D1 Capital Shuts Hedge Fund, Flags AI Energy Play in Siemens Energy

D1 Capital关闭对冲基金，看好西门子能源受益于AI用电需求 创始人Dan Sundheim正将D1 Capital的重心完全转向私募市场，原因在于规模上限削弱了多空策略回报；同时其押注西门子能源，以利用AI推理驱动的结构性电力供应缺口。 在本次讨论中，Dan Sundheim分享了他对日常作息、资本配置以及D1 Capital横跨公开市场股票和私募风投的投资策略的见解。他解释了看重欧洲市场的原因，详细阐述了公司对人才的评估标准，并剖析了对Clean Harbors和西门子能源等长期精选个股的投资逻辑。此外，Sundheim还反思了风险管理在心理层面的真实状况、GameStop事件后的做空环境，以及私营企业在无需面对公开市场压力下进行扩张所具备的独特运营优势。 鉴于公开市场问题重重且波动性日益加剧，像Stripe这样成功的私营企业应避免上市。 (0:00) D1 Capital采用基于基本面的自下而上分析策略，投资视野为3至5年，高度依赖基本面指标，而非量化指标或短期技术形态。 (3:48) Dan Sundheim职业生涯中最大的单类失误就是过早卖出了像Costco这样具备长期复利效应的标的。 (6:04) 评估公开股票市场人才所需的时间远长于评估工程师，通常需要大约五年才能衡量一名分析师的能力，且需要三年时间培训私募股权出身的员工掌握公开选股能力。 (10:52) 2021年的GameStop轧空事件凸显了由社交媒体推动的市场新现实，迫使做空者大幅降低仓位规模，以防范散户轧空带来的系统性风险。 (24:27)

Dan Sundheim of D1 Capital argues that successful investing relies on bottom-up fundamental underwriting and long-term multiple expansion, but structural market shifts—including retail-driven volatility, fund scale constraints, and distorted public valuations—are challenging the traditional public hedge fund model. Scale Disadvantages and Public Versus Private Frictions (3:48) D1 Capital manages approximately $25 billion, split roughly two-thirds in private holdings and one-third ($10 billion) in public equities. Sundheim distinguishes the two asset classes by characterizing private investing as a 'one-way door' and public investing as a 'two-way door.' While illiquidity in private markets provides the structural benefit of preventing investors from prematurely selling their highest-conviction compounders, managing large pools of public capital introduces steep scale disadvantages. Because returns negatively correlate with assets under management on the short side—where managing $30 billion makes shorting a $5 billion company without moving its price nearly impossible—D1 decided to close its public-private hedge fund. Sundheim contends that top-tier private businesses such as Stripe benefit from avoiding initial public offerings. In public markets, retail enthusiasm and algorithmic trading can sharply inflate multiples, pulling future returns forward and forcing companies to grant employee stock options at unsustainable peaks. Conversely, staying private allows companies to conduct orderly periodic tender offers that mirror real intrinsic value growth, avoiding the compensation distortions caused by public market swings. Underwriting Discipline, Accountability, and Holding Winners (18:07) D1 targets an approximate three-year double (equating to roughly a 20% plus annual IRR) through fundamental, bottom-up analysis over a three-to-five-year horizon, completely avoiding quantitative models or short-term technical indicators. Sundheim identifies 'selling the Costcos too early' as the single most expensive error in an investor's career. Investors frequently underestimate the compounding durability of high-quality businesses and exit prematurely after initial multiple expansion, as occurred when D1 sold Netflix early following an analyst departure despite correctly underwriting its global flywheel. To execute this strategy, D1 avoids lateral hires from competing public funds, choosing instead to recruit from private equity and spending approximately three years teaching them public market analysis. To eliminate hindsight bias, every analyst must maintain a weekly 'mock portfolio.' Analysts pre-register their exact sizing and capital allocation across their covered names, creating an auditable track record that directly dictates year-end compensation regardless of whether the primary fund executed the trade. Post-GameStop Risk Architecture and Short Sizing (24:27) The 2021 GameStop short squeeze demonstrated how coordinated retail momentum, government stimulus, and social media sentiment could drive heavily shorted equities up 400% in weeks irrespective of underlying economics. While D1 never faced margin calls, the fund suffered steep losses before covering at unfavorable valuations to protect capital. Sundheim and the discussion emphasize that true risk management must occur before a crisis; actions taken after an unexpected shock represent capital destruction rather than risk control. In response to these market dynamics, the fund restructured its short book, expanding from roughly eight concentrated short positions to a diversified portfolio of around 40 smaller positions. Positions are sized so the firm can absorb an irrational 10x to 20x price surge without forced liquidation. The strategy targets companies facing secular terminal value risk, persistent market-share losses, or cyclical over-earning, while strictly avoiding options and derivatives due to the friction of implied volatility and unpredictable timing. Industrial Capital Constraints and Regional Pricing Lags (41:40) Sundheim capitalizes on regional market inefficiencies, noting that European markets exhibit a 'laggier voting machine' than the rapid repricing seen in the United States. For example, Rolls-Royce’s operational turnaround under CEO Tufan Erginbilgic was slow to be recognized by European institutions despite the predictability of its aerospace aftermarket cash flows. Similarly, Siemens Energy trades at half the enterprise value of GE Vernova despite identical revenue bases of approximately $45 billion, matching product portfolios across turbines and grid equipment, and equal exposure to electrical demand growth. This industrial focus aligns with Sundheim's thesis on the global energy bottleneck. After two decades of 0% electricity demand growth in the United States, Sundheim forecasts electricity consumption will expand at 4% annually over the next 20 years, driven by artificial intelligence training, sovereign inference infrastructure, and decentralized grid integration. Because turbine manufacturers like Siemens Energy remain conservative and refuse to rapidly add capacity, suppliers with strong moats—alongside regulatory monopolies like Clean Harbors in domestic hazardous waste incineration—enjoy durable multi-year backlogs. Sundheim illustrates how institutional equity investing requires managing structural tensions between short-term liquidity and multi-year fundamental compounding. While large funds face severe capacity constraints in public short selling and must navigate erratic sentiment, disciplined position sizing, systematic analyst accountability, and identifying mispriced industrial monopolies remain core drivers of long-term outperformance.

D1 Capital创始人Dan Sundheim阐述了其基于基本面的多空投资体系，指出长期超额收益的核心在于识别卓越企业的复利能力并抵制过早卖出的诱惑；同时他透露，由于管理规模扩张严重制约了空头策略的有效性，D1决定在年底前关闭其对冲基金，进一步将战略重心转向一级市场。 投研决策流程与过早抛售赢家的教训 (2:29) D1 Capital坚持自下而上的基本面研究，持有期一般为3至5年，通常以三年翻倍（约20%以上的年化内部收益率）为目标。在交易执行上，Sundheim透露团队往往在正式投资备忘录完成前就已开始建仓，因为在瞬息万变的公开市场中，等待一个半月完成备忘录可能错失理想价格；若分析师在撰写过程中发现致命缺陷，则会立即平仓止损。 Sundheim回顾其职业生涯时坦言，投资中最惨痛的教训并非在糟糕标的上亏损，而是过早卖出类似Costco这样的长期复利型企业。以2018年建仓Netflix为例，当时市场极度担忧其高额内容支出，但D1准确洞察到该模式借助高收益债撬动全球内容飞轮并推高边际利润的逻辑。然而，因主要行业分析师离职及管理层注意力转移，D1过早离场，错失了巨大的后续涨幅。他指出，由于极难精准预测三年后的退出估值乘数，投资者极易低估优秀企业的长期复利价值。 做空框架重塑与事前风险管理的代价 (24:27) 2021年初的GameStop散户抱团挤仓事件对D1造成了严峻考验。在散户补贴、居家隔离及社交媒体情绪共振下，部分基本面缺乏支撑的股票在两周内飙升400%。尽管D1并未收到追加保证金通知（Margin Call），但在估值极度非理性的极端市场环境下，为了确保基金存续与客户资金安全，公司被迫在最不利的价位平仓止损。Sundheim对此反思道，真正的风险管理必须在危机发生前完成；事后被迫采取的防御性平仓并不是风险管理，而是资本的直接毁灭。 经历危机并暂停做空一年后，团队彻底重构了做空框架：将过去集中持有约8只空头标的的模式，分散为约40只小头寸组合，单笔头寸规模严格控制在即使股价出现10至20倍非理性暴涨也不会威胁基金生存的水平。出乎团队意料的是，这一分散化策略产生的空头超额收益并不逊于过去的集中时代。在做空标的选择上，团队更倾向于寻找具有长期终局价值风险（Secular Terminal Risk）的企业，并避开规模低于10亿至20亿美元的小盘欺诈公司，同时不使用期权等衍生品以避免承担额外的时间与杠杆风险。 私募背景人才选拔与“模拟组合”纠偏 (10:52) 在人才选拔方面，D1几乎从不从其他二级市场基金横向招募投资经理，因为既有习惯难以改变；相反，团队几乎全数招募具备顶尖建模与财务分析底子的私募股权（PE）分析师，并平均投入三年时间训练其适应公开市场的投资直觉。Sundheim认为评估一名股票分析师至少需要五年，成功的核心在于对市场的持续敏锐度以及类似查理·芒格所称的先天“商业嗅觉”。 为消除后视镜偏差（即分析师到年底往往只记住成功推荐而遗忘失败案例），D1要求每位分析师每周必须维护一套“模拟组合”，强制其对其覆盖领域的资金分配进行前置登记。这一模拟业绩直接挂钩年终考核与薪酬，用以客观检验分析师的选股能力是否被基金有效转化。而在实际基金运作中，Sundheim本人直接主导了基金95%以上的投资决策与仓位控制。 公开市场机制扭曲与对冲基金规模约束 (84:46) 尽管数字时代信息高度透明，但公开市场的波动性反而显著加剧，定价效率出现倒退。Sundheim以Stripe为例分析指出，优秀科技企业选择长期保持私有化是极其合理的。在公开市场中，极端估值会将企业未来数年的价值一次性透支，迫使管理层在高位发放股票期权或限制性股票（RSU），这实质上是“掠夺未来员工的价值来补贴当前员工”，导致薪酬机制脱离企业真实内在价值；而在私有状态下，企业可通过定期的要约收购（Tender Offer）实现更合理的价值分配。 与此同时，D1宣布将在年底前关闭旗下对冲基金业务，其管理的约250亿美元资产目前已有约三分之二配置于一级市场。Sundheim解释称，对冲基金的收益率与管理规模呈显著负相关，尤其体现在做空领域。在300亿美元的总规模压力下，基金几乎无法在不引起市场剧烈波动的前提下做空市值在50亿美元左右的中盘股，导致传统多空策略在极端规模下难以为继。 能源基建机遇、区域定价滞后与中国市场 (59:24) 在长期资产配置逻辑上，Sundheim避开技术迭代过快的科技行业，转向具有极高护城河的公用与重工业基础设施。例如，持有危废焚烧核心资产的Clean Harbors因严苛监管与邻避效应（NIMBY）拥有极强的垄断护城河，并直接受益于美国制造业回流。对于AI浪潮，他预测美国电力需求将从过去20年的0%增速提升至未来20年的年均4%，而燃气轮机设备商因历史上行业周期的重创而在产能扩张上极为保守，这使具备涡轮机与电网改造双重优势的西门子能源（Siemens Energy）处于极佳竞争位置。 Sundheim特别指出欧洲与美国市场在定价效率上的显著差异：西门子能源与通用电气电气化业务（GE Vernova）营收相仿（均约450亿美元）且产品线一致，但后者的企业价值却是前者的两倍，反映出欧洲市场因宏观增长悲观情绪而存在严重的定价滞后。谈及亚洲市场时，Sundheim表示D1约在三年前已彻底停止对中国市场的投资，主因在于行政干预与政策突变（如课外辅导禁令）打破了市场对法治程序的预期，甚至导致部分互联网企业为了避免监管打击而刻意不敢超出盈利预期，破坏了股权投资的风险收益对称性。 Dan Sundheim的实践揭示了现代大资金管理者在基本面超额收益与市场结构扭曲之间的权衡。随着公开市场在流动性冲击下波动加剧，叠加庞大资产规模对灵活做空策略构成的物理上限，单纯依赖二级市场多空博弈已难以实现最优资本复利，这也最终倒逼其将资本重心全面转向一级市场与长期现实资产。

[0:00] Dan Sundheim: if I ran a private company like Stripe, I wouldn't go public. [2:29] Dan Sundheim: In the public markets, you know, if you have a good idea, you could take a month and a half to like write a memo, but by that time, the price may have moved [3:48] Dan Sundheim: private is a one-way door and public is a two-way door. [33:02] Gabe Plotkin: risk management after the fact is a bad idea [33:02] Gabe Plotkin: I didn't know a 500 mile per hour hurricane existed [33:50] Gabe Plotkin: I don't know when a stock is going to work [54:07] Dan Sundheim: time is your friend [55:20] Dan Sundheim: I like to think of myself as being a little more academic than vibe trading [59:24] Dan Sundheim: Tech just changes too quickly [59:24] Dan Sundheim: predicting things beyond three years is incredibly difficult [80:23] Elon Musk: just get the cost down, get the cost down, get the cost down [84:46] returns are negatively correlated with size [86:24] robs future employees to pay current employees [95:41] The market will be your mentor.

[0:00] I don't think I've ever had a Guinness. [0:01] Unlike Scott Wu, [0:03] I've had many, many, many beers, perhaps thousands, [0:06] but I'm not sure I've ever had a Guinness. [0:08] Without an alarm clock, I will wake up [0:10] almost exactly [0:11] between 2:55 and 3 a.m. almost every day. [0:13] Wow. So you wake up ahead of earnings. [0:16] No, no. Forget earnings. [0:18] Every day. [0:19] You just want to see what happens at the open. [0:20] No, I don't want to wake up. [0:22] I actually don't want to wake up. [0:24] I mean, if I ran a private company like Stripe, [0:27] I wouldn't go public. [0:28] I think the public markets— [0:31] It's ironic because you're a public markets investor. [0:32] Yeah. [0:33] I think the public markets [0:34] are kind of problematic at this point. [0:37] Should we try to split the G? [0:39] Oh my God. No. [0:40] We cover this in an episode with an actual Irishman. [0:42] It's a completely—it's an invasive species. [0:44] It's a made up TikTok thing. [0:45] Yeah, but how did that start? [0:46] But you can. [0:47] You're a crass American so you can, [0:48] and maybe we should get it on camera, [0:49] and that would be funny. [0:50] We really trying to do this? Split the G? [0:53] Only Mike is. [0:54] Only Mike is, unless you... [0:57] Not close. [0:57] Oh! [0:58] That's not bad. That's not bad. [1:00] Can we zoom in here, please? [1:02] Camera roll three. [1:05] Dan Sundheim is one of the world's [1:06] top hedge fund managers. [1:07] He runs D1 Capital, which invests in public [1:09] and private companies across a bunch [1:10] of totally different industries. [1:12] Dan's one of the smartest investors I know. [1:13] So Daniel Gross and I try to get into his head [1:15] and figure out how he analyzes companies. [1:18] Cheers! Cheers. [1:21] Okay, Dan and I were wondering, [1:22] what trading app do you actually use? [1:24] Like, if you are trading underneath the table [1:26] during the interview, like, are you in Interactive Brokers? [1:29] Are you WhatsApping someone who executes this? [1:32] I'm just emailing. Okay. [1:34] I'm emailing the trader with that, yeah. [1:35] I mean, I'm checking Bloomberg all day long. [1:37] Yeah. Yeah. That's the app on my phone. [1:38] Yes. But I'm just emailing, [1:41] texting my traders. [1:42] And if the trader's like sleeping or something, [1:44] is there an app that you use? [1:46] Or is— If he's sleeping [1:47] during the day... [1:51] We haven't encountered that problem. [1:52] That'd be a different problem, [1:53] but if he's sleeping at night, [1:55] we actually have overnight traders in Asia. [1:57] So, once you told me sometimes when you struggle [2:00] to sleep at night, there's nothing better to do [2:01] than wake up and start trading Asia. [2:04] So, is that your East Coast trader [2:05] or do you have another person? [2:07] Yeah, we have another person that we outsource. [2:11] I see, yeah. I see. [2:13] And then how, what fraction of AUM [2:15] is just Dan Sundheim trades? [2:18] Or Dan Sundheim decisions? [2:19] 95% plus. [2:22] Oh, so all of it? [2:23] Yeah. Okay. [2:24] I mean, people have trading authority, [2:26] but most of the trades, I put in myself. [2:29] Okay, so memos come up to you [2:31] and then you ultimately size them [2:34] and decide whether to do them or not or something like that? [2:37] Yeah, I mean, you know, [2:39] the process starts way before the memo. [2:42] I'm in dialogue with the team about the idea, [2:44] why they like it. [2:46] You know, we're having conversations [2:48] way before it gets to the memo stage. [2:49] But then we get to the memo stage, [2:51] and usually we start buying it before, [2:53] because in the public markets, you know, [2:56] if you have a good idea, [2:57] you could take a month and a half to like write a memo, [2:59] but by that time, the price may have moved. [3:02] So usually we start buying it before the memo's done, [3:05] and then the memo is kind of the final, you know, [3:08] compilation of the due diligence. [3:10] And that's when the best thing that could happen [3:14] is we're buying it and then the stock keeps going down. [3:17] So by the time you have the memo [3:18] and have even the most conviction, you can buy more. [3:21] A lot of times it's going the wrong way. [3:23] But what fraction of the time, like, [3:25] do you finish writing the memo and you're like, [3:26] "Oh no, this is a terrible company. [3:29] This is not good." [3:31] Yeah, yeah. No, like, that doesn't happen [3:32] because usually they'll start writing the memo [3:33] and then they'll come to me and be like, [3:35] "I think I made a mistake," and we'll sell it. [3:37] Like rarely have I—I don't think I've ever [3:40] owned a position because we had discussed it, [3:42] then read the memo and been like, [3:44] "Oh my God, what are we doing here?" [3:48] That'd be a bad sign for the analysts. [3:49] We jumped right off the deep end. [3:51] Describe what D1 does. AUM, strategy, for the uninitiated. [3:55] Yeah. So we invest in public and private companies. [3:58] We do fundamental analysis. [4:01] And so, deep research, trying to understand business models, [4:05] trying to understand company prospects, [4:07] choosing the right management team, [4:09] and whether it's public or private, [4:11] we're investing with the horizon of three to five years. [4:16] And we apply the same due diligence process [4:19] to both public and private. [4:21] Obviously private is a one-way door [4:24] and public is a two-way door. [4:25] So, you know, that's different. [4:27] Is private actually a one-way door, [4:28] even as just the private markets have matured [4:32] and there's so much more secondary activity [4:33] and things like that? [4:34] Like, do you still have to treat it as a one-way door? [4:36] Definitely. [4:37] I mean, you can sell like, [4:38] the best companies you can sell easily. [4:40] But those aren't the companies you want to sell. [4:42] So, you know, we rarely transact in the secondary market [4:47] because we don't want to sell the best companies. [4:50] And it's hard, like, people think you can just transact [4:54] in the secondary market but if you want to sell $50 million, [4:57] $100 million, people have to get information rights [5:00] from the company and then it's like... [5:01] And it's viewed as a signal if you're like large— [5:03] Exactly, they go to the management team and say, [5:05] "I want to sell," and the person doesn't follow through. [5:07] It's not great. [5:09] So, the AUM is about $25 billion, about two-thirds private, [5:14] $10 million public. [5:16] Public is long-short, you know, all bottom up. [5:21] No, you know, no quant. [5:23] Really kind of the same thing people were doing [5:24] in terms of stock picking 30 years ago. [5:27] And when you say no quant, that is, you're contrasting [5:30] to some of the more short-term technical firms [5:34] that are just trading based on what'll happen the next day, [5:36] the next week, the next month? [5:38] Exactly. [5:39] We're not using any computer programs to guide our trading. [5:44] You know, we are not trading based on quarters. [5:47] You could be doing my job [5:50] the same way 20 years ago. [5:51] You wouldn't have exactly the same tools, [5:53] but it's just, you know. [5:54] Do you think of any of those signals [5:56] that people talk about when you think of when [5:57] to enter or exit a trade? [5:58] Like, the stock would be oversold or overbought? [6:02] No. Okay. [6:04] What I find interesting about you guys [6:05] is you're long term in that you're not, [6:08] like you were saying, some of the quant guys [6:11] who are just trying to predict what happens [6:13] in the next, you know, what's the next buyer doing [6:15] or the high frequency guys or something like that. [6:17] You're instead just thinking, like, [6:19] is this company selling good product? [6:21] Will their sales exceed, you know, what people expect, [6:23] and will that add up to a good business? [6:25] And so, the work is evaluating future earnings [6:28] of the company, but there are some people [6:31] who are really buy and hold. [6:32] Like Costco seems to attract a lot of shareholders [6:35] who think there is no price [6:37] that is too expensive for Costco, [6:39] and there's kind of a huge loyalty there. [6:41] Whereas, you can have a price on both ends where, [6:44] at this price, I am an enthusiastic buyer [6:47] and at this price, you know, God bless 'em. [6:49] I'm selling the entire position. [6:50] And you kind of have that price boundary at both ends. [6:53] Yeah, practically we may say, [6:56] "We think that in three years this company [6:58] is going to be worth double." [7:00] Right? So kind of roughly 20-something percent IRR. [7:03] But if we're right about the company, [7:05] usually it doesn't go 24%, 24%, 24%. [7:08] People will pull forward the IRR. [7:10] The stock may go up 50% in the first year [7:13] if things go well, then the forward IRR [7:16] all the sudden looks a lot, you know, worse [7:19] and you're moving capital to the next opportunity. [7:23] I think the biggest mistakes, if we would look back, [7:25] I mean, it's easy to think about the mistakes [7:28] where you lost money, and you think about those a lot, [7:30] but the biggest mistakes are selling the Costcos too early. [7:33] Because, you know, the IRR is totally dependent upon [7:37] what you assume the exit multiple is. [7:40] And that is difficult to be precise about, [7:42] what the right exit multiple is for a business. [7:45] So let's talk about painful mistakes. [7:47] Like, can we talk about Netflix or something like that? [7:49] Yeah, so Netflix, when we started the firm, [7:53] almost every LP I met with, [7:55] they always want to hear a stock pitch. [7:56] So I pitched them Netflix, [7:58] and actually when we were interviewing candidates, [8:02] what we did was we gave every single interviewee [8:04] the same case study. [8:06] We said—this is in 2018. [8:09] So we said, "Look at Netflix, look at Spotify. [8:13] If you had to buy one of these businesses [8:16] and hold them for five years, [8:17] which one would it be and why?" [8:19] So, it was like Netflix all the time [8:22] because I was talking about Netflix with LPs, [8:25] I was talking about it with interview candidates, [8:27] and ultimately the thesis was correct [8:31] and I didn't hold it long enough. [8:33] And so, in 2018, the reason Netflix was contrarian [8:36] was it was obviously a great product that people loved, [8:38] but they were burning a lot of money. [8:39] And so it was just not clear [8:41] was this like a classic tech company, [8:42] where, would they ever make money? [8:43] Like, that's why it might've been contrarian with LPs. [8:46] Yeah, I think the issue is [8:48] there's very few tech companies [8:49] that are massively capital-intensive, right? [8:52] Almost every tech company loses money for a period of time. [8:55] But the typical software company, you're looking at, like, [8:58] there's operating losses, you leverage it, [9:00] and then people are very used to that business model. [9:03] Up into the LLMs, there are very few tech companies [9:05] where it's like a huge fixed investment, [9:08] and then the incremental margins on the sales [9:12] are extremely high. [9:14] And so, what that meant was that Netflix [9:17] was investing heavily in content, which is a fixed cost, [9:21] and then they were selling that to consumers. [9:23] The next year, they were investing more in content [9:27] and selling it to more consumers. [9:28] But you're constantly investing more and more in content. [9:30] And the skeptics thought this would have [9:32] to keep going up forever and they'd never make money. [9:33] Yeah, because cash flow just looks worse and worse [9:35] every year because you're investing more and more. [9:37] But ultimately, that created a moat that was [9:44] I think the defining aspect of the business model [9:48] that made Netflix what it is today, [9:50] meaning that it's like a flywheel. [9:52] You invest a ton before anyone else. [9:55] Actually what happened here is the media companies [9:57] enabled Netflix by selling them content [9:59] and then Netflix invested a ton, sold to people globally, [10:03] took that money, invested more in content, [10:05] actually borrowed in the high yield market, [10:06] invested more in content, sold it to more people, [10:09] and then the flywheel then, you know, [10:10] within five years they're investing way more [10:12] in content than anyone else. [10:14] They're selling it to way more consumers. [10:15] The incremental cost of selling that content [10:18] is very, very low. [10:19] And so you have this fixed asset [10:22] that you've built up that... [10:25] And what ultimately caused you to sell it? [10:28] We had transitions on our team. [10:30] And, look, we cover every sector, right? [10:32] So, there's like, [10:34] we're covering at any given time 300 stocks, [10:36] and we had transitions on our team [10:38] and our media analyst left and [10:44] I was focused on other things. [10:46] It was not excusable, because if I look back, [10:48] we get plenty of things wrong. But Netflix, [10:51] we got exactly right. Yeah, how do you think [10:52] about, there's like a sense of, do you ever have this issue [10:53] where like people want to pitch new things [10:55] because they're like new and exciting, [10:57] and, I guess, the good old Occam's razor idea— [11:00] just hold what you have. [11:02] Do you ever have to push back against that? [11:05] Yeah, I probably should push back more. [11:07] I mean, you know, I think it's like a human tendency, [11:09] like, "Oh, this is a new company. [11:10] I love investing. [11:11] Let's learn about this new company. [11:13] And this sounds super exciting." [11:14] But at the end of the day, [11:16] there's like only so many amazing companies, [11:19] and trying to [11:22] sell them and move into something else [11:24] is almost always a bad decision. [11:26] That's the nice thing about the private markets [11:27] is that, once you invest, you can't sell. [11:30] And usually for the best companies, [11:32] that's a huge benefit, ultimately. [11:34] Yeah. I just want to go back to starting D1 for a second. [11:37] So the lore is that prior to D1, [11:41] you were at a firm called Viking. [11:43] Is that right? Yep. [11:44] And the lore is that almost all of Viking returns [11:48] increasingly came from Dan Sundheim offhandedly [11:52] making a comment to someone at the [11:54] water cooler that, "Oh, that could be a good stock," [11:56] or "Definitely don't invest in that," [11:58] and people would be like taking those trades [12:00] as their own for the most part. [12:02] Those are your words, not mine. [12:04] That's the lore. [12:05] I never said that. [12:06] That's the lore. [12:07] And I guess my question is, when someone comes up to you, [12:10] an analyst on your team today, and says, [12:12] "Oh hey, what do you think of X, Y, Z?" [12:14] And you do that sort of Dan Sundheim 30-second, [12:17] 10-second take. What is going through your head exactly [12:22] that gives you just this binary sense of [12:24] looks good, not good? [12:27] Well, I mean, every situation's different. [12:29] I think the biggest risk is sometimes people will come to me [12:32] with an idea, and like, [12:35] and they pitch it for a little bit, [12:36] and I think about it, and I give them an answer, [12:38] and I don't have enough information [12:41] to give them a really solid answer, [12:42] but whatever I say they think is like, [12:44] "Okay, well if he said that's not good, [12:46] then I'm just going to forget about it." [12:47] And that's actually not, you know, that's not constructive. [12:49] Like, I am wrong all the time. [12:51] Even when I've done a ton of work. [12:52] And if it's just like a 30-second pitch at the water cooler, [12:55] I can say like, "That sounds interesting," [12:57] or I could say, "That doesn't sound interesting," [12:59] but I'd say my hit rate there is dramatically lower [13:02] than once we've done all the work. [13:04] And there is a risk, like, analysts are like, [13:06] they try to take my temperature and if I'm going to be, [13:08] if it's going to be something that interests me [13:10] and if it's not going to interest me, [13:11] don't waste time on it. [13:12] That's actually not great. [13:13] But I think what Daniel is getting at is, [13:15] you seem to have an intuitive sense for companies [13:18] and in this underwriting. [13:20] Like, it's not just all about does the model [13:22] spit out a 19% or a 21% IRR. What is that, [13:27] that the Spidey sense is picking up on? [13:30] I mean, I think it's like anything else in life. [13:33] It's pattern recognition, right? [13:34] Part of it is just understanding business models, [13:38] understanding what kind of valuations [13:41] companies should be trading at. [13:43] Part of it is just having invested in this capacity [13:50] for 20-something years, [13:52] you get a sense. Like, if somebody comes to you [13:55] and pitches an idea to allocate a ton of R&D resources [13:59] to some engineering project, [14:01] you probably have a pretty good sense upfront, [14:04] whether or not like they should go write a big memo [14:07] proposing that or not. [14:08] When you interview young, younger— [14:10] It's an art, not a science. [14:11] Right, so when you interview younger, [14:12] say, portfolio managers, or when you meet— [14:15] do you have a sense, fairly quickly, [14:17] if they have whatever this is? [14:19] Like, do you think this is a thing [14:21] just people have or they don't? [14:23] I think the answer is yes. [14:25] So, a couple things. [14:26] One, we don't hire portfolio managers. [14:29] I pretty much only hire people [14:31] who've never done public equity before, [14:34] which has pros and cons. So what have they [14:35] done before? [14:38] Piano teacher. [14:40] Hey, glad we're pouring Guinness. [14:43] Let's split the G. [14:44] We typically hire from private equity [14:47] because then they have the analytical skills, [14:50] they understand accounting, [14:51] they understand financial modeling, [14:52] and then we can teach them the stock picking. [14:55] And if I hire somebody laterally [14:58] who's a portfolio manager at another fund, [15:00] that's rarely been successful. [15:03] Almost never. [15:04] There's some negative transfer? [15:05] Like, they have expectations that are wrong? [15:06] Everybody has a different, [15:07] like, every firm has a different approach to investing [15:10] and getting people to change their habits [15:13] to align more with how we invest is incredibly difficult. [15:17] Now, if I hire somebody from private equity, [15:21] it takes about three years for them [15:23] to really be like contributing to D1. [15:26] So, obviously it would be much faster [15:28] for me to just hire an external [15:30] and there's funds— And what is that? [15:31] So like they come in, you say it takes three years. [15:34] What's happening? [15:36] You know, say they're three months in [15:38] and they're trying to speak up in a meeting, [15:39] and you must be thinking to yourself, [15:40] "This isn't really the D1 way." [15:43] What is that exactly that changes in them [15:46] over the course of the three years? [15:48] Yeah, I mean look, I would tell you that I wish that— [15:50] it is incredibly difficult to figure out [15:54] who is going to be great. [15:57] And even after three years, [16:00] I'd say our hit rate still is not as high [16:05] as you would expect. [16:06] I imagine if you hire an engineer, after three years, [16:10] you probably have a pretty good sense [16:11] if that guy's a good engineer, or girl's a good engineer. [16:14] I think that I have a pretty good sense, [16:19] but it takes even five years to really get a sense [16:24] because some people start out [16:27] where they're analytically really solid [16:30] and they work incredibly hard, [16:32] but the actual intuition of stock picking [16:37] hasn't come to them yet. [16:39] Is it because you're saying like the job [16:40] is so pattern recognition-oriented [16:43] that you just need some time [16:44] to build up the pattern recognition? [16:46] Is that what they're learning? [16:48] Once in a while, people come in and, [16:50] you know, right out of thee the gate you're like, [16:51] "Wow, this person, like, sees it. [16:53] They see the ball really clearly." [16:56] But that is a very small fraction of the time. [16:59] Most people get great over time [17:01] and they have to learn an industry. [17:02] So when they come in, we will say, [17:05] "You're going to cover fintech," [17:08] and it just takes them a year just [17:09] to understand fintech, right? [17:11] Then they have to see, you know, well why is X, Y, Z stock [17:15] trading at this multiple? [17:16] Why is this stock trading at this multiple? [17:17] What's the market saying? [17:19] And yeah, I think it's like the market [17:21] is constantly giving you data points. [17:24] Some of them are, you know, false signals [17:26] and some of them are good signals, [17:27] and over the long term, they're all good signals. [17:29] And the people who do great at this job, [17:32] you have to have some commercial sense, [17:35] which I think is like probably just, [17:37] you're either born with it or not. [17:39] Charlie Munger's money-making gene. [17:40] Yeah, I wish I could like test for that. [17:43] It's impossible. [17:44] But you also have to love the job. [17:46] Like, you have to wake up in the morning in the shower, [17:48] you're thinking about, like, your stocks. [17:51] This isn't the kind of job where [17:53] you close your laptop, you go home, [17:55] and you forget about it. [17:56] It's, kind of, you have to always be on [17:58] and there's people who love it, [17:59] and if you love it and you have that commercial instinct, [18:04] that's, you know. [18:05] And you have to be analytically sharp. [18:07] Do you have a leaderboard of individual people's [18:10] portfolios and how they're doing compared to each other? [18:12] Yes. [18:13] Everyone has what we call a mock portfolio. [18:16] So, every week they have to take the positions they cover [18:19] and they have to say, "If I was managing capital, [18:21] here's how I would allocate capital [18:24] among the names I..." Wow. [18:25] And then what happens at the end of the year [18:27] is that sometimes it's like I own these five stocks, [18:32] my mock portfolio was up, and we take it very seriously. [18:35] Like, you know, the mock portfolio [18:37] actually goes into people's comp. [18:39] Sometimes it's like [18:40] that person did phenomenally well in their ideas [18:43] and I didn't monetize them. [18:46] And then sometimes it's the opposite. [18:48] But like, it takes care of a lot of the typical [18:52] hedge fund thing of, you know, [18:54] at the end of the year, like, oh, everyone like remembers [18:56] the things that they wanted to do that worked out, [18:57] and forgets the—you know. Ah. [19:00] But people have to pre-register [19:02] what they're going to come to you at the end of the year [19:03] and say, "See I was right." [19:04] Yeah, exactly. [19:05] So, I know, did I— [19:09] Do you ever have a situation with the mock portfolio [19:11] where someone goes short something you're long [19:13] or vice versa? [19:16] That hasn't happened yet? [19:17] No, that hasn't hasn't happened yet. [19:19] But there are big discrepancies because [19:22] sometimes analysts will be like, [19:24] "I think this is a 5% position," [19:25] and I'll be like, "I think you're crazy. [19:27] I think this should be double that." [19:28] Do you have a mock portfolio yourself [19:31] or is D1 your mock portfolio? [19:34] That wouldn't be very useful. [19:35] D1 is my mock portfolio. [19:37] Yes, and then, you talked about sizing positions. [19:40] Everyone's coming to you with all these good ideas, [19:44] because, presumably, they're not bringing you [19:45] their bad ideas, and they're like totally idiosyncratic. [19:48] Like, some are super safe bets, [19:51] some are kind of flyers, [19:52] some are industrials and tech and banks and everything. [19:55] How do you size positions [19:57] and how do you just assemble that into a sensible portfolio? [20:00] Yeah, that is really difficult to answer [20:02] because it is, you know, it is entirely an art [20:07] that we're always trying to get better at, [20:09] and we look at tons of data just to [20:12] see what we could do better. [20:13] But it is really just to feel like, look, [20:16] for every stock there is a target price [20:19] but then there's also what kind of risk? [20:21] There's a risk-reward calculation. [20:23] Like, okay, maybe you could make 50% [20:25] but you also could lose 50% if you're wrong. [20:28] Sometimes you can make 25% [20:29] but the risk of losing money is quite low over time. [20:32] And it's a matter of like, [20:35] you have to compare those two ideas which are [20:38] inherently very, very different [20:41] and construct a portfolio which has— [20:44] you don't want all names where you could make [20:46] a hundred percent but lose 40, [20:48] but you probably don't want to have a portfolio [20:51] full of names where you're only [20:51] going to make 20% and lose three. [20:53] So, it's a lot of things that go into it. [20:56] There's the risk-reward, [20:57] there's the ultimate upside. [20:59] There's a sense of [21:00] what's currently happening in the business. [21:02] Like, if a business we think [21:05] currently has a lot of momentum, [21:07] I'm not predicting quarters. [21:08] But if the business currently has a lot of momentum, [21:11] the chance of you losing money in the short term, [21:13] assuming it's at a reasonable valuation, is much lower [21:15] than if a business—we're often buying businesses [21:18] when they're going through really tough periods, right? [21:20] You have to be very careful because [21:24] often that's the best buying opportunity [21:25] but that's the riskiest time too, [21:28] because sometimes it's hard to call the bottom. [21:32] When you say you have to be really careful, [21:33] isn't there a principle agent issue here [21:36] where when you analyze a company, you try to see, [21:41] you try to predict what its earnings will be [21:43] three years hence, and be more right than everyone else's. [21:47] Maybe there are too pessimistic on the company [21:49] and you think it'll actually really over-earn [21:50] compared to people's expectations. [21:53] But, you're actually really sensitive to what people think [21:57] in the short term because your LPs are grading you [22:02] on the performance in the year [22:04] and the performance in the quarter. [22:05] Correct. And you're judging [22:08] this poor analyst and their job [22:10] on the mock portfolio, which you're saying [22:11] is like updated weekly. [22:13] And so, aren't you not actually looking at companies [22:17] on a three-year time horizon because they have to [22:19] perfectly perform at every step along the way [22:21] and they can't be misunderstood for any short period? [22:23] Yeah, the way I think about it is, [22:25] any given time we are planting seeds [22:27] and then we're harvesting. [22:28] Like, there's some companies that we invested in a year ago [22:31] and our thesis is starting to play out [22:33] and you're making money, hopefully, on that idea, [22:37] and maybe you're selling that idea at that point, [22:41] and then you're putting in new ideas in the portfolio [22:43] that might take another year to play out. [22:45] And so, it's not like if we started out [22:47] with a bunch of ideas that all had three-year targets [22:49] all at once, like, yes. [22:51] But like, the portfolio's a living thing where [22:54] you have a range of positions. [22:56] Some you've had for a year or two and you think the [22:58] business is going to turn the stocks into work, [23:02] and then there's other companies [23:04] that you're buying now because they're really depressed [23:07] and really out of favor and [23:08] you know they're not going to go up, [23:09] you don't think they're going to go up [23:10] in the next three or six months, [23:12] but over any medium term timeframe they will. [23:14] So, it's like always— [23:15] So you've limited capacity in your portfolio [23:17] for the out-of-favor stuff where, [23:18] ultimately, you're smoothing this for LPs [23:20] where you always need some companies [23:22] that are performing and are popular. [23:24] No, because like, [23:26] I may have bought a company a year ago [23:28] that was out of favor, and maybe it's still out of favor, [23:31] but eventually, the way economic cycles [23:34] or industry cycles work, the company I bought a year ago, [23:36] maybe I didn't make money on it for the first nine months [23:39] and now it's starting to inflect. [23:42] And I'm putting a new company in the portfolio, [23:43] which is out of favor, and that will start to work, [23:46] it's impossible to predict, that'll start to work [23:48] 6, 12, 18 months in the future. [23:51] But if you have a portfolio of things [23:53] that you've built over time, [23:55] our monthly returns and quarterly returns [23:57] are purely an output. [23:58] Like, it is arbitrary in some respects. [24:02] I think Jeff Bezos once said, [24:05] when people congratulate him on the quarter, he said like, [24:08] "Yeah, but that quarter was cemented like three years ago [24:11] when I made this decision to do X, Y, Z." [24:14] It's kind of the same thing with us. [24:15] When the stock's working, we probably were investing [24:18] in that stock a year, 18 months ago. [24:20] And then, you know, now we're investing in a new stock [24:23] which hopefully will pay off in 18 months. [24:27] So one thing you're kind of famous for is pain tolerance. [24:31] You know, you go through phases of euphoria in markets. [24:36] You go through phases of pain in markets. [24:38] And I was just wondering if you'd like walk us through, [24:40] maybe across D1's history, the most painful phase [24:45] and just how you, yeah. [24:46] How was that for you? [24:48] That's a very oblique way of asking about GameStop. [24:50] Yeah, yeah, yeah. [24:53] I do have a high pain tolerance. [24:57] So, the first three years of our fund, [25:00] our fund did phenomenally well. [25:02] Way better than I would've expected. [25:04] There's always some mean reversion. You expect, [25:06] okay, the next 12 months are going to be harder [25:10] because you've just made so much money. [25:13] As we were entering 2021, we had an amazing run. [25:17] Everybody in our firm was elated. [25:19] Everyone had done well, [25:21] they were part of a winning team. [25:23] And, for me, I'm always a little nervous [25:26] in the back of my head when things are going too well. [25:28] What I didn't expect was that the mean reversion [25:31] would happen within three weeks and would, you know, [25:34] be so dramatic that, like, it... [25:38] We never came close to going out of business, [25:40] never had margin calls. [25:42] But it was an absolutely insane, [25:46] harrowing experience. So what happened? [25:46] What three weeks was that? [25:48] Okay, so, 2020 ends, everything's great. [25:51] We're up by, I don't even remember how much we were up, [25:54] but very, very high. A lot. [25:56] And back then, [25:57] we are fairly aggressive short sellers. [26:00] I like shorting stocks. [26:03] It's masochistic, but I love doing it [26:05] and I've always liked shorting stocks. [26:08] This is at a time when, [26:13] we go back to like 2021, think in your head, [26:14] the government was mailing everybody checks, [26:16] everybody was sitting at home, [26:18] they weren't going to the office. [26:19] Reddit was starting to become a big thing. [26:22] And all those things came together to create [26:27] a massive short squeeze on the order [26:29] that I've never seen before. [26:30] The way I've always thought about stocks is [26:33] let's say I short a stock. I think it's going to earn X. [26:37] I know that if I'm wrong, it's going to go up this much [26:41] and probably goes up more than it should fundamentally [26:43] because people are short and they're going to have to cover. [26:46] That's my whole career. [26:48] I've never had, in my risk-reward framework, [26:50] nothing happens fundamentally [26:52] and a stock goes up 400% in two weeks. [26:56] That's what happened. [26:57] And if you think about what shorting is, [27:01] shorting is leverage, right? [27:02] It's just borrowing shares that you have to pay back. [27:05] Imagine borrowing a lot of money [27:08] and the amount that you have to pay back goes up [27:10] by a hundred percent every week. [27:12] That's very stressful. [27:15] It's a development market deathtrap kind of. [27:18] That was quite stressful. [27:20] And then imagine that also at the same time, [27:23] the American populace is basically all coming together [27:28] to think that like, these are the people you should attack, [27:32] these short sellers. [27:34] And so, it was like CNBC was like— [27:36] it felt like everybody was behind this massive short squeeze [27:41] and felt like it was a great thing [27:43] because these hedge fund managers were paying [27:44] for what— [27:46] That was the surprising thing, right? [27:47] Because it's the age-old "in a crisis, [27:50] all the correlations go to one," [27:52] and things that are supposedly in your model uncorrelated [27:54] actually turn out not to be. [27:56] This is the long-term capital management issue. [27:58] It's like, oh crap, everything's correlated. [27:59] And similarly here, wasn't it that like any one stock [28:02] spiking 400% would've been fine [28:07] but it was the fact that this new retail phenomenon [28:09] of all of the highly-shorted stocks by hedge funds [28:13] melting up together was maybe the thing that [28:16] you couldn't have predicted. Exactly. [28:17] People say GameStop, but GameStop is just like, [28:19] that's a name that people use [28:21] because it was the most prominent name. [28:23] But like, what was the experience like? [28:24] So it's week one, it's up a hundred percent. [28:26] The week was like, the experience was, [28:29] I was almost like in shock. [28:30] We were losing an amount of money that was almost hard [28:35] to even fathom that you could lose that much money [28:38] in a short period of time. [28:40] And imagine, like, we were going from being [28:42] on top of the world, to within three weeks we were like, [28:45] people were saying like, [28:47] "Oh, these funds are going to go out of business." [28:49] And I am generally pretty even-keeled [28:51] when it comes to markets. [28:53] I've been doing this a long time. [28:54] I've seen a lot of things. [28:55] That was the toughest period I've ever had. [29:01] And it was really stressful to the point of like, [29:10] I felt like I was in emotional shock. [29:12] Now, I was still laser-focused [29:13] and I knew I'm on top of it, [29:17] you know, I'm acting completely rationally. [29:19] But it was very, very, very hard. [29:24] And I had friends who were getting bailed out by other funds [29:27] and you felt like the whole world was against you [29:30] because you know, there's all these people— [29:31] It kind of was. [29:32] Yeah, it kind of was. [29:34] It's like, it was really hard. [29:35] What I'd say about this job is that, [29:39] there's like an asymmetry of emotion, right? [29:42] When you're making a lot of money, [29:44] it's not like I'm like, [29:46] tap dancing around my house, like euphoric. [29:49] My lifestyle doesn't change. [29:51] I feel like, yes, I'm happy that we're succeeding, [29:54] but, you know, I'm happy but I'm not euphoric. [29:58] When you go through something like that, [30:01] it is like— Yeah, it's not 1 over x. [30:03] So painful and stressful [30:06] and it's not because I'm losing my own money. [30:08] It's more about losing other people's money. [30:10] You know, it's more about [30:12] you know that what's happening [30:14] is completely economically irrational [30:16] and how long can you actually stick with it? [30:20] Yeah, is that the painful thing, [30:22] that if you're covering the short, [30:23] you know that you're covering it at a bad time [30:25] because this is not a long-term sustainable price level? [30:27] Yeah, so we covered everything at the worst time [30:32] and I knew it was the worst time, [30:34] but if it kept going on for two weeks, [30:38] all of a sudden there's a real problem. [30:40] And I wasn't going to put the firm at risk. [30:42] And I told my investors, I said, "This is the worst time." [30:43] Now, if I had held that short portfolio, [30:46] forget like from the top. [30:47] If I had just held from the end of 2020, [30:50] I mean, they underperformed massively. [30:53] Like, it would've been like— [30:54] You were right on the shorts. [30:55] I was right on the shorts [30:56] but I was wrong on their risk management. [31:00] So again, there's this new phenomenon [31:01] of retail melt-ups in stocks, and in particular, [31:05] certain stocks seem to attract the retail crowd [31:11] and the price movement of the company stock [31:14] just seems to, like, I was just looking now, [31:17] there's a lot of retail interest in Opendoor. [31:19] And so the price movement just kind of takes off [31:21] and starts doing its own thing [31:23] that's kind of different to what's been happening, [31:24] how it's been valued previously. [31:26] How does this new retail phenomenon [31:29] change the short-selling game for you? [31:32] I would say that the opportunity for short selling [31:35] is better than it's ever been in my career. [31:38] However... [31:40] Oh, damn it. [31:40] Like, however. There's a "however." [31:42] However, you cannot capitalize on that opportunity. [31:47] Like, if we went back to 2019 [31:49] and the same thing was happening, [31:51] I would be massively short these companies. [31:54] But after going through GameStop, [31:57] there's a ton of opportunity [31:58] but you have to be much smaller in each individual name [32:02] so that you don't subject yourself to [32:05] what happened in January of 2021. [32:08] But is this a new kind of science [32:10] just like understanding what the crowd is going to do? [32:13] Or is that just like, it's too random. [32:15] There's no science behind that. It's like— [32:17] But you could have the Reddit analytics [32:19] and following the tweets and— [32:21] Yeah, you could follow [32:22] Reddit tweets all day long, [32:24] but like, what does that tell me? [32:26] Like, okay, the stock went up a hundred percent. [32:28] There's a bunch of people who like the stock. [32:30] Obviously, if I read the Reddit, [32:33] duh, they like the stock. [32:34] It went up a hundred percent. [32:37] There's no information signal there. [32:40] I know that people are buying the stock, [32:42] I know that they're buying the stock for a reason [32:44] that I think is kind of silly. [32:48] The most important thing to me [32:50] is that we just have to be able to ride it out. [32:52] So, whatever happens, [32:53] a risk management cannot happen after the fact. [32:57] Like, anything you do after the fact is not risk management. [33:00] It's actually just destroying capital. [33:02] Risk management has to happen before the fact, [33:04] meaning you size these positions [33:06] so that when it goes up, you don't have to cover. [33:09] Anyone who says, [33:10] "Oh, I like risk-managed the portfolio [33:13] because things were going against me." [33:15] If you're a bad stock picker, sure. [33:17] Yeah, it's like I cut my losses. [33:18] If you're any good at stock picking, [33:19] risk management after the fact is a bad idea. [33:21] So we now size things. [33:23] So I said, look, I'm not trying to figure out [33:27] what's happening on Reddit or this or that. [33:33] It's like people get excited about, [33:36] one stock is in fashion one day, [33:37] then the next week they move on to another one. [33:39] I can't predict that. [33:41] All I can do is size a position such that [33:44] if they get excited and the stock goes crazy, [33:46] we don't have to cover. [33:47] We just kind of let it go and ride it through. [33:49] So is it something like previously you could have had [33:51] a more concentrated short portfolio where you could have had [33:54] a few shorts that you really like, [33:56] and you would have said, "Oh, it can go crazy [33:59] and it can 2x if it wants, and I'm still fine." [34:02] Whereas now you need to have many more [34:03] individual short positions where any of them can 10x or 20x [34:08] and you're still able to ride it out. [34:10] Correct. 10x or 20x is a little extreme. [34:12] We don't have that very often. [34:13] But this kind of stuff happens! [34:14] It does happen. [34:15] Yeah, we don't trade like $1 stocks. [34:17] But yes, that's exactly right. [34:19] After GameStop, we took about a year off of short selling, [34:22] and then I reengaged in short selling. [34:24] It's like someone going through a tough breakup [34:26] and saying, "I'm not dating right now." [34:28] Yeah, it was like, okay, we're going to step back. [34:30] I didn't know a 500 mile per hour hurricane existed. [34:34] I didn't know that could happen. It happened. [34:35] Okay, so now we have to figure out [34:37] what kind of windows we're going to put in our house, right? [34:39] So, we took about a year off from shorting, [34:41] which hurt us a lot in 2022. [34:43] When we got back into it, I said, [34:45] "Look, we're going to have to be more diverse. [34:47] We're going to have to be less aggressive." [34:51] And my view going in is like, [34:55] logically you should expect that your returns will be worse. [34:58] And then the question is like, okay, so shorting [35:02] is a really hard thing to do. [35:03] So, at some point it's like, okay, [35:05] if you're going to diversify so much [35:06] and your returns are going to go down, [35:07] like is it even worth it? [35:09] And if you asked me at that time, my LPs asked me, [35:11] I'd say like, "I don't know. [35:13] I like shorting stocks, I think we can do it, [35:16] but this new strategy, not quite sure [35:18] if the juice is worth the squeeze." [35:20] As it turns out, if you look at our short alpha [35:24] since that time and since we implemented [35:27] the more diverse portfolio you referenced, [35:29] our short alpha has been as good or better. [35:31] And I think the reason is because— [35:33] As your long alpha or as the prior short alpha? [35:35] No, as the prior short alpha [35:36] when we were more concentrated. [35:37] I think the reason is is that you can't be as big, [35:40] but there's so many more mispriced stocks [35:43] that you can have 40 of them [35:45] instead of having 8 of them. [35:50] What makes a good short? [35:52] Good question. [35:54] Yeah, like, I mean there's a lot [35:55] of different kinds of shorts. [35:58] So, the ones we've been talking about are like stocks [36:02] where usually what gets the retail crowd excited [36:07] on social media. Usually it's stocks that have some story [36:11] which is novel and exciting. [36:13] Maybe it's new technology, the revenue growth is really good [36:18] and you're kind of analyzing the business model [36:21] and saying this technology just doesn't work, [36:25] or this company's actually, [36:27] you know, the founder's horrible. [36:29] That's its own thing. [36:31] And then you have shorts where, [36:34] shorts that we think have like real terminal value risk, [36:37] where we look at the company and say, [36:39] "I actually don't think [36:40] this company will be around in 15 years." [36:44] And that's usually because something secular [36:46] is happening in the economy. [36:48] But does Kodak type stuff, [36:50] where like the business will go away, [36:51] as opposed to Theranos type stuff, [36:54] usually where it's like pure fraud. [36:56] Exactly. [36:57] Theranos would be like it's own thing. [36:58] So there's frauds, [37:01] which I used to do a lot more when I was younger. [37:04] It's hard to find frauds at $10 billion companies. [37:07] It happens, but we're not— [37:08] US capital markets are pretty good [37:10] for eliminating frauds from the... [37:12] Yeah, you'll find them [37:13] but they're usually in companies [37:14] like that are $1 billion market caps, [37:16] or $2 billion market caps, [37:16] which is kind of below the threshold where we play. [37:19] Because you can't get enough exposure. [37:20] Yeah, I mean a fraud rarely gets [37:22] to be, like, Enron size. [37:25] So you have the retail stocks, [37:27] which are just like story stocks that have, you know, [37:30] there's really no real substance to the company. [37:33] Then you have terminal value stocks, [37:34] like Kodak would be good example, [37:38] where you just look and say, like, [37:39] here is where I see the economy going [37:42] because of some secular trend or [37:46] some technological innovation, [37:48] and this company's business model [37:50] is going to be severely compromised over the next decade. [37:55] Then you have stocks where you just think [37:57] that the business from a competitive standpoint [38:00] is just really disadvantaged [38:02] and it's going to cede market share [38:04] for a very long period of time. [38:06] And then I'd say the last and least attractive shorts [38:09] are the ones where it's like, cyclically, [38:12] they're just over-earning a lot [38:14] and you believe that the real earnings, [38:18] if you were to look through a cycle, [38:20] are much lower than people think. [38:22] And how do you think about, so there's a company, [38:24] maybe they're in a bad place in the economy, [38:26] it just doesn't seem like it'll work out, [38:27] but there's always this fear [38:28] that they get bought out, [38:30] or that there's a sudden management change, [38:33] and do you just size your books [38:35] such that that's okay for you? [38:37] Or do you use options? [38:38] Take over. [38:39] I find that, generally, companies [38:42] that have secular risks almost never get bought. [38:46] Rarely do I ever see a CEO say, [38:51] "I want to go buy something that grows slower than I do." [38:54] Right? That just doesn't happen, right? [38:57] Nobody wants to take the time [38:59] to do an acquisition that's going to [39:01] invite a bunch of secular risk into their business. [39:03] Maybe Ron Perelman did that back in the day [39:05] but that doesn't happen. [39:09] What can happen is if you have a company, [39:14] like the example I gave where there's a company [39:16] in the industry that you just believe [39:18] is going to cede market share for a long period of time [39:20] because of their positioning or management or strategy. [39:24] And then what can happen is an activist comes in, [39:29] takes a big stake in the company, and says, [39:31] "We're going to replace the management team [39:33] and once we do that, [39:36] the new management team will pursue a different strategy [39:39] and they will no longer lose market share." [39:43] That is a big risk in that category, right? [39:47] That's kind of the only category where like— [39:50] And to manage that risk, [39:52] you just managed your book. [39:54] Do you use derivatives at all? [39:56] I really don't use derivatives at all. [39:58] The problem with derivatives is, look, [40:00] at the end of the day when you break down [40:02] what is a derivative? [40:03] It's just leverage and implied volatility. [40:06] If I want leverage, I can go [40:09] get leverage from a prime broker. [40:11] I don't have any view on implied volatility. [40:13] And the problem with derivatives [40:14] is there's typically a timeframe [40:16] and I'm not good at timeframes. [40:18] I don't know when a stock is going to work. [40:21] That's very difficult. [40:22] So, I don't like having something that, [40:25] making a bet that something's going to happen [40:26] in a certain period of time. [40:28] We keep it pretty simple [40:30] and just trade underlying stocks. [40:31] So, I think that like nobody buys businesses [40:34] that grow slower than them, or rarely. [40:37] People usually don't buy companies [40:39] that are market share donors. [40:43] With CEOs, I really want to like buy this. [40:45] It's like, that's a headache, right? [40:48] So there's a stage of the life cycle [40:49] where companies kind of become unacquirable, kind of? [40:52] Those companies, like, [40:53] I don't worry about them being acquired. [40:54] What you worry about, as I said, activist comes in, [40:57] replaces management. [40:59] Usually the new management team can't fix it, [41:01] but sometimes they can and that's a risk. [41:06] In terms of the companies [41:08] where I said they have no terminal value, [41:09] the risk is just that they usually trade [41:12] at low multiples, right? [41:13] And so, you're basically just DCF-ing the cash flows, [41:17] and if the cash flows, if something happens [41:19] and people perceive the cash flow is going to [41:20] last a little bit longer, [41:22] because the starting valuation's low, they can go up. [41:27] Cyclical shorts are just entirely different. [41:30] The risk is just that you're early [41:33] or you're just wrong about the cycle. [41:34] Like, there's something about the cycle [41:35] which is different this time. [41:38] But those are kind of the main categories of shorts. [41:40] Speaking of cyclicals, why has Rolls done so well? [41:43] Like, they're up 5x, 10x over the last few years? [41:47] This is Rolls-Royce, the jet engine [41:49] and turbine manufacturer. [41:50] Yeah, yeah, yeah. [41:51] Main business is aircraft engines for wide-body jets. [41:54] They're not big in AI, IGTs? [41:57] Not as far as I know. [41:59] It is like a little bit. [42:01] It's only a matter of time. They have an SMR business. [42:02] Okay. [42:03] So, that could eventually help. [42:05] Rolls is not very cyclical. [42:08] The reason why— Aren't engine makers [42:09] historically cyclical? [42:11] Not as much as you would think [42:13] because most of the business is, [42:15] I sell you an engine, [42:16] I sell you an engine for not much money, [42:18] but I make a lot of money in aftermarket, [42:19] which is much more predictable. [42:20] So the air framers are cyclical, [42:22] but the engine makers aren't? [42:23] Air framers are cyclical, but pretty good secular growth, [42:30] and it's a duopoly. [42:32] Rolls-Royce was just horribly managed for a long time. [42:35] Making jet engines is incredibly difficult. [42:39] It sounds hard. [42:40] A new jet engine. It looks hard. [42:41] You're like, wow. [42:42] A kind of idea you think might be hard. [42:44] It probably takes 5 to 10 years to do the R&D [42:47] to develop a new jet engine. [42:48] So, Rolls-Royce actually had good technology. [42:50] It was just very, very poorly managed. [42:53] And they signed a bunch of contracts with airlines [42:56] that were very unfavorable. [43:00] And they had a new CEO come in, [43:03] and he operationally turned the business around [43:08] in a pretty fantastic way. [43:10] You guys were long Rolls-Royce? [43:12] Right? Yes. [43:13] So, that's a good example of like, how do you, [43:15] because every management team says— [43:17] Yeah, how'd you figure that out? [43:18] Yeah, every management team says, [43:19] "We're going to turn this thing around, [43:21] and every management team has a projection [43:23] that looks good. [43:25] And so, how did you determine that now finally [43:28] they were going to turn it around? [43:29] Okay, so let me start by saying [43:31] the US and Europe are very different in this respect. [43:33] I find that if we—let's take a US company [43:36] that had a turnaround, an industrial company like 3M. [43:40] 3M had been a horrible stock for a very long time. [43:43] Wasn't well managed. [43:46] New CEO comes in, puts up one or two good quarters, [43:51] and we owned it. Everybody basically understands [43:54] what's happening and the stock goes to fair value, [43:58] with the assumption that the margins [44:00] are going to go to where they should go. [44:01] The US is pretty quick at seeing change happening [44:05] and then pricing in that change. [44:08] In Europe, I find once a company, [44:12] like a company like Rolls had underperformed for so long. [44:15] I guess that European mutual funds, [44:18] they just kind of got in their head [44:20] that like Rolls is something we just don't want to touch. [44:23] So the voting machine is laggier in Europe? [44:27] Laggier. [44:28] Why is the information connectivity higher in the US? [44:30] Like, what's going on? [44:33] I think it's hard for me to explain [44:35] but I've seen it over and over again in Europe. [44:37] It's almost like when Rolls was being turned around, [44:45] it was pretty clear after the first year [44:47] that what he was doing was going to work. [44:50] And it wasn't really that difficult to— [44:55] Based on earnings, based on deliveries, [44:57] based on talk to customers? [44:58] Based on... You meet with the management team, [45:01] they say, "Here's our plan, [45:03] here's what we're going to do." [45:05] You see things playing out, [45:06] you see the income saving progressing [45:09] as the person said, you get a sense [45:12] for like Tufan, name of the CEO. [45:15] You get a sense for, okay, is this person good? [45:17] The same way you would assess— [45:18] Yeah, yeah. [45:19] Is the right framework that it's US and rest of world? [45:22] Or is Europe kind of uniquely bad at this [45:25] compared to Latin America, compared to Asia? [45:28] I don't do enough in Latin America to have a strong view. [45:33] But like a Japanese turnaround, would that— [45:34] I think Japanese turnaround would be closer [45:36] to Europe than the US [45:38] but I've done a lot more turnarounds in Europe. [45:42] If Rolls-Royce was traded in the US, [45:44] I'm fairly confident that after the first few quarters [45:47] and people meeting the CEO, it was very clear to me [45:50] that the CEO was excellent. [45:52] And why is that, functionally? [45:54] Like, do hedge fund managers, [45:56] or most of them presumably are in the US [45:57] and they mostly like to buy American stocks, or? [46:01] I think the American markets are just much more efficient. [46:04] There's just a lot more capital, and... [46:08] But there are no hedge funds that have, [46:09] very few hedge funds presumably only invest in the US. [46:14] You'd be surprised. [46:15] I think you'd be surprised. [46:16] I think Europe is generally viewed, [46:18] appropriately so, as a extremely low-GDP, [46:21] unexciting place Wow. [46:22] to deploy capital. For peak growth. [46:24] There's an issue where like presenting that to your LPs [46:27] is kind of embarrassing. [46:29] Is that part of the problem? [46:30] Like, oh, we took— Present what? [46:32] Oh, "We took a position in a European company. [46:34] I'm not sure we would want to do that." [46:35] No, no, no. [46:37] Obviously you don't, but I'm just trying to understand [46:39] why the average hedge fund manager [46:40] doesn't just back up into Rolls. [46:43] But isn't it just a home country bias? [46:45] Like, people invest in what they know? [46:46] Yeah, I mean I think it's, let's see. [46:50] I think it's much easier [46:52] from the US to invest in US companies. [46:54] You understand the accounting, [46:56] it's US GAAP. [46:59] You don't have to stay up all night, [47:01] in the middle of the night, to follow the stocks [47:03] when they report earnings. [47:06] Do you do that? [47:07] You're assuming like hedge funds— [47:08] If you have a big position in another time zone, [47:10] do you stay up to watch earnings? [47:12] Or do you wake up to watch earnings? [47:14] I actually, like, without an alarm clock, [47:16] I will wake up almost exactly [47:18] between 2:55 and 3 a.m. almost every day. [47:20] Wow. That's amazing. [47:21] Wait, let's dig in. [47:23] Yeah, there's so much to unpack there. [47:26] So you wake up ahead of earnings. [47:28] No, no. Forget earnings. [47:29] Like, every day. Oh. Oh, okay. [47:31] Like, every day. [47:32] Just because I've been doing it for, [47:33] who knows, 20 years? [47:34] Like, the European market opens— [47:36] depending on Daylight Savings Time—opens at 3 a.m. [47:40] You just want to see what happens at the open. [47:43] No, I don't want to wake up. [47:45] I actually don't want to wake up. [47:47] Yeah, but you're there. [47:48] But it's just like old habits die hard. [47:49] So you wake up, check Yahoo! Finance, and go back to sleep? [47:53] Yeah, sometimes I'll send out [47:55] a bunch of texts to— Thoughts. [47:57] Thoughts for the team. [47:57] People who have no interest in getting a text at 3 a.m., [48:00] and you know, I like, think of things and then, you know. [48:04] Does your team have like a special setting [48:05] on their iPhones where it doesn't just like [48:08] chirp at exactly— Boss calling. [48:09] Yeah, like it just tracks an alarm. [48:10] I've never actually asked them. [48:11] I don't expect—if the company reports earnings, [48:14] I expect the analyst to be awake. [48:16] Wow. And I'm awake, too. [48:18] You can't... That would be like— [48:20] Wow. That's happened. [48:21] Like, one time I heard— [48:22] Do you go back to sleep then? [48:24] I try to, but like— But don't. [48:26] I try to but, look, it's like— [48:28] I just want to have a split screen. [48:30] So, you are in Miami, you're up. [48:33] The analyst is in a one-bedroom in New York City, [48:37] laptop in the bed, on the phone with you, [48:40] it's three o'clock in the morning. [48:41] Company's about to report earnings. [48:43] His girlfriend doesn't understand why. [48:45] She's already at the couch in the other room. [48:48] And then it gaps down, it's gapping down, it's red. [48:51] We could see the red reflected on the analyst's face. [48:55] This would make a great movie. [48:55] Yeah, yeah, yeah, yeah. [48:56] And then what are you telling them to do right now? [48:59] Are you just saying, "Why is it red? [49:00] It should be green." [49:02] Well, usually it's red for the right reason, [49:06] and then it's just a matter of understanding, [49:08] okay well, is the stock overreacting? [49:11] Or, what actually happened? [49:14] Does this actually change our view [49:16] of the intrinsic value of the company or not? [49:19] And that's a matter of, [49:21] you know, the analyst is in real-time, [49:24] we're discussing what happened, [49:26] and sometimes we're buying, [49:29] rarely we're selling, but sometimes [49:31] there's a quarter where it's like [49:33] your whole thesis is just wrong. [49:35] Usually, it's like your thesis isn't totally broken. [49:39] Maybe the stock is down 5, 10%, [49:43] and you kind of say like, "I understand why, [49:45] but it doesn't really change my long term." [49:48] So it's been 30 minutes now, you did your stuff. [49:50] Now it's four o'clock in the morning. [49:53] Do you go back to sleep? [49:56] Yeah, it depends. Because the analyst [49:57] is not going back to sleep, I'm telling you that. [49:59] Yeah, it depends. [50:00] Ideally I go back to sleep. [50:03] It depends how red it is. [50:04] If it's really red, then it's hard. [50:07] Just push through. [50:08] Then there's more... We have to like, [50:11] if it's down 20%, it's not like, [50:12] "Oh, that's interesting, the stock's down 20%." [50:15] What's for breakfast? [50:17] Yeah, I'm going to go back to bed. [50:18] Catch you in the morning. [50:20] Then we're like, game on. [50:21] Then it might as well [50:23] be three in the afternoon because we're [50:28] deeply trying to understand [50:30] where we were wrong or if we were wrong, or— [50:35] You might be right at that moment [50:36] and it's time to buy maybe? [50:37] Yeah, sometimes. [50:38] Yeah. Yeah. [50:39] Okay, let's talk about, so you wake up. [50:41] So now it's— [50:43] You went back to bed [50:44] because it was up 2% or something. [50:45] You go back to bed. Now it's what, 7 a.m., 6 a.m.? [50:49] You wake up again. And then you reach for your phone, [50:53] and what's in your inbox? [50:54] What are you reading? [50:55] Because you said you're not CNBC. [50:56] At that point, like the analyst... [50:58] Usually when the earnings report comes out, [51:00] it's stressful for the analyst [51:01] because the company reports. [51:03] You're being judged [51:04] on your position. The company reports, [51:05] and I'm on the phone with them, [51:05] and I'm like— Think! [51:06] "What's happening?" [51:07] And they're like, Do the math! [51:09] "Give me a second. [51:10] Let me digest the information." [51:11] I'm like, "Think faster." [51:13] I'm kidding. [51:15] But not really. But in real time, [51:16] but not really. [51:18] In real time, the analyst is like looking at it [51:19] and telling me [51:21] his perception of what's happening. [51:24] And then by the time I wake up in the morning, [51:26] there's usually a thoughtful earnings review [51:29] which goes through, in detail, what happened. [51:32] And I only go back to bed if [51:36] there's nothing that has to happen that's like... [51:38] If we're not going to have to make dramatic changes [51:40] one way or another to the position, I'll go back to bed. [51:43] And in the morning, I'll read [51:45] what is like the analyst has time to step back, [51:49] think about it, write things up, [51:51] and then we, by that time, [51:54] we have a few more hours of trading in Europe. [51:57] We may do something, we may not. [51:58] Let's talk about the, across all markets, [52:00] there is this after hours thing, [52:03] and you would know better than me [52:05] but I feel like it's gotten even crazier in recent years [52:09] where it reports, the stock can occasionally [52:13] act extremely erratically in both directions. [52:16] I think after APP reported earnings last quarter, [52:19] it was down 13% and then it opened the next day up 13%. [52:23] Yeah. [52:24] What's going on there? [52:25] And you'd think it's thin enough. [52:27] At some point, someone should come in and make markets. [52:30] Well, look, for sure this is not what we do [52:33] but there are bonds that, [52:35] as soon as the earnings report goes out, [52:37] they are effectively having AI [52:40] read the earnings report and interpret, [52:42] and then it's just like, it's just trading. [52:44] When you're staring at that, [52:46] I presume you sometimes will wait [52:48] because you kind of anticipate this thing [52:49] to flop around for a while? [52:51] I mean, we transact in the aftermarket but it depends. [52:53] Like, if NVIDIA reports, [52:55] the aftermarket liquidity is going to be huge. [52:57] If a company that's $10 million reports, [53:00] it's a waste of time. [53:02] If you try to buy the stock, [53:03] you're just going to send it up too much. [53:04] It's better to wait. [53:08] Dan's in the business of money management, [53:11] efficiently allocating capital [53:12] to companies around the world. [53:14] That's the job he signed up for [53:15] and seems to be going pretty well for him so far. [53:18] But what we see at Stripe is that as companies scale, [53:20] every founder, every CFO, every finance team, [53:22] they are becoming money managers, too. [53:24] And not in the way that they want or that Dan does. [53:26] Managing a fleet of international bank accounts, [53:29] dealing with trapped liquidity, paying global teams, [53:31] it's the kind of complexity [53:33] that creeps up on fast-growing businesses. [53:35] At Stripe, we believe that treasury management [53:38] should be fast, globally native, and programmable. [53:41] And that's why we're building a new financial home [53:43] for your business right inside your Stripe dashboard. [53:46] You can now do things like hold balances [53:48] in pounds and euros and dollars [53:49] and convert between them instantly 24/7 and at low cost. [53:53] If you need to pay a partner, [53:54] you can send money to over 150 countries [53:56] with just an email address. [53:58] Explore Stripe's new money management capabilities [54:00] at stripe.com/treasury. [54:07] Another thing funds will often do is, [54:10] as they get big, they'll expand [54:11] to other kinds of products. Macro. [54:15] Is that ever something you think about? [54:17] Definitely not macro. [54:18] First of all, it's tempting to think [54:20] that because you're good at one thing, [54:22] you're going to be good at all these different things. [54:24] And I think that, in and of itself, is a mistake. [54:28] But the thing about my business that I like [54:32] is that if you buy a great company, [54:34] you might be wrong for a little bit [54:36] but time is your friend. [54:38] Like you own it, maybe it has a couple bad quarters, [54:41] doesn't go up, blah, blah, blah. [54:42] But as time goes on, the company compounds value, [54:46] you know, it grows. [54:48] If I'm going to short US government bonds, [54:53] I might never be right. [54:55] It's not like, oh, over time it has to go back to 5%. [55:00] You're just basically making a bet, [55:03] where like in my job, I like the fact that [55:06] if you buy bad businesses you don't have that. [55:08] But if you buy great businesses, [55:10] you overpay for them a little bit. [55:12] You get the earnings wrong a little bit in the short term. [55:13] Over time, value compounds [55:17] and time is your friend. [55:18] Macro, that's not the case. [55:20] It's more of a binary bet. [55:22] So you have this story of like, [55:24] it basically takes time and then the market catches up. [55:26] Do you have like any particular stocks [55:28] that you can share where the markets didn't get it [55:32] for a very long time and it actually took quite some time [55:35] for it to catch up? [55:37] Yeah, I mean sometimes it takes the markets [55:40] like years to really get it. [55:44] What's an example? [55:45] I mean, Netflix is an example. [55:47] If you really like— [55:50] And is the "get it" moment the multiple expansion moment? [55:53] Yes. Okay. [55:54] So that's really [55:55] the thing, right? Yeah. [55:56] The interesting thing about [55:58] when people don't get it. [55:59] So, if I told you what's logical, okay? [56:03] Let's say you have a company, here's a good example. [56:07] Booking.com, which was called Priceline [56:10] back when I owned it. [56:11] It was growing 40% a year [56:13] and it traded at nine times earnings. [56:15] Why did it trade at nine times earnings? [56:17] It traded nine times when it was growing 40% a year? [56:19] Correct. [56:20] That's not meant to be an occupied corner of the— [56:25] Yes, it didn't make a lot of sense. [56:27] But like the narrative back then was like, okay, [56:29] they're just like arbing Google [56:30] and this is like a flimsy business model, whatever. [56:33] Now, you know, years go by. [56:37] I don't know how fast Booking is growing [56:39] right off the top of my head [56:40] but I imagine it's single digits. [56:41] High teens, I think. [56:42] Is it that fast? Okay. [56:44] I'd be surprised if it's still growing high teens, [56:46] but it's a lot slower than it used to be [56:49] and it trades at like 3x the multiple, right? [56:52] So like, the multiple should be correlated with growth rate, [56:55] but sometimes the market is just so skeptical [56:58] about the business model that even as the business slows, [57:01] the multiple goes up. [57:02] So would it be fair to say that D1 makes- [57:04] I think Meta is trading at a higher multiple [57:06] than it did for the last 10 years. [57:08] And we'd have to look, like don't take that as fact [57:12] but I would say Meta's multiple has expanded over time, [57:16] even though, just with the law of large numbers, [57:18] it's still growing incredibly fast for its size, [57:22] but it's growing slower than it was 10 years ago. [57:24] But now, I think for Meta's whole life [57:28] there was a lot of question about the moat [57:31] and TikTok and this and that, [57:33] and now it's growing slower, [57:35] but it has a much higher multiple. [57:37] But would it be fair to say that D1 makes [57:39] most of its money on this concept of multiple expansion? [57:42] Like, is that the actual trade? [57:44] Yeah, so that's what I wanted to ask about, [57:45] because Ben Graham talks about the voting machine [57:48] versus the weighing machine, and if I read a D1 memo, [57:50] it'll all talk about company fundamentals [57:54] voting machine stuff where [57:56] Siemens Energy is a great business [57:58] and like there's going to be so many turbine orders [57:59] because of AI and the wind thing is complex [58:01] but it's actually grand, and whatever. [58:04] But over a one-year period, the largest part of stock moves, [58:09] as Daniel is saying, will be multiple expansion, [58:12] which is basically sentiment. [58:13] And so, aren't you actually in the sentiment— [58:16] Is it vibes? Are you kind of vibe trading? [58:19] Yeah, I mean I like to think of myself [58:21] as being a little more academic than vibe trading. [58:25] But, look, I think that you're always going to [58:30] make the most money from multiple expansion. [58:32] Just like it's hard to have that differentiated a view [58:36] on a company's growth rate. [58:38] You can have differentiated view on the longevity [58:41] of the growth rate or the ultimate margin profile [58:45] of the company based upon some work you've done. [58:49] But the multiple over the long term [58:53] is a function of the perception of the stability [58:57] of the cash flows over the long term. [58:59] And sometimes people are skeptical of a business model, [59:02] and then over time it gets proven out [59:04] that the business model doesn't deserve [59:05] to have that skepticism, and then the multiple expands. [59:08] You do make most of your money on multiple expansion. [59:11] And that's why I said at the beginning, [59:12] we talked about [59:14] like actually gauging what is the right multiple [59:17] for a business is quite hard. [59:18] And when I sell companies too early, [59:20] it's almost always because we use too low [59:23] of a multiple at exit. [59:24] If I'm a non-profit that you want to help out [59:27] and I say we need to put all our money into a single stock [59:30] for the next, and we're not going to touch it, [59:32] we're actually going to lock it up contractually [59:34] for the next 10 years. What would you recommend? [59:39] I think if we start like narrowing it down, [59:41] there's very few tech companies I would say [59:43] I feel comfortable saying [59:45] because I just think tech just changes too quickly. [59:48] So, it wouldn't be a tech company. [59:49] It would have to be a company with like a moat [59:53] that I just felt like would be incredibly difficult [59:56] to penetrate with a growth rate that was probably [60:00] not like super high but well above GDP [60:03] for a long period of time. [60:06] And you're including valuation in this equation? [60:09] Yeah, we're just going to buy a stock [60:10] for our non-profit endowment now, [60:12] and then in 10 years time we're going to sell it. [60:15] It's an odd trading strategy [60:17] but we're in the non-profit business, [60:19] not the investment business. [60:20] I like Siemens Energy quite a bit. [60:23] I think that you may get a lot of the multiple expansion [60:26] in the next two years, and then the next eight years [60:28] might be less exciting, but— [60:30] We're not even going to look at it over the 10-year period. [60:32] There's a company called Clean Harbors [60:35] which I like a lot, which is they do hazardous waste [60:39] and they own the majority of the incinerators [60:41] in the United States. [60:42] You can't really build more incinerators because of NIMBY. [60:45] And as you have more onshoring, [60:49] there's going to be more hazardous waste, [60:51] and they have both the incinerators [60:53] and they have the network to go collect. [60:56] And so it's just a very, very good business. [61:00] And the starting multiple is pretty reasonable. [61:02] And I think I can grow earnings [61:04] well into the teens for 10 years. [61:06] I would tell you like I think that predicting things [61:11] beyond three years is incredibly difficult. [61:13] I've seen more instances in my career [61:15] where things change in a way you just couldn't expect. [61:22] And that happens all the time. [61:27] So, 10 years is hard. [61:31] Why do you like Siemens Energy so much? [61:33] I think that over the next 5 to 10 years, [61:39] first we started with [61:41] training these AI models, and I think we're still [61:45] in the early stages of that, it's early innings. [61:48] At some point, the scaling laws [61:49] will asymptote out but I think we still have years [61:52] of high energy uses to train models, [61:55] and then there's going to a really long tail of inference. [61:59] And so I think that, [62:02] and interestingly the gas turbine companies, [62:05] being Siemens Energy, GE Vernova, Mitsubishi, [62:09] they're not AI native, right? [62:12] So when they look at this, they say like, [62:14] "Hmm, this is like every other cycle we've seen. [62:16] This is, you know, we're not going to do it again. [62:19] We're not going to build capacity, [62:20] and then all these crazy people [62:22] who are building these data centers [62:23] all of a sudden realize it was a bad investment [62:25] and we're stuck with the capacity." [62:26] So, I think that there's a disconnect [62:29] between what people in Silicon Valley [62:32] think is going to happen in the next 10 years [62:33] and what gas turbine manufacturers think is going to happen. [62:36] I think that gas turbines are going to be in shortage [62:39] for a very long time. [62:42] Because all of the producers are conservative by nature. [62:45] Conservative, yeah, their conservative nature. [62:47] Siemens are literally German, you know. [62:49] And by the way, it's not crazy. [62:50] You look at it, you're sitting there in Germany, [62:52] you're like, okay, Masa's putting in [62:55] tens of billions of dollars. [62:56] Every day there's an announcement [62:58] of a hundred million dollars. [62:58] This sounds insane. I am not getting on board with this. [63:02] I tend to think that the turbine manufacturers [63:04] are more wrong than... [63:07] Look, I think that the electrical grid, [63:10] in the United States and globally, [63:12] people have not invested in the electrical grid at all [63:15] because electricity demand in the United States [63:17] grew at 0% for the last 20 years. [63:20] And so, if you think about the next 20 years, [63:23] I think electricity demand will grow at 4%. [63:26] It doesn't sound like a lot, [63:27] but if you compound 4% over 20 years, [63:30] and you haven't invested for 20 years before that, [63:32] it's very material. [63:34] The second thing with, they have a few businesses, [63:38] but the main ones that are important are gas turbines [63:39] and then products that upgrade electrical grids. [63:43] The other thing about electrical grids is not only [63:45] have they not been invested in, [63:46] as you start to introduce solar farms and wind farms, [63:52] and people having solar on house, [63:55] all of a sudden you need to have a much more dynamic— [63:59] it used to be you have a power plant, [64:00] that power plant sends power out through power lines, [64:03] to houses or businesses, and that's it. [64:05] Now all of a sudden you have solar, wind-producing power [64:10] in far away places that then have to connect to the grid. [64:14] Now the grid is handling [64:17] both consumers actually consuming energy, [64:20] and actually sending energy into the grid. [64:22] So, the grid just wasn't set up to one, [64:25] handle the amount of energy that's going to be consumed [64:27] in the United States and globally, [64:28] and two, it wasn't set up for all these renewables [64:31] which are going to fragment the electrical production, [64:36] in all kinds of different places, [64:38] and require that it's not just like one power plant [64:41] sending it this way. [64:42] It's going to be like a pretty dynamic [64:45] distribution of electricity. [64:47] How should we be thinking of, [64:48] so if we take Siemens as an example, [64:50] so they make the turbines... [64:51] Siemens Energy. [64:52] Siemens Energy. [64:53] The turbines are in short supply, so the price goes up. [64:56] The commodity producer, in this case Siemens, [64:59] does not want to produce more of the commodity [65:00] because of their fear of cycles. [65:02] And, so in theory, the amount they can charge should go up, [65:05] the shortage goes up. [65:07] But then I would assume if they don't produce more [65:09] and demand continues to go up, [65:11] the year at which you would get your turbine [65:13] should also elongate. [65:14] So it goes from '27 to '28 to '29, [65:17] and then you start to ask the question, [65:19] are there other sources of energy we should be using? [65:22] And then, you get into the whole battery thing [65:24] and the whatever. [65:25] The solution to long lead times is long lead times. [65:26] Yeah. [65:27] My question is, as you think about [65:30] what the right price for the stock is, [65:31] is there a point at which like the lead time [65:33] becomes so extreme that... [65:36] Because at the end of the day, [65:37] people don't want the turbines, they want the energy, [65:40] that like other forms of energy [65:42] become more exciting somehow? [65:43] Yeah, I think it's two things. [65:45] I think one, eventually, [65:47] let's assume that Silicon Valley is correct [65:49] and that all this investment is going to be high ROI, [65:52] and therefore they're going to keep doing it. [65:54] Eventually, I think Siemens and Mitsubishi and GE will say, [65:57] "You know what, we have to build some more capacity." [66:00] That will happen. I see. [66:02] I think over a very long period of time, [66:05] nuclear will be probably a more viable option for energy. [66:09] That's probably the best [66:11] in terms of taking everything into account, [66:13] like steadiness of like the electricity, [66:17] environmental factors, nuclear is much better than gas. [66:20] But nuclear plants take a really long time to bring on [66:23] and have all kinds of other complexities. [66:25] So, I think that two things will happen. [66:27] I think over the very long term there will be more nuclear [66:31] but I think that like people miss that, [66:33] like the United States has been pretty early [66:36] in building out these huge clusters. [66:40] Most countries in the world that have developed economies [66:42] are going to need to build out. [66:45] Maybe they don't need to have like the same kind of clusters [66:47] that OpenAI is building in Abilene, [66:49] but they're going to need to have, probably [66:51] for national security reasons, [66:52] their own training clusters, and at a minimum, [66:54] they're going to have their own inference clusters. [66:56] So, I think that everyone's focused on the US, [66:59] and the US is important, [67:01] but this is going to happen everywhere. [67:03] AI is not contained to the US. [67:05] It's going to happen in Japan, [67:06] it's going to happen in Korea. [67:08] So, I think that the tail on this, [67:11] it may be like any kind of, like the internet, [67:14] it may, there may be— [67:15] Ebb and flow. Overbuild and then they... [67:18] But if you look over 20 years, [67:20] I just think that electricity is basically intelligence [67:28] and intelligence is what everyone's investing in, [67:31] and ultimately the bottleneck is going to be electricity. [67:35] So companies that enable electricity production, [67:38] assuming they have some moats, [67:40] are going to be good investments. [67:43] And I would tell you, interestingly, [67:44] going back to the Europe comment, [67:48] Siemens Energy and GE Vernova [67:51] are basically the same company. [67:52] There's no two companies I've ever looked at [67:54] that are more similar. [67:56] They have almost the same revenue, dollar revenue, [67:59] I think it's like 45-something. [68:00] The products they sell are pretty equivalent. [68:01] The products they sell are pretty equivalent. [68:03] Like, nobody would say like this one's dramatically... [68:05] They're in the same business as turbines, electrical grid, [68:10] and then wind energy. [68:12] GE Vernova trades at 2x the enterprise value [68:17] of Siemens Energy. [68:18] Just to give you a sense of like, [68:19] and ultimately there's no reason why the margin [68:23] should be higher at GE Vernova. [68:24] The revenue base is the same. [68:26] So, I think the Europeans are implicitly [68:29] just a lot more skeptical of the sustainability [68:34] and longevity of this AI cycle. [68:38] And that has to narrow one way or another. [68:43] I don't think you can have two companies [68:45] and one trade at half the price of the other one, [68:47] and there's nothing about it that's actually worse. [68:49] There's also the other open mystery, [68:51] which is I believe the onshore TSMC trades [68:55] at a 20% discount to the US ADR. [69:00] What's going on there? [69:01] How is that possible? [69:03] That is also extreme, [69:08] and it just shows you [69:10] the US equity markets either have more liquidity [69:15] or just a different perception [69:16] of the risk inherent in buying TSMC. [69:20] And 20% is quite meaningful for the same company. [69:24] But Siemens energy is 50% discount. [69:27] Now, I understand TSMC is the same company, literally. [69:29] Yeah. [69:30] But like these are pretty close. [69:32] Well, okay, you raised different regions [69:35] trading at different multiples, and you raised Asia [69:40] which is a great segue. [69:43] Asian multiples. [69:44] Yeah, everything in China is really cheap right now. [69:45] You know, ByteDance trades at a really low multiple, [69:48] Alibaba trades at a really low multiple. [69:49] Do you guys do much in China? [69:51] Yeah, what's going on there? [69:52] Is that just a value trap? What's going on there? [69:54] Yeah, we stopped investing in China about three years ago. [69:58] So, look, everything has a price, right? [70:02] I think that the reason we stopped investing there [70:04] is, fundamentally, I have an issue [70:08] with the way their economy works. [70:10] I think the government has way too much influence [70:12] on how resources are allocated and— [70:14] Sorry, you have an issue as a moral matter [70:18] or you just have an issue as a practical, capitalistic... [70:20] Practical, capitalistic. [70:22] They can, you know. [70:23] I'm not making moral judgements on China. [70:25] I personally don't think it's the best. [70:26] You're just saying you don't know how to predict [70:28] how the market will perform. [70:29] Look, I think that what markets don't like is uncertainty, [70:33] and the problem with China is twofold. [70:35] One is that the government has way too much influence [70:38] on how resources are allocated. [70:41] I don't think they're particularly good [70:42] at making those decisions as you saw during COVID [70:46] when they shut everything down for no reason for, you know. [70:49] And they will arbitrarily decide [70:51] which industries are important [70:54] and which ones are not important. [70:55] And if you make too much money, that's bad. [71:00] And so all of those factors, you know, [71:02] for a while we were seeing Chinese internet companies, [71:04] they didn't want to beat earnings [71:06] because if you beat earnings, your stock goes up too much. [71:10] You might be flagging that, like, [71:13] you're over earning and you know... [71:16] Things have a target. [71:17] If you look at a company. [71:20] If you know they're not going to beat earnings, [71:22] what are you left with? [71:23] You're left with a distribution [71:25] which is not particularly attractive. [71:27] And so I think that the economy there is, [71:31] it's just very, very difficult. [71:34] That being said, I think the valuations are extremely cheap, [71:38] and I think the Chinese people [71:41] are probably the most commercial, hardworking, smart. [71:48] If you look all over Southeast Asia, [71:51] you know the best companies are usually run by Chinese. [71:55] Chinese people are incredibly industrious, [71:57] incredibly smart, incredibly commercial. [71:59] So, I find it kind of sad that [72:03] what's happening in China is stifling innovation [72:06] that probably would be better than... [72:08] Before this all happened, if we went back 10 years ago, [72:12] I think we would probably all sit here and say, [72:15] these Chinese internet companies are actually [72:17] probably technologically ahead of where we are. [72:20] Like, TikTok was... obviously came out of China. [72:23] Like, WeChat was like really probably [72:26] the most advanced form of social media globally. [72:30] But, you know, the way governments behave [72:34] is incredibly impactful to companies. [72:37] So you think the tech is very cool, but just fundamentally [72:39] being an equity holder there is hard? [72:40] The tech is very cool. [72:42] I just don't think it's progressing [72:44] at the same rate it used to [72:45] because a lot of the great entrepreneurs have left [72:48] because they felt like the environment [72:50] for starting companies, for [72:54] driving new innovation that [72:55] was going to make a lot of money, [72:57] like that wasn't something that was [73:00] looked favorably upon by the government. [73:01] So, the best people left. [73:03] A lot of them went to Singapore. [73:04] And that to me, I think that's a shame [73:06] because with the right government [73:09] and with the right economic system, [73:12] China should be [73:16] the most important economy in the world, [73:19] and it should be a technological leader, [73:21] and I think the geopolitical tensions [73:23] would be entirely different, [73:25] and I have a ton of respect for [73:27] what China did from 1970s to 2020. [73:31] I think it's pretty much an economic miracle. [73:33] But what we've seen the last five years is, [73:36] governments matter and [73:38] it matters what your political system is, [73:40] and it matters how the government intervenes, [73:42] and due process matters. [73:43] There's the famous case where many [73:46] of the digital education apps were banned overnight. [73:51] Did that really update your view of China? [73:52] Yeah, that like, investors hate uncertainty. [73:55] And you want to, okay, like in the US, [73:57] can something bad happen to social media companies? [74:01] They get called in front of Congress, [74:03] but there's due process, right? [74:05] Like, you know that like there is a system [74:07] that you can't have, [74:09] Donald Trump can't just wave his hand and get rid of Meta [74:12] because he doesn't like the fact that [74:13] they're allowing people [74:15] to say bad things about him. [74:16] And so, when you have a government [74:18] that can act capriciously, [74:21] that has a humongous gravitational pull on valuations. [74:26] Some big picture questions related to China. [74:28] So, there's all these charts one could look at [74:31] of the concentration of Mag 7 into the S&P 500, [74:35] and you could look at the 1970s equivalent [74:36] of the Nifty Fifty. [74:39] There's the inflation, which seems to be tracking [74:41] the double hump story. [74:43] First of all, do you [74:45] care at all about these analogies? [74:47] Like, do these analogies mean anything to you? [74:48] These charts with all the JPEG artifacts on them. [74:50] They're always very convincing. [74:52] I think there's, like, look. [74:54] History rhymes, right? [74:56] And the same thing in the market, [74:57] so there is like, it's interesting to look [75:01] at what's happened historically to these companies. [75:04] It's dangerous to say, like, [75:06] look at it and just extrapolate to the point [75:10] where you say— So you can overcorrect [75:10] on the chart. [75:11] But then my question would be, [75:13] what would be the sign for you actually [75:15] that the bubble is close to peaking? [75:18] Is there something, an anecdote [75:20] or an actual piece of evidence you'd look at, [75:21] you'd say like, "Okay, maybe we all believe in AI, but we—" [75:24] So you're saying AI is a bubble? [75:26] No, I'm saying I believe in cyclicality. [75:28] Like, the internet worked but it was also a bubble. [75:31] It was also a bubble, yeah. [75:32] Derek Thompson said this recently in a way that I liked [75:34] which is people associate using an AI, [75:38] that AI may, currently be or in future be a bubble [75:43] as like some kind of negative statement. [75:46] But because every major tech change— [75:49] the canals, the railroads, the internet, whatever— [75:52] has been accompanied by a huge speculative bubble. [75:54] Because it has to [75:55] when you think about it. It has to. [75:56] Yeah. [75:58] Because Americans are optimistic. [75:59] Right, but it's an AI-optimistic statement to make. [76:02] Yes, that's right. [76:02] That there will be a bubble. [76:03] We are actually celebrating how important it is. [76:04] So Daniel's saying he's an AGI believer, [76:06] and AI's clearly going to be a big thing. [76:08] Therefore, is it this year, is it next year? [76:11] Like, when's the bubble? [76:12] If you look at like the checklist of things, [76:14] you want to like look at for like what's going to be closer [76:17] to the end of the bubble. [76:18] If you went back like a year ago, you'd be like, okay, [76:20] yes there's a lot of money being spent, [76:24] but these companies have massive amounts of cash flow. [76:26] Microsoft kicks off a ton of cash flow. [76:28] Once you start having debt-fueled investment, [76:32] that's usually a bad leading indicator. [76:36] Because, obviously, when you have a lot of debt, [76:38] there's not much room to make mistakes. [76:43] And you are seeing, like, some of these latest projects [76:46] are debt finance. [76:48] I think that nobody knows, like, I think even if you spoke [76:51] to... I know when Dario was sitting here, [76:54] it's like they keep investing in training, [76:58] and maybe it's pre-training, it's post-training, [77:00] but at some point the returns on that training [77:03] are going to be disappointing. [77:06] And then it's that people will be freaked out a little bit, [77:09] but usually then the next time it's like back to, okay, [77:10] the next model's good. [77:13] At some point people are going to say like, [77:14] "Well, we're asymptoting out. [77:16] The returns on this new investment [77:18] are actually not working." [77:20] That's when I think you'll see a pretty big correction. [77:26] Okay, so there's the theory that there's a correction [77:28] because the returns to pre-training start asymptoting, [77:32] and then there's a theory of bubbles. [77:34] If you look at the 1970s, and to some extent dot com, [77:37] which I think was really hampered [77:38] by the fact that the Fed raised rates into '99 and 2000. [77:43] There's another view of bubbles which is, [77:44] there's an awesome thing happening, [77:46] everyone gets overexcited, [77:47] but then some random thing hits the bubble, [77:51] and because things have run up so much, [77:53] everyone starts panic selling all the way down [77:54] because there's all the people that bought in, [77:56] their cost bases of, you know, NVIDIA [77:58] not to pick on anyone, is like 4 trillion. [78:00] And so my question is, do you think we're in that mode, [78:04] and is this a productive line? [78:06] Like, at D1, do you guys ever think, [78:09] oh, what could be that thing, [78:11] and how do we prepare for that thing? [78:12] Or do you sort of think like— [78:14] I think to our detriment. [78:16] We should have owned more AI stocks. [78:18] There are definitely things, [78:19] you go down the list of what constitutes a bubble, [78:22] and you can check some things like [78:26] massive debt-fueled investment, [78:28] just like people were building— [78:29] What else is on that checklist? [78:31] Valuations, [78:34] and bad companies trading crazy valuations. [78:37] Circular financing. [78:38] I would tell you like NVIDIA is, [78:40] we can have a debate about what the earnings [78:42] are going to be in a few years, but it's not expensive. [78:46] Like, NVIDIA trades at 20-something times multiple. [78:51] That's, I think, within reason [78:55] and I don't think there's anything I see [78:57] in the public markets which is bubble-like [78:59] from a valuation standpoint. [79:02] I think if you look back in history, [79:03] 70% of the time you have this [79:06] major breakthrough technology, [79:07] there is a stock market bubble. [79:09] And maybe what you're seeing right now [79:10] in some of these retail stocks, like where [79:13] Opendoor goes from 1 to 10. [79:14] Maybe that is like things are starting to, [79:17] you know, bubble. [79:19] But we haven't seen like large cap... [79:23] I don't think the large cap AI stocks [79:25] are trading at crazy valuations at all. [79:27] We're not seeing that yet, [79:29] so maybe it's 1997, 1996 and by the time that [79:37] we're in the equivalent of 1999, NVIDIA [79:39] will be three times higher, and it's possible. [79:43] Someone said to me recently as well that [79:44] for a proper kind of crisis you also need things [79:49] that people thought were safe to not turn out to be safe. [79:52] Like in 2000 when the NASDAQ went down 85%, [79:56] it's like well that's a real bummer, [79:57] but we did know that we were buying these [79:59] highly-gassed tech stocks. [80:01] Whereas, when the debt actually turns out to not be safe... [80:05] The debt turns out to be equity. [80:06] Exactly. Yeah. [80:08] That's when you get real issues. [80:10] It was very different in 1999, though. [80:11] Like, there were horrible companies [80:13] which had no real economic prospects [80:15] trading crazy valuations. [80:16] I'm just not seeing that in the public markets now. [80:19] I am, just not in AI. [80:23] You're a huge fan of SpaceX and a big holder. [80:26] What are you so excited about? [80:27] What's going on there? [80:29] Everyone understands Elon is, you know, [80:31] an amazing inventor, an amazing entrepreneur. [80:35] I think people underestimate [80:36] how good of a business person he is in that like— [80:39] Been my theory, yeah. [80:40] It's like okay, yes, does he invent great things? [80:44] He does, but he is ruthless about bringing down costs [80:49] to a point where his business becomes [80:52] a natural monopoly [80:54] because it is the low-cost provider. [80:56] With SpaceX, the whole problem [80:58] with space in general, doing anything in space historically [81:01] was that it was very expensive [81:04] to launch anything into space because, one, [81:06] the rocket blew up and, therefore, [81:11] it better be really high value [81:13] if you're going to send it up there, [81:14] because if you're going to take a plane [81:16] from New York to LA and every time you do it, [81:18] the 747 blows up, the plane tickets [81:20] will be really expensive. [81:21] You better really want to go up to LA, right? [81:23] I think, so the first thing is [81:25] the idea of bringing down the cost dramatically [81:27] by making things reusable. [81:28] Now I think like he was like 5 to 10 years [81:32] ahead of everybody else. [81:32] Now, obviously, it's more reusable [81:36] and the scale of the rockets is getting so big, [81:41] and each rocket will be able [81:43] to be flown more than a hundred times. [81:44] And so, the cost per ton of launching things into space [81:49] is going to go down by relative to [81:53] five years ago, it's going to go down like 99.9%. [81:55] It's dramatic, the cost. [81:58] Once you bring down the costs that way. [82:02] Obviously, the first order effect [82:04] is like you launch a bunch of LEO satellites, [82:07] which gave us this great Starlink. [82:09] And I think that will continue in a way [82:13] that probably surprises people [82:15] in that I think SpaceX will probably capture [82:18] more of the global telecommunications market [82:21] than people expect. [82:22] You saw them buying Spectrum from EchoStar, [82:24] but then it's like, okay, so if you really step back [82:29] and say like, "Well what else should we be doing in space [82:31] if the cost per ton is like dramatically lower?" [82:34] I value the company based on the launch business, [82:38] which is a near monopoly and a fantastic business, [82:41] and then our expectation, like what happens with Starlink, [82:44] which is just a matter of like P times Q subs. [82:47] But when you bring the cost per ton down so dramatically, [82:51] you open up markets— There's option value. [82:53] They would say, yeah, you fly from LA to Sydney [82:55] in 30 minutes, or we start having solar panels [82:58] in space that are, you know, [83:01] people say data centers in space. [83:03] I don't know what it's going to be. [83:04] Certainly there will be US defense deployments [83:08] in space, right? [83:10] For sure. [83:11] And then I think there's a lot of option value. [83:13] I look at it and say like, [83:14] I think the business can triple without having [83:17] to make any big bets about people traveling [83:21] 30 minutes across the world or [83:23] data centers in space. [83:25] But there are legitimate, like a lot of medication [83:27] is better manufactured in space than it is on Earth. [83:30] There will be plenty of allocation over long term, [83:33] and that's why I think like it's the equivalent [83:35] of Tesla's Robotaxi. [83:37] There is this option value out there that, [83:40] no, Tesla's not an EV company. [83:42] It is going to own global transportation [83:44] because it's a low-cost provider and it's, you know, [83:47] and so I think you have that with SpaceX. [83:50] I like it because you have the downside of the cash flows [83:53] from Starlink and the launch business, [83:55] but you still have the option value [83:56] of all the things we could dream about, [83:58] which are impossible to quantify. [84:00] I totally agree on the underappreciated aspect of Elon [84:04] being the business savvy where, [84:06] let's not forget Tesla invented a new mode [84:07] of selling cars direct to consumer. [84:09] They had to get the laws changed in certain states [84:11] because the current dealership system was enshrined, [84:14] similarly with SpaceX. [84:15] For what they do, [84:17] it's actually a very capital-efficient business [84:19] and they've run it in a, you know, [84:20] they've built it with profits [84:22] rather than venture funding, [84:24] or they've built it with revenues, [84:25] customer revenues rather than venture funding. [84:26] And so, I think that's really underappreciated. [84:28] I think he's a better business person than he is inventor. [84:30] He's amazing at both, [84:31] but like people don't understand, [84:33] like he just naturally understands, [84:35] just get the cost down, get the cost down, [84:37] get the cost down. [84:38] And that's very hard to compete against. [84:40] Do you want another? [84:42] Sure, I'll have another. Sure. Great. [84:44] You keep going there. [84:46] You know, they say Stripe is actually controlled [84:48] by an Irish mafia behind the scenes. [84:52] Did you actually, I mean, it's quite interesting, [84:54] I didn't realize until you said it [84:56] that D1 is mostly privates now, by size. [85:00] Yes, by size. [85:01] Did you expect that to happen? When you started? [85:05] No, I did not. I did not expect that to happen. [85:09] It's actually just more a function [85:11] of how things played out than a deliberate strategy [85:15] for us to grow the private business [85:18] at the expense of the public business. [85:19] Our public business went through some big ups [85:22] and some big downs. [85:23] I think our public business... We are, [85:26] and look, there is a limit to how much we can manage [85:29] on the public side. [85:30] We announced a few weeks ago that at the end of this year [85:34] we're going to be closing the hedge fund. [85:36] We may replace redemptions, [85:37] but the most important thing is in our business, [85:41] returns are negatively correlated with size. [85:44] Especially on the short side. [85:46] Because you just own too much of a company to, like, [85:47] you move the stock when you try to trade it and stuff? [85:49] It's not so much on the long side, it's on the short side. [85:51] As you turn things over. [85:53] And it's like, if I'm $30 billion, [85:56] you can't really short $5 billion companies, and most— [85:59] Could you not grow AUM and be more long exposure? [86:04] Probably what you'd want to do is, [86:06] and what we may do in the future is, [86:07] and what other funds have done, [86:09] is you have your hedge fund, which does long, short, [86:12] and then you have a separate fund [86:13] which is like just long only, [86:16] and that can scale quite a lot. [86:19] because we tend to hold them longer [86:21] and you can buy very large cap companies. [86:22] So, privates. [86:24] So what was your first private position for D1? [86:28] That's a good question. [86:29] We launched, I think our first private position [86:33] at launch was I rolled a stake. [86:36] I had taken a stake starting in like 2010 [86:40] in a company called Lineage, [86:41] which is a cold storage warehouse company [86:44] that I rolled off of my personal balance sheet [86:47] into the fund, and then we made subsequent investments. [86:50] And then in the first couple years, [86:55] we did a variety of different things. [86:58] Like, some I think worked out not as well. [87:03] JUUL was like a, I'm not quite sure if that was, [87:06] it was good and bad at times. [87:09] I think it's going to be good now. [87:10] And then, we did Ramp really early. [87:13] That was good. [87:14] John was nice enough to allow us to invest [87:17] in Stripe in 2019. Yeah, how did you [87:19] underwrite Stripe? [87:20] Because again, we weren't profitable at the time, [87:22] and I don't know, yeah, what does the process look like [87:24] for something like that? [87:25] I actually started out my career [87:28] as a financial service analyst. [87:29] So, I had looked at all the card networks, [87:33] all the processors. [87:34] It was pretty clear to me that the competitive set [87:36] in merchant processing was very mediocre at best, [87:41] and that their technology was not conducive [87:46] to most internet companies. [87:50] Nor did they have the tech stack that would allow them [87:52] to adjust to what was happening [87:54] in terms of e-commerce. [87:55] And, look, at the end of the day [87:58] it was like, I think that there's going to be one, [88:01] maybe two companies that actually can provide the technology [88:06] for companies to enable e-commerce [88:09] or any online transactions, [88:12] and you guys seem pretty smart. [88:16] We do a lot of work on management teams [88:19] and, you know, huge market, great management team, [88:23] weak competitive set, is like, [88:29] perfect recipe for making a lot of money. [88:31] So you've got Ramp, Stripe, SpaceX, [88:38] and maybe one day of course [88:39] all these companies go public, [88:40] but it would seem as if there's like a lot [88:42] of later-stage private companies now. [88:45] Why do you think that's happening, [88:47] and where do you think that's going [88:48] to go in a couple of years? [88:51] Are public markets just going to be the laggards [88:54] and all the new hot stuff will be private? [88:56] Or will it rebalance one day? [88:57] I mean, if I ran a private company like Stripe, [89:00] I wouldn't go public. [89:02] I think the public markets— [89:03] Which is kind of ironic [89:05] because you're a public markets investor. [89:06] Yeah, I think the public markets [89:07] are kind of problematic at this point. [89:09] Let's just take Stripe for example, [89:11] and I won't speak for John. But basically, [89:14] Stripe grows earnings, cash flow at some amount. [89:18] Value compounds, and they do tender offers, [89:22] and the tender offers are relatively in line [89:24] with the value creation, [89:25] and therefore the people who are working at the company [89:28] and are creating that value, get paid for that value [89:30] because the stock price goes up in line with value creation. [89:33] Now what we see in public markets [89:35] is you take your company public and, [89:39] depending on what the retail crowd is doing that day, [89:43] the stock may trade at some insane value, right? [89:47] And most people are high-fiving. [89:49] "This is amazing. [89:50] Our stock is like trading 2x where it should be. [89:54] This is great. We're all rich." [89:55] The problem with that is that— [89:57] Cuts both ways. [89:58] You've now pulled forward a ton of value. [90:02] And so all the people working at the company now [90:04] are being overpaid because [90:07] they didn't actually create this value, [90:08] that's not... [90:11] and then the people who you're hiring, [90:13] those people are probably more likely to just cash out [90:16] because they've just made too much money. [90:17] You're robbing future employees to pay current employees. [90:20] Exactly. [90:21] And then your future employees, [90:22] now you have to give them stock options [90:24] or RSUs at a stock price you don't really believe in. [90:27] And so, the stock is so volatile that [90:31] you're actually not being paid [90:33] as an employee based on value creation. [90:35] You're being paid arbitrarily based upon multiples [90:38] which have nothing to do with the true intrinsic value [90:40] of the company. [90:41] So, obviously it's bad [90:44] to be undervalued as a company, [90:46] because then you're issuing stock to employees [90:47] at too low of a value, [90:49] and then they don't appreciate it usually. [90:52] But it's pretty bad to be overvalued, too. [90:54] Because like, employees, [90:57] if a stock doesn't go up, [90:58] they will definitely come back to you [90:59] and ask for more options. [91:01] If the stock goes up way more than it should, [91:02] they're not going to come back to you and be like, [91:03] "Oh, you know what? [91:05] Hey, I made too much money." [91:08] And so you end up having this asymmetric... [91:12] I think it's really not a healthy dynamic [91:14] to be a public company. [91:15] Is there anything that should be changed [91:17] about the public markets to make it better? [91:19] So, for example, [91:21] Robinhood got rid of commissions. [91:24] Is zero the correct amount of friction [91:27] for entering and exiting trades? [91:29] This is hard. [91:30] I had breakfast with Vlad this morning [91:31] and I really like Vlad. [91:32] I think that it's a moment in time. [91:35] My view is that over the long term, [91:37] stocks will go to intrinsic value. [91:39] It's taken longer than I've expected [91:41] for some of these things. [91:44] Some of the shorts you were— [91:45] I still believe it. I believe it. [91:48] I can't tell you I have a lot evidence that's the case, [91:51] but that will happen. [91:53] That doesn't necessarily help a company like Stripe [91:55] if they go public and if eventually in five years [91:58] it's in fair value, but in the meantime [91:59] they're just kind of like... [92:00] Whips up and down. [92:01] Yeah, whips up and down. That's bad. [92:02] I'm not sure you can do anything to change markets. [92:07] Markets are inherently volatile. [92:09] It just so happens that at this moment they are more, [92:13] like, you would think that in the current world [92:17] if I told you that like we just have perfect information, [92:19] everybody has all the information, [92:22] it's at your fingertips. [92:24] Everything should be more efficient. [92:25] Yeah, yeah. Wow, stocks are going to be so correctly priced. [92:27] It should be correctly priced [92:28] because everyone has access to the same information. [92:31] It's actually gotten less efficient over time, for sure. [92:35] And I don't know. [92:36] I don't think you can just necessarily fix that. [92:38] You mentioned starting your career as a banking analyst. [92:40] How has the banking industry changed? [92:43] Most of the banks [92:47] tend to be very dominant in one geography. [92:53] They're not like tech companies like Google or Meta [92:56] where they're just dominant. [92:57] Like, JP Morgan is dominant in the US, [92:59] but like Europe doesn't matter. [93:00] Same thing with the European banks. [93:02] Up until now, I think the legacy banks have more or less [93:07] in most geographies been able to keep their market share. [93:13] However, you increasingly are seeing banks [93:16] like Nubank or Revolut that don't have the tech debt [93:21] of like mainframes and old code. [93:24] And just offer better customer experiences. [93:26] Don't have branches. [93:27] Iterate on product faster, have better engineers. [93:30] And I think that those banks are going [93:35] to increasingly take market share. [93:37] Neobanks have happened in Brazil with Nubank, [93:39] have happened in Europe with Revolut and Monzo [93:42] and people like that. [93:44] Haven't happened in the US, really, [93:46] and there's probably other geos where they have and haven't. [93:50] Do you have a view on will it happen in all markets? [93:54] Are certain markets more impervious than others? [93:57] I think that it depends how good [93:59] the incumbent bank is. [94:01] Like, I think JP Morgan is a very well-run bank [94:03] and the big banks are well-run, [94:05] but do I think that they are vulnerable to disruption? [94:10] Definitely. [94:11] But you still buy bank stock, [94:12] so how do you get comfortable? [94:14] Then there's like a more theoretical question of, [94:15] okay, well if all these AI agents, [94:20] if before Revolut and Nubank just hired the best engineers, [94:23] and so they were just naturally going to beat JP Morgan. [94:26] But if like AI agents make, you know, [94:28] now everybody has the best engineers [94:30] because best engineers are actually not people, [94:32] they're just agents. Maybe JP Morgan can be as good [94:35] as some other companies. [94:36] That's theoretical. It's probably not correct. [94:38] In the US I haven't seen someone come in [94:40] and be that disruptive. [94:42] But is that because of market structure reasons, [94:44] or just we haven't seen the great founder yet? [94:48] It's a good question. [94:49] I mean, banking is not, it's not like, [94:52] you have both sides. [94:53] Like you need deposits, you need to provide credit, right? [94:56] And you need a lot of scale. [94:58] And a market like the US is much more difficult [95:01] to penetrate because it's so big [95:04] and the competitors have so much capital to invest. [95:07] Whereas smaller countries, [95:10] like Revolut has low single-digit [95:13] market share in every country. [95:15] I think that'll keep growing, [95:17] but they don't actually provide credit. [95:18] So, I think we're in the early days of disruption. [95:21] I think you roll forward like 20 years, [95:25] there's going to be some companies [95:26] that didn't exist 10 years ago [95:29] but are going to become enormously large banks. [95:32] But I also think the incumbent banks [95:34] are probably going to innovate enough [95:36] that they're not going to [95:39] go the way of like JCPenney. [95:41] Last question. [95:41] If you're a youngster interested in investing, [95:44] you have views on companies but you don't feel confident yet [95:47] in how to underwrite and construct a model [95:50] and things like that, what advice would you give them? [95:53] Someone who's interested in this stuff [95:55] but still getting their feet wet? [95:56] I think that pretty much with [96:00] anything you want to do in life, [96:01] I just believe that reading, just incessantly [96:06] is the way to get ahead. [96:09] And investing's so different. [96:12] It's just like, read stock pitches, [96:15] over and over and over again, [96:18] and then watch those stocks, see how things play out. [96:20] The market will be your mentor. [96:24] But I think you need to just like, really, like... [96:27] I learned by, there's a website called Value Investors Club, [96:30] and like, I just read everything that people pitched. [96:33] You want to have some framework. [96:35] I liked reading Buffett's books because, [96:39] he has some faults obviously, [96:40] but he just he has a way of like distilling down [96:43] the complicated into very simple ideas. [96:46] And so I would read a lot of Buffett books, [96:48] I'd read a lot of stock pitches. [96:50] Anything I could get my hands on with regard [96:51] to like what's happening in technology, [96:54] what's happening in the economy. [96:55] And to me it's just like, the more you read, [96:57] the better you are. [96:59] Obviously, the Buffett stuff is very worth reading, [97:02] and there's the Cunningham book [97:02] that takes all the letters [97:05] and smushes them together because there's [97:06] a lot of repetition. That's my favorite book. [97:07] Exactly. Yeah, yeah. [97:09] So that is a classic and a lot of people listening [97:10] have probably read it. [97:11] So, all the Berkshire letters. [97:13] What I actually only read recently for the first time [97:15] is if you go back and read the Buffett partnership letters, [97:19] so this was, the partnership he had [97:21] with which he bought Berkshire Hathaway [97:23] and turns into, but this was like a fund, [97:25] more of a hedge fund than [97:27] the C corp that is Berkshire. [97:29] What's interesting is I find it [97:30] stylistically very different. [97:31] This is the late '50s, early '60s [97:34] and it's before he got so polished. [97:37] You know, it's before he got so folksy and approachable [97:41] and careful in what he said, [97:44] and a little more of the raw ambition is on display [97:48] before he sanded that off. [97:50] And it's, I don't know if you've gone back and read them. [97:52] I haven't read them. [97:53] Oh, it's awesome. I'll send it to you. [97:54] It's really good reading. [97:55] But the original Buffett partnership letters are, [97:59] I mean obviously it's Buffett so it's similar in a way [98:01] to the Berkshire letters, [98:02] but I actually think they're better in certain ways. [98:04] Yeah, he is a brilliant guy, but he actually does like [98:08] to portray himself in a certain light. [98:10] The Buffett partnership letters felt more authentically [98:10] Buffett to me. Yeah, they shine a light [98:13] on like who he truly is, but it's been good for his business [98:17] to portray himself in that light. [98:18] It makes sense. [98:19] You can't argue with the results. Yeah. [98:21] All right, Dan, Daniel, thank you, guys. [98:23] All right. Thank you.