Alphabet had a busier week than on earnings day.
It closed Friday at $295.77, down 0.98%. It briefly surged to $304.97 intraday before being pushed back, but still gained about 4.7% for the week. It remains 27.6% below its 52-week high of $408.61, with the 200-day moving average at $338.45 hanging overhead. Its relative strength index is 40.58, still languishing in weak territory.
First, let’s do the math. Third-quarter results were announced on October 28: revenue of $127.6 billion, up 25% year over year; Google Cloud revenue of $30.7 billion, more than doubling year over year; operating income of $40.6 billion; and an operating margin of 31.8%. The figures looked great on paper. But buried in the same report were two figures: quarterly capital expenditures of $51.4 billion and free cash flow of negative $9.1 billion, compared with expectations of just negative $2.7 billion. The next day, the market voted with its feet: the stock fell about 12.6% in a single day to close at $283.82, wiping out more than $500 billion in market value as it plunged from $3.97 trillion to $3.46 trillion—its worst day since going public. Revenue growth of 25% couldn’t save the day; the real reason for the selloff was negative cash flow.
Three developments this week brought those numbers back into focus.
First, on November 12, four nonprofits sued the AI divisions of Google, OpenAI, Anthropic, Meta, and SpaceX in San Francisco Superior Court. They asked the court to order a freeze on frontier-model development, barring the release of any model more powerful than those available on November 12, citing California’s Unfair Competition Law. The 63-page complaint seeks only this one form of relief, with no damages. Realistically, the case is unlikely to have an immediate impact: none of the five defendants has publicly responded, no hearing date has been set, and U.S. courts have never issued an injunction prohibiting the development of more powerful models. But it put a question on the table: as models get more expensive to build, who will pay for the race to build them faster?
Second, on the same day, Google went ahead with its usual product launch: Gemini 4 Hydrogen went live across Google Search, the Gemini app, Google AI Studio, and enterprise products. The lawsuit is the lawsuit; product launches continue, with the pace completely undisturbed.
Third, and most substantial, Google agreed to invest $10 billion in about 135 million newly issued Anthropic shares at the offering price. It’s part of Anthropic’s $26 billion private financing, priced on November 18, with trading starting on November 19. Google is both Anthropic’s cloud provider and now one of its shareholders—a presence across the supply chain.
The chain of events is really quite simple: after NVIDIA CEO Jensen Huang resigned on November 12, Wall Street began seriously calculating the potential of a second provider of AI computing power. Morgan Stanley estimates that direct sales of Google’s Tensor Processing Units (TPUs, its in-house AI chips) could reach $84 billion next year, compared with about $8 billion this year—a tenfold jump. If that plays out, Google Cloud would no longer be a “cash-burning side business” but a second growth engine, giving the company’s full-year capital expenditure of $195 billion to $205 billion a clear rationale. On the other hand, if TPUs don’t sell, negative free cash flow won’t be a one-quarter issue.
There are two risk factors, too. A trust belonging to founder Larry Page reported selling about $148 million worth of shares on November 12: 253,800 each of Class A and Class C shares, under a monthly sale plan established back in 2020. It doesn’t affect control and, taken on its own, isn’t a bearish signal. But it coincides with the period of negative free cash flow, making it easy for the market to read “the founder is selling according to plan” as “insiders are also waiting for a better price.” The other concern is the valuation anchor: the 52-week low of $272.11 is only about 8.7% below the current price. If the stock breaks below the $300 psychological level again, the next support is around $272, followed by the base formed late last year.
Three levels to watch: whether the stock can reclaim the $300 psychological level; the 200-day moving average at $338.45 as the medium-term bull-bear divide; and Anthropic’s pricing and listing on November 18–19 as the nearest catalyst.
Google’s story has shifted from “search company” to “computing power company.” Whether the stock recovers depends on whether the TPU numbers make it from research reports into earnings.
#Alphabet #MagnificentSevenStocks
$GOOGL $NVDA $QQQ
It closed Friday at $295.77, down 0.98%. It briefly surged to $304.97 intraday before being pushed back, but still gained about 4.7% for the week. It remains 27.6% below its 52-week high of $408.61, with the 200-day moving average at $338.45 hanging overhead. Its relative strength index is 40.58, still languishing in weak territory.
First, let’s do the math. Third-quarter results were announced on October 28: revenue of $127.6 billion, up 25% year over year; Google Cloud revenue of $30.7 billion, more than doubling year over year; operating income of $40.6 billion; and an operating margin of 31.8%. The figures looked great on paper. But buried in the same report were two figures: quarterly capital expenditures of $51.4 billion and free cash flow of negative $9.1 billion, compared with expectations of just negative $2.7 billion. The next day, the market voted with its feet: the stock fell about 12.6% in a single day to close at $283.82, wiping out more than $500 billion in market value as it plunged from $3.97 trillion to $3.46 trillion—its worst day since going public. Revenue growth of 25% couldn’t save the day; the real reason for the selloff was negative cash flow.
Three developments this week brought those numbers back into focus.
First, on November 12, four nonprofits sued the AI divisions of Google, OpenAI, Anthropic, Meta, and SpaceX in San Francisco Superior Court. They asked the court to order a freeze on frontier-model development, barring the release of any model more powerful than those available on November 12, citing California’s Unfair Competition Law. The 63-page complaint seeks only this one form of relief, with no damages. Realistically, the case is unlikely to have an immediate impact: none of the five defendants has publicly responded, no hearing date has been set, and U.S. courts have never issued an injunction prohibiting the development of more powerful models. But it put a question on the table: as models get more expensive to build, who will pay for the race to build them faster?
Second, on the same day, Google went ahead with its usual product launch: Gemini 4 Hydrogen went live across Google Search, the Gemini app, Google AI Studio, and enterprise products. The lawsuit is the lawsuit; product launches continue, with the pace completely undisturbed.
Third, and most substantial, Google agreed to invest $10 billion in about 135 million newly issued Anthropic shares at the offering price. It’s part of Anthropic’s $26 billion private financing, priced on November 18, with trading starting on November 19. Google is both Anthropic’s cloud provider and now one of its shareholders—a presence across the supply chain.
The chain of events is really quite simple: after NVIDIA CEO Jensen Huang resigned on November 12, Wall Street began seriously calculating the potential of a second provider of AI computing power. Morgan Stanley estimates that direct sales of Google’s Tensor Processing Units (TPUs, its in-house AI chips) could reach $84 billion next year, compared with about $8 billion this year—a tenfold jump. If that plays out, Google Cloud would no longer be a “cash-burning side business” but a second growth engine, giving the company’s full-year capital expenditure of $195 billion to $205 billion a clear rationale. On the other hand, if TPUs don’t sell, negative free cash flow won’t be a one-quarter issue.
There are two risk factors, too. A trust belonging to founder Larry Page reported selling about $148 million worth of shares on November 12: 253,800 each of Class A and Class C shares, under a monthly sale plan established back in 2020. It doesn’t affect control and, taken on its own, isn’t a bearish signal. But it coincides with the period of negative free cash flow, making it easy for the market to read “the founder is selling according to plan” as “insiders are also waiting for a better price.” The other concern is the valuation anchor: the 52-week low of $272.11 is only about 8.7% below the current price. If the stock breaks below the $300 psychological level again, the next support is around $272, followed by the base formed late last year.
Three levels to watch: whether the stock can reclaim the $300 psychological level; the 200-day moving average at $338.45 as the medium-term bull-bear divide; and Anthropic’s pricing and listing on November 18–19 as the nearest catalyst.
Google’s story has shifted from “search company” to “computing power company.” Whether the stock recovers depends on whether the TPU numbers make it from research reports into earnings.
#Alphabet #MagnificentSevenStocks
$GOOGL $NVDA $QQQ