
上周三,Alphabet一天之内连创两项纪录。
首先,斩获史上最高单季利润。财报显示,Alphabet当季实现利润1,120亿美元,这是公司历史上首次单季利润突破千亿美元大关。不过,其中69%来自其持有SpaceX和Anthropic股权产生的未实现账面收益,而非核心业务。华尔街早已看穿了这一成绩背后的真相。
即便剔除这部分收益,谷歌(Google)该季度的业绩依然相当亮眼。如今已成为公司核心业务的云计算部门收入暴增82%。那么,为什么投资者当天仍然大举抛售Alphabet股票,导致其股价大跌近7%,创下关税风波以来的最差单日表现?
原因在于它创下的另一项纪录:公司历史上首次出现现金流转负,也就是说,上季度公司现金流出大于流入。管理层还警告称,2027年的资本支出将“大幅”增加,这进一步加剧了华尔街的焦虑。Alphabet上周四提交的最新文件显示,公司已承担超过8,000亿美元的采购履约义务及其他债务。这一惊人的数字还暴露出一些更隐蔽的支出,例如,Alphabet拿出了约510亿美元,为其他公司的数据中心提供财务兜底。
至少有六家机构因此下调了Alphabet的目标股价,包括派珀·桑德勒(Piper Sandler,下调至395美元)、瑞银(UBS,下调至379美元)以及D.A. Davidson,后者给出的350美元目标价是华尔街最悲观的预测之一。只有巴克莱(Barclays)逆势上调了目标价。这场抛售迅速蔓延,波及微软(Microsoft)、亚马逊(Amazon)、英伟达(Nvidia)以及整个芯片板块。上周,市场对新一波中国开源权重模型的担忧进一步加剧了这种恐慌。特斯拉(Tesla)同样交出了利润率下滑和资本支出高企的成绩单,导致其股价重挫15%。
D.A. Davidson科技研究主管吉尔·卢里亚在接受《财富》杂志采访时表示:“现金流转负是一个负面里程碑,市场现在正对此作出反应。此前市场普遍认为这种情况绝不会发生,如今已然成真。”
不过,尽管下调了目标价,卢里亚认为市场反应过度。他表示,自己担心的从来不是支出,而是估值。今年早些时候,当Alphabet股价接近400美元时,“市场像是在按照‘唯一赢家’为它定价”。而如今股价回落至320美元左右,他表示:“谷歌现在的估值合理得多,因为这是对‘赢家之一’而非‘唯一赢家’的估值。”他预计,未来几年,微软、亚马逊和Alphabet都将从提供AI算力中持续获利。
他认为,看多Alphabet的理由其实十分清晰:Google Cloud“增长势头惊人”。他预计,今年Alphabet将直接从AI算力基础设施建设中获得150亿至200亿美元收入。“只不过,在这个出现负面里程碑的日子里,首席财务官刚刚预警情况好转之前会先恶化,眼下没有人愿意关注这样的消息。”
不过,卢里亚担心的是,那些将AI巨头与自身客户深度绑定的复杂投资关系:谷歌、亚马逊和Anthropic之间的“三角关系”,微软与OpenAI的绑定,以及英伟达与CoreWeave之间的联系。他表示:“整个生态系统都在互相支撑。”而那些资产负债表外的兜底安排尤其令他警惕:“‘财务兜底’这个词,是在安然(Enron)事件之后才广为流传的,所以我一听到这个词就会心生警惕。”不过,他并不认为这是一种虚假的循环交易。他指出,如今消费者和企业在AI上的年化支出已经达到1,200亿美元,“而两年前,这个数字还是零。”“这是真金白银的支出,和循环交易根本不是一回事。”(财富中文网)
译者:刘进龙
审校:汪皓
上周三,Alphabet一天之内连创两项纪录。
首先,斩获史上最高单季利润。财报显示,Alphabet当季实现利润1,120亿美元,这是公司历史上首次单季利润突破千亿美元大关。不过,其中69%来自其持有SpaceX和Anthropic股权产生的未实现账面收益,而非核心业务。华尔街早已看穿了这一成绩背后的真相。
即便剔除这部分收益,谷歌(Google)该季度的业绩依然相当亮眼。如今已成为公司核心业务的云计算部门收入暴增82%。那么,为什么投资者当天仍然大举抛售Alphabet股票,导致其股价大跌近7%,创下关税风波以来的最差单日表现?
原因在于它创下的另一项纪录:公司历史上首次出现现金流转负,也就是说,上季度公司现金流出大于流入。管理层还警告称,2027年的资本支出将“大幅”增加,这进一步加剧了华尔街的焦虑。Alphabet上周四提交的最新文件显示,公司已承担超过8,000亿美元的采购履约义务及其他债务。这一惊人的数字还暴露出一些更隐蔽的支出,例如,Alphabet拿出了约510亿美元,为其他公司的数据中心提供财务兜底。
至少有六家机构因此下调了Alphabet的目标股价,包括派珀·桑德勒(Piper Sandler,下调至395美元)、瑞银(UBS,下调至379美元)以及D.A. Davidson,后者给出的350美元目标价是华尔街最悲观的预测之一。只有巴克莱(Barclays)逆势上调了目标价。这场抛售迅速蔓延,波及微软(Microsoft)、亚马逊(Amazon)、英伟达(Nvidia)以及整个芯片板块。上周,市场对新一波中国开源权重模型的担忧进一步加剧了这种恐慌。特斯拉(Tesla)同样交出了利润率下滑和资本支出高企的成绩单,导致其股价重挫15%。
D.A. Davidson科技研究主管吉尔·卢里亚在接受《财富》杂志采访时表示:“现金流转负是一个负面里程碑,市场现在正对此作出反应。此前市场普遍认为这种情况绝不会发生,如今已然成真。”
不过,尽管下调了目标价,卢里亚认为市场反应过度。他表示,自己担心的从来不是支出,而是估值。今年早些时候,当Alphabet股价接近400美元时,“市场像是在按照‘唯一赢家’为它定价”。而如今股价回落至320美元左右,他表示:“谷歌现在的估值合理得多,因为这是对‘赢家之一’而非‘唯一赢家’的估值。”他预计,未来几年,微软、亚马逊和Alphabet都将从提供AI算力中持续获利。
他认为,看多Alphabet的理由其实十分清晰:Google Cloud“增长势头惊人”。他预计,今年Alphabet将直接从AI算力基础设施建设中获得150亿至200亿美元收入。“只不过,在这个出现负面里程碑的日子里,首席财务官刚刚预警情况好转之前会先恶化,眼下没有人愿意关注这样的消息。”
不过,卢里亚担心的是,那些将AI巨头与自身客户深度绑定的复杂投资关系:谷歌、亚马逊和Anthropic之间的“三角关系”,微软与OpenAI的绑定,以及英伟达与CoreWeave之间的联系。他表示:“整个生态系统都在互相支撑。”而那些资产负债表外的兜底安排尤其令他警惕:“‘财务兜底’这个词,是在安然(Enron)事件之后才广为流传的,所以我一听到这个词就会心生警惕。”不过,他并不认为这是一种虚假的循环交易。他指出,如今消费者和企业在AI上的年化支出已经达到1,200亿美元,“而两年前,这个数字还是零。”“这是真金白银的支出,和循环交易根本不是一回事。”(财富中文网)
译者:刘进龙
审校:汪皓
Alphabet made history in two ways on Wednesday.
First, it had its most-profitable quarter in corporate history. Reporting $112 billion in profit, that’s its first 12-figured quarterly profit in history. But 69% of that came from unrealized paper gains on its stakes in SpaceX and Anthropic, not the core business, and Wall Street looked straight through it.
Stripping that out, however, Google still ostensibly had an excellent quarter; its cloud computing business, now the core of the company, soared 82%. So why did investors punish Alphabet today, sending their shares down nearly 7%, the worst day since tariffs?
Because of the other record it set: for the first time in the company’s history, it became cash flow negative, meaning less cash entered the company than left it last quarter. Management also warned that 2027 capital expenditures would be “significantly” higher, further deepening anxiety on the Street. The company’s latest filing Thursday shows more than $800 billion in purchase commitments and other obligations, an eye-popping number that reveals some sneakier expenses like some $51 billion spent backstopping other companies’ data centers.
At least six firms cut their price targets for Alphabet in response, including Piper Sandler (to $395), UBS (to $379) and D.A. Davidson, whose $350 target was one of the most bearish on the Street. Only Barclays raised its target. The damage was contagious, spreading outward to Microsoft, Amazon, Nvidia and the chip complex, and compounded this week by jitters over a new wave of powerful Chinese open-weight models. Tesla also reported a profit margin decline and high capital expenditure, sending its stock down 15%.
“The fact that we crossed over to negative cash flow was kind of a negative milestone, and people are reacting to that,” Gil Luria, head of technology research at D.A. Davidson, told Fortune. “Maybe there was a sense that we’d never get to that point, and we just did.”
Yet Luria, despite cutting his target, said the market was overreacting. His concern was never the spending but the valuation; earlier this year, with the stock near $400, “it was being valued as if it was the only winner.” At around $320, he said, “Google’s valuation is a lot more reasonable—because it’s the valuation of a company that’s a winner, as opposed to the winner.” He expects Microsoft, Amazon and Alphabet all to profit from supplying AI compute for years.
The bullish case, he said, is hiding in plain sight: Google Cloud’s “out of this world” growth. He estimated Alphabet will earn $15 billion to $20 billion this year directly from its compute buildout. “Nobody wants to hear that on a negative-milestone day where the CFO just told us it’s going to get worse before it gets better.”
Luria, however, worried about that tangle of investments which bind the AI giants to their own customers; the love triangle of Google, Amazon and Anthropic; Microsoft and OpenAI, Nvidia and CoreWeave. “This ecosystem is propping itself up,” he said, and the off-balance-sheet backstops give him pause: “That term came into parlance around Enron, which is why I never like hearing it.” But he drew a line at calling it circular. Consumers and businesses are now spending at a $120 billion annual rate on AI, he noted—”that number was zero two years ago.” “That’s real spend. There’s nothing circular about that.”