Alphabet tucked a nearly $98 billion windfall into a single line item. In its second quarter earnings report, the company said its “other income” totaled $98 billion, and that the figure came from unrealized gains on its investments. Alphabet did not disclose which investments drove the gain, but the business context makes the likely sources pretty clear: major equity stakes in companies like SpaceX and Anthropic.

So the question investors were really asking was not “where did the $98 billion come from?” It was “why is the market still mad?” Alphabet’s stock closed down about 1.24% after the report, despite the large paper gain. The market chose to focus on what management is spending to build its AI future and what has not yet materialized in that future: rising capital expenditures and ongoing struggles in the AI race, including continued delays around the next big AI chatbot that some rivals have mocked online.

This is where the one-line windfall becomes a governance and capital allocation story, not a finance trivia contest. The gain is “unrealized,” which means it is recognized on paper as the value of investments rises, but it is not necessarily cash in the bank. Alphabet also does not have an obligation to disclose exactly where those gains come from, and it did not. That matters because shareholders want clarity on durability. Paper gains can reverse. Cash generation has to show up in operating performance over time.

The most obvious reason the $98 billion number still “feels” different from revenue is that Alphabet is simultaneously telling a story of heavy investment. The tech giant hiked its capital expenditures to a maximum of $205 billion this year as it races to compete on AI. Capex at that scale is not a side quest. It typically signals that the company is pouring money into compute, infrastructure, and the broader stack required to train and run AI systems efficiently. In other words, even if equity investments are marking up in value, the core business still needs to prove it can translate that strategy into product traction and sustained results.

Alphabet is not new to this style of disclosure. In April 2025, it disclosed a similar $8 billion paper gain. That pattern tells you something about how Alphabet views investor relations: it will report large equity marks through financial statement categories, but it will not necessarily map each one to every underlying holding in the earnings narrative. In a way, it is the financial equivalent of letting your biggest donors be known by reputation, not by name.

The likely underlying investments are the same names that have been driving Alphabet’s “other income” chatter for a while. Google was an early investor in SpaceX, buying about 7% of the company in 2015. SpaceX also uses Google Cloud for its Starlink service, which creates a commercial relationship on top of the equity stake. The source notes that SpaceX is currently worth about $1.5 trillion dollars since its IPO last month. Back in the earlier valuation, Google invested when SpaceX was worth only about $12 billion, implying a 133x return. Those are exactly the kinds of multi-year marks that can create enormous unrealized gains when valuations move.

Alphabet’s stake in Anthropic is another plausible driver. The source states that Alphabet owns about a 14% stake in Anthropic as of last March, according to filings seen by the New York Times. It also points out that Anthropic was valued at almost $1 trillion in a massive $65 billion funding round in May, with some investors thinking it is already worth $1.2 trillion. Add the Databricks angle, too: Alphabet is also an investor in Databricks, which was valued at $188 billion in a funding round earlier this month. Together, these holdings illustrate why a single “other income” line can balloon to eye-popping amounts. SpaceX, Anthropic, and Databricks did not respond to requests for comment.

But for executives, the important takeaway is not that Alphabet is a great investor. The market already knows Alphabet is good at picking winners. The market is asking a different question: can Alphabet’s AI build-out justify the spend before the next platform shift locks in on competitors? The source flags a key tension. Alphabet has advantages in distribution and chipmaking, yet its efforts to build a leading AI model have not paid off yet. Management continues to delay its next major AI chatbot, and rivals have been mocking those delays online. That combination matters because customers and developers reward momentum. If the product roadmap slips, even an outstanding capital markets performance might not be enough to keep the stock from reacting.

Still, analysts are not treating the quarter as a wash. The source notes that many analysts remain bullish on Alphabet’s fundamentals. Revenue jumped by almost 25% compared to last year, driven by strong ads and cloud sales, which are also being boosted by AI. Emarketer principal analyst Nate Elliott is quoted as saying: “Another impressive quarter for Google.” The result is a split-screen market: equity gains that look like rocket-fuel on paper, and operating and AI execution pressure that looks like runway lights that keep flickering.

For boards and senior leadership at other companies in the AI era, the second-order lesson is sharp: valuation marks can calm some nerves, but they do not replace product proof. A company can be winning in private markets while still losing in public-market expectations. In Alphabet’s case, the $98 billion “other income” line did not turn into investor confidence because the next question was always going to be about capex, AI delivery timing, and whether today’s spend becomes tomorrow’s advantage.