首页  >>  来自播客: Motley Fool Money 更新   反馈  

Motley Fool Money - When Leverage Goes Wrong on Wall Street

发布时间:   原节目
以下是内容的中文翻译: 《莫特利·富尔隐藏瑰宝投资》本期节目涵盖了多项重要的市场事件和投资考量,其中包括一家对冲基金的戏剧性崩溃、超大规模(hyperscaler)公司盈利表现的分化,以及一个比较不同行业领导者的“你更愿意选择谁”(would you rather)环节。 **情境感知对冲基金的追加保证金通知** 讨论始于本周的头条新闻:人工智能对冲基金“情境感知”(Situational Awareness)收到了追加保证金通知。该基金由Leopold Aschenbrenner(前FTX和OpenAI员工)创立,资产从数亿美元增至200亿美元,上半年回报率高达440%,并严重依赖杠杆(据报道为4倍)。然而,人工智能基础设施股票(如美光、SK海力士)的下跌,以及其针对软件公司的空头头寸反转,导致了巨额亏损。主持人解释说,在4倍杠杆下,资产价值下跌25%即可抹去所有股本。这触发了银行的追加保证金通知,导致资产被迅速出售给Citadel。尽管Leopold保留了Anthropic的50亿美元股权,但这一事件生动地提醒了杠杆的危险性。主持人强调,个人投资者应避免使用保证金交易,并指出当“拥挤交易”解体时,市场动态往往涉及其他基金“嗅到血腥味”。在这种情况下,Citadel并非完全被视为反派角色,而是一个通过阻止800亿美元资产更大规模、更无序的清算来稳定市场的实体。对个人投资者的主要启示是:关注基本面,理解大型机构投资者的运作方式不同,并抵制在市场低迷时恐慌的冲动,因为通常“最好的行动就是不行动”。 **超大规模公司盈利分化** 播客接着讨论了近期主要科技公司的财报,注意到股票表现出现显著分化:微软和亚马逊股价大幅上涨,而苹果和Meta则下跌。Lou Whiteman认为,市场对这些公司的主要问题是“何时实现投资资本回报率(ROIC)?”。微软和亚马逊更好地阐明了其人工智能投资的变现路径,并拥有多元化的收入来源。相比之下,Meta尽管拥有庞大的用户基础,但在其人工智能支出回报方面,市场对其信任度较低,尤其考虑到其过去的元宇宙投资。Meta的人工智能投资大多被视为仅利好其核心广告业务,而非像亚马逊网络服务(AWS)或微软Azure那样,创建广泛且可变现的平台,后者正受到Anthropic等公司的需求驱动。尽管市场目前对某些公司(例如微软每年1750亿美元的资本支出)增加资本支出表现出热情,但人们普遍担心“前沿模型”(frontier models)能否带来足够的回报率(ROIC)。 **“你更愿意选择谁”投资选择** 主持人就以下行业配对进行了“你更愿意选择谁”的游戏: * **特斯拉(11倍销售额,160倍市盈率)对比通用汽车(1倍销售额,6倍市盈率):** Lou和Jason都选择了特斯拉,尽管其估值很高。他们的理由是特斯拉在核心汽车业务之外的“期权价值”(optionality)以及埃隆·马斯克的过往业绩,尽管他们承认通用汽车近期在股票回报方面表现出色。 * **礼来(14倍企业价值/销售额,40倍市盈率)对比诺和诺德(3.5倍企业价值/销售额,11倍市盈率):** 两位主持人都倾向于礼来。他们指出礼来在GLP-1减肥药(Zepbound、Mounjaro)领域拥有强大地位,以及其利用重磅药物的现金流,通过大量交易积极构建未来产品管线的战略,即使它并非“价值型”投资标的。 * **摩根大通(2.5倍市净率,三年增长率8%)对比SoFi(1.9倍市净率,三年增长率27%):** Lou和Jason都更青睐摩根大通。他们对SoFi成熟后能否保持高增长持怀疑态度,并质疑其报告的每客户平均建立1.5个“关系”作为变现指标的有效性。他们更看重摩根大通的规模、杰米·戴蒙领导下的强大管理团队、稳定的股息、股票回购和稳健的资产负债表。 * **Costco(48倍市盈率)对比Target(19倍市盈率):** 两位主持人都选择了Costco。他们强调Costco高粘性的会员模式(续订率超过90%)、忠实客户群和明确的价值主张,并将其与Target在竞争激烈的零售市场中不那么明确的“特色”形成对比。 **特斯拉出售中国业务/SpaceX合并传闻** 《华尔街日报》报道称,特斯拉可能出售其中国业务,以促成与SpaceX的合并,但埃隆·马斯克称此传闻为“假新闻”。Lou和Jason讨论了此举背后的逻辑:将SpaceX这样的国防承包商与一家在中国拥有大量业务的公司整合,可能会面临美国政府的严格审查。他们猜测马斯克可能旨在在本届政府任期内完成此类合并。此举也将标志着特斯拉身份的演变,从电动汽车制造商转向人工智能和机器人领域。 **关注股票** * **是德科技(Keysight Technologies, KEYS):** Jason Moser的选择。是德科技提供用于通信、航空航天、汽车、半导体等多个领域的技术测试和部署的硬件、软件和服务。其软件和服务现在占收入的36%,提供更高利润率的经常性收入。 * **L3哈里斯科技(L3 Harris Technologies, LHX):** Lou Whiteman的选择。这家主要国防承包商的盈利超出预期并上调了业绩指引,但在宣布其导弹业务首次公开募股(IPO)推迟(至2027年)后,股价下跌了10%以上。Lou认为此次抛售反应过度,指出L3哈里斯拥有420亿美元的积压订单、专注于高利润的国防电子产品,以及更强的国际影响力,因此,目前估值与通用动力(General Dynamics)等同行相当,是一个有吸引力的买入机会。

This episode of Motley Fool Hidden Gems Investing covered several significant market events and investment considerations, including a hedge fund's dramatic collapse, diverging hyperscaler earnings, and a "would you rather" segment comparing different industry leaders. **Situational Awareness Hedge Fund's Margin Call** The discussion began with the week's headline news: the margin call for Situational Awareness, an AI-focused hedge fund founded by Leopold Aschenbrenner (formerly of FTX and OpenAI). The fund, which grew from a few hundred million to $20 billion in assets and saw returns of 440% in the first half of the year, relied heavily on leverage (reportedly 4x). A downturn in AI infrastructure stocks (like Micron, SK Hynix) and a reversal in its short positions against software companies led to significant losses. The hosts explained that with 4x leverage, a 25% drop in asset value can wipe out all equity. This triggered margin calls from banks, leading to a rapid-fire sale of assets to Citadel. While Leopold retained a $5 billion stake in Anthropic, the event served as a stark reminder of the dangers of leverage. The hosts emphasized that individual investors should avoid margin, and that market dynamics often involve other funds "smelling blood" when crowded trades unravel. Citadel, in this scenario, was viewed not necessarily as a villain, but as an entity that potentially stabilized the market by preventing a larger, more disorderly liquidation of $80 billion in assets. The key takeaway for individual investors: focus on fundamentals, understand that large institutional players operate differently, and resist the urge to panic during downturns, as often "the best action is inaction." **Hyperscaler Earnings Divergence** The podcast then moved to the recent earnings reports from major tech companies, noting a significant divergence in stock performance: Microsoft and Amazon were up substantially, while Apple and Meta were down. Lou Whiteman argued that the market's primary question for these companies is "when ROIC?" (Return on Invested Capital). Microsoft and Amazon have better articulated a path to monetizing their AI investments and have diversified revenue streams. In contrast, Meta, despite its vast user base, has less credibility with the market regarding the return on its AI spending, especially given past metaverse investments. Its AI investments are largely seen as benefiting its core ad business, rather than creating a broad, monetizable platform like Amazon Web Services or Microsoft Azure, which are seeing demand from companies like Anthropic. While the market is currently enthusiastic about increased CapEx for some (e.g., Microsoft's $175 billion annual spend), there's an underlying concern about whether the "frontier models" will deliver sufficient ROIC. **"Would You Rather" Investment Choices** The hosts played "Would You Rather" with various industry pairs: * **Tesla (11x sales, 160x P/E) vs. General Motors (1x sales, 6x P/E):** Lou and Jason both chose Tesla, despite its high valuation. Their reasoning centered on Tesla's "optionality" beyond the core auto business and Elon Musk's track record, even acknowledging GM's recent outperformance in stock returns. * **Eli Lilly (14x EV/Sales, 40x P/E) vs. Novo Nordisk (3.5x EV/Sales, 11x P/E):** Both hosts sided with Eli Lilly. They noted Lilly's strong position with GLP-1 weight-loss drugs (Zepbound, Mounjaro) and its strategy of using cash flow from blockbusters to aggressively build out its future pipeline through numerous deals, even if it's less of a "value" play. * **JP Morgan Chase (2.5x P/B, 8% 3yr growth) vs. SoFi (1.9x P/B, 27% 3yr growth):** Lou and Jason both preferred JP Morgan. They expressed skepticism about SoFi's ability to maintain high growth as it matures and questioned its reported 1.5 "relationships" per customer as a metric for monetization. They favored JPM's scale, strong management under Jamie Dimon, consistent dividend, share buybacks, and robust balance sheet. * **Costco (48x P/E) vs. Target (19x P/E):** Both hosts chose Costco. They highlighted Costco's sticky membership model (over 90% renewal rates), loyal customer base, and clear value proposition, contrasting it with Target's less defined "specialty" in the competitive retail landscape. **Tesla China Sale / SpaceX Merger Rumors** A Wall Street Journal report suggested Tesla might sell its China business to facilitate a merger with SpaceX, a rumor Elon Musk labeled "fake news." Lou and Jason discussed the logic behind such a move: integrating a defense contractor like SpaceX with a company having significant Chinese operations would likely face intense scrutiny from the U.S. government. They speculated that Musk might aim to complete such a merger during the current administration. This move would also signify Tesla's evolving identity, moving beyond just electric vehicles towards AI and robotics. **Stocks on Radar** * **Keysight Technologies (KEYS):** Jason Moser's pick. Keysight provides hardware, software, and services for testing and deploying technology across various sectors (communications, aerospace, automotive, semiconductors). Its software and services now account for 36% of revenue, offering higher-margin recurring income. * **L3 Harris Technologies (LHX):** Lou Whiteman's pick. This defense prime beat earnings and raised guidance but saw its stock drop over 10% after announcing a delay (until 2027) in the IPO of its missile business. Lou believes this sell-off is an overreaction, noting L3 Harris's $42 billion backlog, focus on high-margin defense electronics, and stronger international presence, making it an attractive buy at its current valuation, which is on par with peers like General Dynamics.