff doesn't happen without a token, and that in and of itself is why crypto makes something like this possible.
Yeah, I would distinguish Hyperliquid from something like VVV. I think Hyperliquid is clearly not super decentralized right now, but it is built on crypto rails in a way that something like VVV just isn't. There is potentially a pathway for it to become more decentralized, and it has been moving in that direction over time.
I agree there's an opportunity set in front of you. For me, when I came across crypto early on, I thought this was the biggest thing I'd ever come across—the biggest narrative—and there were all these dreams that we had for it. I think in many ways we exceeded those: stablecoins and DEXs have crushed it.
But there are many other things I sort of don't believe in anymore, kind of like Kyle's tweet. A lot of Web3 I just don't believe in anymore. If I have scarce attention units, I'm always interested in the biggest story. I like the biggest-upside story.
That's what crypto taught me, too: if you're in the industry that's growing the fastest, you just have more opportunities, and even if you screw up, you kind of win anyway. You're punished way less for losing. I just don't see that as crypto right now.
None of these good token stories, like VVV, changed my mind on it, because I don't feel like it's really crypto. I would rather be in some of these other markets. I do think there are interesting opportunities within crypto, but now it's more about deploying things that we already know work—whether it's perps, stablecoins, or open APIs for RWAs—and scaling them, rather than these very interesting 0-to-1 primitives.
Other than things like Zcash and MetaDAO, which are super-interesting crypto-native experiments, there are definitely a few of them out there. But I don't think it's the most interesting story right now, personally.
Yeah. I think you kind of created this Venn diagram where it needs to satisfy all of these individual beliefs, desires, and all this stuff. Through that lens, sure, nothing fits the filter. But I—
Jose's hierarchy of needs.
Yeah.
No, if we go—
Yeah. This bar is too high.
No, I just think there are sectors that are growing at such an absurd pace, and people are building such legitimately, insanely interesting stuff.
Yeah, for sure. People don't have access to all of that stuff, and that's kind of part of the point: this allows for much broader access to these assets.
Yeah, you want to buy robotics, you have to go buy Kang's FOF strategy.
And honestly, I think a lot of people talk about how it's trading so much above NAV, but those are private marks, right? I don't know what the right premium over NAV for it to trade at is, but it's obviously not 1-to-1, considering those marks are from raises a year ago, and people are pretty bullish on robotics.
But to your point, with robotics, you can probably do well in robotics over the next few years even if it turns out to be like the car industry in 10 years, right? As you're going through this, you can kind of be wrong and get away with it. That's definitely true in crypto. You definitely can't be wrong anymore.
Robotics, AI generally—inference, infrastructure, defense, even biotech—is having a bunch of—
But yeah, so much of this is private markets, right?
Yeah, for sure. It's all private.
Here's a question for you, Jose. If you didn't have access to private markets, would your opinion change on where it's easiest to do well?
I think if I didn't have access to private markets, I would be doing the poor man's Aschenbrenner and just trading AI.
I think memory was a relatively obvious trade. A lot of people talked about it. If you were on Twitter using Claude, it was very clear that this was a bottleneck that was going to happen. I think people underestimated AI in public markets for ages, and I think they still do.
Even though semis have gotten really hot, I think some of the other bottlenecks always have a way to go further ahead on that. Personally—and again, there are different ways to play the game, right?—you can try to find softer games that maybe have less upside but where you have more of an edge, or you can always try to play in the biggest game.
The way I'm built and the way I like to invest is that I always want to play in the biggest game. Whatever the biggest trend is that I see happening, I want to be as close to it as possible. So I think I'd probably be doing that.
But realistically, everyone can have access to private markets. It's a little bit of a skill issue, right? If you really wanted to put in the time, you could have access to private markets. If you're not doing this full-time, that's the other thing.
Yeah, for sure. If you're not doing it full-time, you—
And that's most people. So—
Yeah, but you probably shouldn't be buying tokens either, man. You probably shouldn't be aping into tokens either.
Well, if you have somebody you trust giving you information, because that's what most of this stuff is, it's like—
It's like a friend chat with one dude who's full-time in crypto shilling his enormous—
For sure.
But then they have access, and so—
For perspective, a family office[?].
Dude, from my perspective, I just feel like I don't have a ton of time to go looking into private markets. I'm trying to run the research business at Delphi, and naturally, yes, I can be looking at tokens all the time because that's kind of what my job is.
There are obviously other people in tons of different areas. Now, if you're just some unemployed trader—
Mm-hmm.
Then sure, maybe. But going into a bunch of private markets as someone unemployed could work, but it could also—
Yeah, I don't know.
Be pretty bad.
I feel like a lot of people probably don't have the money to effectively invest in private markets. Most people don't have millions of dollars.
Also, private markets historically have done worse than just buying the index fund. There are obviously a lot of good private investments now, but again, it's about selection and—
So, let's say you are someone who goes into private markets and then you're buying all these SPVs that end up getting rugged too. There's so much that you need to—
I agree.
Like—
Yeah, and it's also about quality, right? The highest-quality deals are still very much an access game, especially—
If you had a younger cousin, Jose, who had, I don't know, $50,000 or $100,000 saved up and wanted to get into the markets, which markets would you tell them to get involved with? Would crypto still be there, given a smaller portfolio, which probably applies to the majority of people in retail?
No, I don't. Unless you love investing, I'd probably just tell them to buy a Mag 7 kind of thing—Amazon, Google, these kinds of companies—that I think are pretty safe from AI.
In a bull case, they do pretty well, maybe not as well as the CoreWeave and memory stocks and stuff like that, but I think they do pretty well. In a bear case, they're trading at teens or low-20s earnings, so I think you don't get absolutely crushed in those names.
And then if you're— I mean, yeah, it's hard. It's tough. I do think public markets are more interesting than ever, and that's the competition for crypto. I don't know that crypto offers that interesting an opportunity set for $50,000 to $100,000.
There's still alpha because these things get so mispriced, like Yan just showed with VVV, calling that super early. It's clearly a fundamental story, but I'm not sure that I think there aren't people crushing it in public markets too, in trades that are not that hard.
The memory trade wasn't that hard. Now it's maybe harder, and things are more expensive, but yeah.
I think the $50,000-to-$100,000 portfolio size was very attractive in crypto when there were a lot of new things launching because of the airdrop component.
That lumpiness was pretty meaningful, especially if you got creative and really optimized on the farming front, but—
For sure, that—
There just aren't as many launches happening. It definitely hurts.
Crypto is a much harder game to play, and you also have the issue that the only people who survive are obviously going to be better than the average user. It becomes more PvP. They're also the ones who tend to sell, so rallies get sold off and there's just a lot of chop. But when risk returns, they do really well.
I think it's definitely hard for a new person to get in because they're not going to be conditioned the right way to survive right now.
Yeah, I think it's really hard to—yeah, I think it's hard.
And it's true. A lot of the best-performing coins are hype. What are you looking for here?
A 2–3x on VVV was a pretty big multiple if you got in right at the pico bottom in December, but right now, it's more reasonable. Same with Zcash. I know people are throwing out $10,000, but realistically, it's going to be tough, in my opinion, to get to $2,000 or something.
These are all the strongest-performing assets, right? Whereas one meme coin did a 3x recently or something. In the past, you had 10–50xs everywhere, plus the airdrops.
To Jose's earlier point, it was hard to screw up if you were just long. So many people, including myself, thought we were geniuses during DeFi Summer in 2020. But what was I doing? I was just throwing money around in different smart contracts and farming Yams, and somehow I didn't get rugged.
Maybe that Based coin was the dumbest, but there was just a lot of stuff like that back then. Maybe it'll come back. It's just definitely a lot more concentrated.
Yeah.
Yeah. It was way more fun when you could just farm Yams and make money.
5. Advice for young investors entering markets today
What's the equivalent right now? I guess the sectors or whatever that are most mispriced—if you're a smart, hungry 20-something—
In crypto.
Starting from a low base, I think in crypto in general—
Yeah, or in general. I know, Yan, you're super interested in this, and we're spending a lot of time on the defense side, which I think is a pretty interesting sector. Obviously, I don't know if we need to get into that on the Hive Mind right now, but there are opportunities like that always.
Yeah, I do think there are opportunities like that always.
But it depends. Do we caveat it with whether you're full-time in investing or a casual participant? What would you do if you had a full-time job where you did well, but it wasn't investing? I think then I'd just do—
I'd have to—you'd have to just do the public market, big—
Big index or big names, I think.
I don't know. I don't think you're going to do well casually in any market right now. There's not a—
Unless you're in Yan's family office.
I just don't think it's going to be easy. I more so mean if you're a really hungry 20-something who's willing to dedicate some time, wants to really make it, and doesn't want to be part of the permanent underclass.
I do think defense is one of the most interesting places to mine for opportunities right now. AI is obviously super interesting, but it's very fully priced and kind of hard to break into from scratch.
But in that sense, are you kind of trying to pitch yourself to a fund to get hired? That's because you don't necessarily have the capital base to just invest on longer horizons.
I think you do the Delphi playbook, right? You just start learning, writing, trying to have alpha, sharing it with people, gradually building your reputation, gradually building your network, and finding whatever the right opportunity is.
Eventually, you want to be in an investing seat. That's the way to go—or a founder seat, if you have the skill set, or joining an early-stage company. But you kind of have to be in the investor seat first so you understand what company to join and what makes the most sense.
I think you probably just do the Delphi playbook. It's still the way to make it in whatever sector. But yeah, if you don't have money, for sure, if you don't have money—
So basically, work hard.
Grind.
Yeah, work hard. Exactly. Yeah.
I feel like defense is one area that someone grinding and working hard as a regular person can't really get into on the private side. So much of this stuff is heavily influenced by government interests, right? National defense is different from just investing in an AI company. So, yeah, no, I agree.
Bio is one I know very little about, but we've spoken to a lot of very smart people who have been super excited about bio over the last few years in terms of price and where it's going. But, yeah, Kevin, I don't know. Do you want to get us back on track? Talk about some layoffs or Cerebras?
No, I think this is exactly what I wanted to talk about because I agree. I think that comes back to the competition for attention, capital, and liquidity. Now the public markets are exciting again, and I'm pulling up charts like this where you look at—I mean, this is just the iShares U.S. Aerospace & Defense ETF. There are much better select plays or ways to get exposure, but you have semiconductors and memory stocks.
To your point, there's so much happening. It's almost these exponential tech trends coming to a head at once. If anything, the hard part of the game is figuring out what the signal is from the noise because there are just so many things that, over the last 6 to 12 months, seem to be vying for your attention if you consider us all growth investors, which I think we all are.
There's so much vying for attention right now.
For sure. Yeah, I'm long this index, actually, and I do think defense is super interesting because you always want to be in sectors where there are big tailwinds, obviously. With defense, there's this same thesis that got people into crypto, or at least is a bull case for crypto. One of my bull cases for crypto is just that the world is becoming more multipolar.
I think it's almost consensus now that the U.S. isn't going to be the police force of the world. Everyone has to spend on their own defense, and you're seeing Anduril-like companies in every geography. Every geography will want to have its own defense companies and not be buying from potential adversaries.
So there's that tailwind of everyone both spending more and wanting to spend more on companies in their geography. Then there's the second tailwind, which is that I think drones—the Ukraine war has created this paradigm shift in defense. I'd love for Yan to jump in because I think he's even deeper on this than I am.
In simple terms, you can blow up billions or hundreds of millions of dollars' worth of equipment with 2 or 3 orders of magnitude less cost. A lot of this equipment becomes very difficult to use, and the game just moves to drone and counter-drone warfare. It's almost like a reset, right? A lot of the investments that countries have already made in this kind of stuff become much more redundant.
I think a lot of the spending has to shift to drones, counter-drones, and a bunch of the ancillary technologies around them. You have this other tailwind of spending shifting, which I think is pretty huge. It's almost like a reset.
One of the founders I spoke to put it this way: He has this theory that the social order is dictated by the dominant military order. You had knights and feudalism, and the whole system was centered around knights. Then you invented gunpowder, and you no longer needed 20 years to train and equip a knight. Any peasant could grab a gun and blow his head off, you know?
Drones, to some extent, are shifting this balance again by lowering the cost of defense—and of just being annoying—by so much. But, yeah, I'm curious to hear what you think, Yan.
Yeah, it's like lowering the cost of attack. Right now, you no longer need the most cutting-edge tech to be a real nuisance to a world power, right? That widens the surface area of potential attackers.
Both in terms of cost and proximity. They no longer need to set up a very obvious missile system. It's something much more portable and unnoticeable. And so what that means is that a lot of existing defense doesn't really work for it, right? You need to rearm in a different way.
The economics of this are also very flipped. Before, it was an expensive thing shooting down an expensive thing, but if all your defenses are in the form of expensive missiles, using them against cheap tech is much harder, especially because of the swarm element. You kind of need to rethink it.
You're seeing quite a bit of that, and you're seeing counter-drone tech in a variety of ways.
A big one is ACS. I think they're raising 2 billion now, and we're not investors in that, but they've built this cool system where they allow existing guns—50-caliber, 30 mm, and machine guns—to connect to the system. The system is a tracking system, so you can use these guns to shoot down drones.
You're basically using existing ammunition and all that on that front. Another big area has been lasers, and that's one we're definitely excited about and have invested in. It's not announced yet, so I'll refrain from sharing names, but basically, it's using lasers to shoot down drones.
Previously, it was difficult because these lasers are pretty fragile, and you're still using a very high-kilowatt laser to shoot down something that's pretty minimal. Lasers also, if you miss, continue into space. They can hit airplanes. I think they're less of an issue relative to bullets, which eventually have to land somewhere.
But I think as the laser tech improves—which it has—it provides a really clean way to defend. Spending on defense is very palatable, right? We're not spending on missiles to blow things up. We're spending on defense to protect.
There isn't really going to be a slowdown in this spending because you always want to shore up more. As we're building out data centers, and even just existing airports and all this stuff, everything is going to need defense.
I think we move to a world where just data centers—like ships, right? Like big cargo ships—I just think a lot more of it is going to be normalized. You'll have some drones on board to do basic defense, or counter-drone systems, depending on where it ends up.
Yeah, counter-drone systems. As that tech gets better, the unit economics continue to improve, and the accuracy improves. Once the accuracy improves, you can also broaden the number of places where you can set these things up.
With lasers, you can design the beam so that when it hits the drone, it's hitting it with just enough force to destroy it, but it begins to dissipate beyond the drone into a harmless light. That means you can set it up in more densely populated areas.
I think there's going to be a lot of rearmament, and none of the existing defense really applies. You'll see this stuff get bolted onto existing larger missile systems. That will help when these things are reloading, or simply as a way to avoid firing something that costs $1 million at something that costs $1,000.
Broadly, these things need to be set up everywhere, and they will be because you don't really have the negative headline of, “This is an attack thing.” No, it's purely defense, so it becomes really easy. Once you have one of these events, everyone rearms massively.
The Cold War was an example, but even something as simple as when an area has a big blackout, everyone buys generators because it's a relatively cheap way to protect against something like that. You would be a fool to run into the same issue again.
If you're thinking about the value of what you're defending, in your home you have electricity. That's obviously not a huge value beyond your family and the convenience of not living without power. But when you're defending incredibly valuable systems like airports or data centers, where any form of outage can be catastrophic or catastrophically expensive, then you can rationalize a pretty massive amount of spending.
You also need to overspend because you're not defending against someone sending 1 drone. You need to defend against the 3-sigma event. If I'm going to attack, I'm going to try to increase the probability of my attack being successful. So you need to defend against a pretty sizable swarm of drones or missiles, or whatever it is.
So, I think there’s going to be a huge buildout on that front.
6. Robotics, Figure AI & the next major investment wave
And maybe—yeah, I don’t know, Kevin, if you want to—where you want to take this. I was thinking of just doing a shout-out to Kang at Robo Strategy. I don’t know if people are aware of how crazy that trade was that they did, or we can talk about some of the AI topics that you had—Cerebras or situational awareness. What do you think? No, do the Robo Strategy one for those who aren’t. Yeah.
Yeah. Robo Strategy is crazy because, for those who don’t know Kang—who entered crypto in the last few years—Kang’s one of the most successful crypto traders, probably of all time, or he’s probably up there. Obviously, we don’t know about a lot of the Bitcoin whales and stuff, but he entered pretty much at the same time as us, around 2017. I think THORChain was both of our first wins.
Yeah, Synthetix. Then 3AC was the one in size, I guess, that got us going. I guess he was early in Synthetix, too. He just absolutely crushed it as a trader and, obviously, with venture through Mechanism, too. Then he got into AI very early and saw robotics as the real defensible area in AI, because he understood early that the cost of software was going to go to 0 and made this bet on robotics.
He ended up doing Figure and Skild extremely early on. I think he did it at $20 million or something like that. I think he did the 2.6 billion round before it got marked up to like 40 a few months later. That’s according to his tweet, at $2.6 billion, like 15x-ing that. That’s obviously an illiquid position. He contributed it to this Robo Strategy fund, right? Now it’s liquid, and it’s at—I mean, I don’t know what it is right now, but sort of 2x to 4x, something like that, marked up to NAV.
The crazy thing is, I think he’s just getting started, but it’s one of the best investments and trades of all time: this parlay of Kang from crypto into robotics. Very impressive to see. I do think people have thought he’s a trader, which he definitely is, but I do think he was extremely bullish on crypto, and being long was the only way to really make money early. I think he just realized what a lot of people are realizing now, earlier, and I think he’s very long robotics. He’s one of the people that I really think can build something like that.
Yeah, generational there.
I’m also a BOT investor. I think the FUD now is basically where it trades relative to NAV. A couple of points there: 1, I think robotics is due for a repricing. I think inference stole a lot of the limelight.
If you think about how much inference and broader cutting-edge tech has repriced since these rounds, the valuations before kind of seem pretty trivial now. At the time, they carried a lot of headline risk, but now—again, I’m an investor in BOT; I also invested in Apptronik—and I do think those are due for repricing. There was always going to be this kind of lull in the trade because the market—
—moved away, and revenue was a bit—okay, so you have stuff that’s already generating a bunch of revenue, and you could see it continuing to scale without slowing down. At the same time—
—didn’t. So everything—and, you know, for the first time, money chased revenue-generating companies instead of hype. I do think, as attention shifts back here, at Figure’s $40 billion and Apptronik at $5 billion, I’m personally more bullish on Apptronik at the relative valuation differences.
I do think they’re going to see repricing as these new models come out. The timelines varied, but I think it was always late 2016, early 2017 when you start to see these robots in pilot programs with some of the initial customer bases. As people see that, they’ll start doing some moon math and extrapolating, and the unit economics of these are pretty impressive. I do think you’ll see repricing there, and part of the NAV, I think, is basically stale, is what I’m saying.
The other element is that it’s natural to assume there’s a premium because this isn’t a vehicle that bought these assets and just sits there, right? If it was one that just sat there and was a holding company for a variety of these, then you would value it closer to NAV, with some liquidity premium on top.
But I think what these guys are doing is they’re going to be issuing shares and investing more. I do think they’re going to get into great deals because of the brand and their ability to help with future rounds. I think they understand the space very well, so there’s going to be good selection and good taste outside of just good access.
That’s effectively the premium you’re paying, and the idea that what they invest in will be able to reprice. You’re selling a share for—you know, call it $1 million worth of shares in the open market—to invest that $1 million into something that you think will appreciate. What the NAV is implying is their ability to turn a market-cap premium into quality investments that appreciate at a bigger rate than the spread over the NAV.
And to me, this is an example of how—what Kang did—you can just reinvent yourself and learn a completely new industry and try to become one of the leading names in that industry. I think the best investors—and I think crypto is going to have some generational investors—are going to just do that over and over again in their careers.
That’s what I was trying to say earlier. I do think, whatever the quote is, “When the facts change, I change my mind. What do you do?” I think that’s what I would say to a lot of crypto people that are still doing crypto.
Kang did a really good job when he switched timing—
—on everything.
It was early—
Very early. And also not getting sucked back into ETH. I think a lot of people in crypto got sucked back into ETH because Tom Lee came back last year. He stuck to his guns on all of that, too, and it’s all played out well for him.
Yeah. Nice. What are you thinking, Kevin? You want to end it here?
7. Bitcoin, macro tailwinds & long-term positioning
Yeah. Yeah. I was going to tie it back, when we were talking drones and defense, to run it all full circle. I was looking at an industry where you’ve got—I think some of the best ones right now are ones where you have massive spending explosions, and a lot of that is government-financed, right? So it’s non-dilutive financing for a bunch of these companies.
Just looking at national defense spending in the U.S., that’s only going up and to the right. This all ties full circle to our big macro thesis and even what Bitcoin’s long-term value proposition was with currency debasement: defense spending is going up, government expenditures are going up, and government revenues aren’t matching that. Deficits are getting bigger. The U.S. is pulling back and forcing people like European allies to bolster their own defense and spend a lot more on that. This world of fiscal deficits, and those expanding, is not going anywhere.
How I’d like to wrap it is with the long-term chart of Bitcoin. When you really zoom out and look past the day-to-day volatility, it’s not that bad-looking of a chart. I would take this chart every day of the week, right? Even the short-term pullbacks and some of the volatility, every now and then I think it’s important to look at these types of charts and just think about the long term.
My thesis around BTC is still as strong as it has ever been because of all the currency debasement and all the spending we’ve been talking about.
I agree. Yeah, I think BTC will probably pump at some point soon, and then everything pumps with it and people get back in. All it takes is people starting to make money again in crypto.
BTC is pumping. It’s just private.
Private BTC is pumping.
Hopefully.
But yeah, we’ve—I think it was a really good conversation. We definitely had some topics we didn’t get to, but that’s what this is for, right? Just to air out what’s most top of mind for us. We’ll have plenty of juice to talk about next time. Appreciate you guys hopping on, joining, and sharing your thoughts.
It was good. Thanks, everyone.
Thank you.
Thanks, guys.