Editorials

Alphabet's capital expenditure has undergone a dramatic acceleration. From $31.5 billion in 2022 to a guidance range of $195–205 billion for 2026, the company has nearly tripled its annual infrastructure spending in just four years. The headline justification is familiar: an $811 billion AI infrastructure commitment, data center buildouts, and the computational demands of competing in large language models and generative AI. But the critical question for capital-efficient investors is whether the marginal dollar of spending is generating returns consistent with Alphabet's historical profitability, or whether it is beginning to outrun the moat.

To evaluate that question, we must examine three dimensions: the trajectory of capex as a percentage of operating cash flow, the spread between the company's return on incremental invested capital and its cost of capital, and whether free cash flow generation is keeping pace with the acceleration in spending.

The raw numbers are striking. Alphabet's fiscal 2025 capex totaled $91.4 billion—a 74% increase from FY 2024's $52.5 billion, and a near tripling from FY 2022's $31.5 billion. For context, that $91.4 billion alone exceeds the annual operating budgets of most Fortune 500 companies. And guidance for 2026 suggests capex will push deeper into the $195–205 billion range—meaning the company will be deploying more in infrastructure spending annually than its entire operating income of a few years prior.

Critically, capex-to-operating-cash-flow has shifted sharply. In FY 2025, capex consumed 55.5% of Alphabet's $164.7 billion in operating cash flow. Compare that to FY 2024, when capex ate up just 41.9% of operating cash flow, or FY 2023, when it was merely 31.8% of $101.7 billion in cash from operations. The company is allocating a steadily larger share of its cash generation to long-term assets—and the trend is accelerating toward a guiding range that could push the ratio well above 60% by 2026.

Here is where the durability question surfaces. Alphabet's operating income grew from $112.4 billion in FY 2024 to $129.0 billion in FY 2025—a gain of $16.6 billion on an incremental capex commitment of $38.9 billion ($91.4B − $52.5B). That implies an approximate pre-tax operating return on the incremental capital deployed of roughly 43%.

That sounds strong. It is strong. But it merits two caveats:

First, lags and lumpiness. Capital-intensive infrastructure spending has long lead times. Many 2025 capex dollars are funding data centers, GPUs, and networking infrastructure that will only generate returns in 2026, 2027, and beyond. The operating income gains we see in FY 2025 reflect the productivity of capital spent in prior years. To correctly assess whether the most recent increment of spending is earning its cost of capital, we must project forward—a task made harder by the fact that Alphabet does not segment capex by business line or expected return rate.

Second, the direction of the trend matters. Alphabet's historical operating margin (operating income as a percentage of revenue) has been remarkably durable, hovering around 32% even as the company has scaled revenue from $257.6 billion (FY 2021) to $402.8 billion (FY 2025). This suggests that prior rounds of capex investment—in data centers, search infrastructure, YouTube, and cloud—have generated returns sufficient to sustain those margins despite competitive and inflationary pressures. But there is no guarantee that the next doubling of capex will sustain that performance. The easy, high-return infrastructure may have already been built.

The clearest warning sign is the trajectory of free cash flow. In FY 2025, Alphabet generated approximately $73.3 billion in free cash flow ($164.7 billion operating cash flow minus $91.4 billion capex). That sounds robust—and it is, in absolute terms. But in growth rate terms, it is decelerating: FY 2024 free cash flow was roughly $72.8 billion ($125.3B − $52.5B), meaning free cash flow grew less than 1% year-over-year despite operating cash flow growing 31%.

This is the mechanical consequence of capex growing faster than operating cash flow. If operating cash flow grows 15% annually but capex grows 75% (as it did from FY 2024 to FY 2025), the residual free cash flow cannot keep pace. Extend the current guidance—where capex could reach $195–205 billion in 2026—and assuming operating cash flow grows at, say, 10–15%, free cash flow growth becomes either flat or negative. A company allocating $200 billion annually to capex would need operating cash flow north of $280 billion to maintain free cash flow of $80 billion.

Operating cash flow of $280 billion would imply revenue approaching $900 billion, assuming historical margins hold. That is not implausible over a multi-year horizon, given Alphabet's 12.5% three-year revenue CAGR and the AI tailwind. But it is also a significant bet: it assumes that the incremental capex will expand the addressable market enough—through cloud AI, search monetization improvements, or new revenue streams—to support a third of Alphabet's entire operating cash flow going to infrastructure maintenance and expansion.

Alphabet's moat remains intact: the company has durable competitive advantages in search, YouTube network effects, and increasingly in cloud infrastructure, supported by extraordinary profitability and cash generation. The capital-allocation question is not whether the company can afford to spend $200 billion annually on infrastructure. It plainly can.

The question is whether it should—and at what rate the returns on that spending will stabilize. Here, the data offers limited clarity:

Historical precedent suggests confidence. Alphabet has proven, through decades of reinvestment, that it can deploy large amounts of capital into infrastructure, software, and platforms and generate strong returns. The company's operating margin expansion, despite massive scaling of revenue, testifies to that.

But margins are plateauing. Operating margins have been stable—not expanding—over the past three years despite a 56% increase in revenue. If AI capex is truly as strategically critical as the $811 billion commitment implies, why has operating income margin not expanded? One interpretation: current capex is still largely exploratory, with returns yet to be realized. Another: the easy returns have been captured, and future capex is buying optionality, not assured returns.

Capital returns are declining. Alphabet repurchased $45.7 billion in stock in FY 2025, down sharply from $62.2 billion in FY 2024. At the same time, capex jumped 75%. This is a deliberate reallocation: management is signaling that infrastructure investment now takes priority over returning capital to shareholders. That choice may be sound—but it is a choice, and it reflects uncertainty about near-term returns on that capital.

For value investors, the frame is this: Alphabet is conducting a massive, multiyear experiment in AI infrastructure. The company has the capital base and cash flow to fund it. And management's track record of capital allocation deserves respect. But the durability of Alphabet's returns is no longer a function of its core search and advertising business—it is a function of whether the capex surge generates returns on incremental capital equal to, or exceeding, the cost of that capital. The company's historical performance does not guarantee that outcome. Investors should watch the next two years closely: if operating income growth continues to lag revenue growth, and free cash flow continues to decelerate, even Alphabet's mighty moat will begin to erode under the weight of unproductive capital deployment.