After Google’s $84.7 billion funding round, the market adjusted, and AI valuations began to focus on how quickly profits can be recouped.

TL;DR

· Alphabet plans to raise approximately $80 billion in equity financing, later increased to $84.75 billion, bringing AI infrastructure funding to the forefront.

· The market's concern isn't about AI failing, but about how much money needs to be spent upfront on data centers, chips, electricity, and network construction, and how long it takes to recoup the investment.

· Related assets: GOOGL, NVDA, AVGO, AMD, MU, DLR, EQIX, as well as assets related to power and data centers.

In recent years, the core question of AI trading has been simple: Will AI change the world? As long as the answer leans toward "yes," the market is willing to assign higher valuations to chipmakers, cloud providers, software companies, and model companies.

Recently, market language has begun to shift. Some semiconductor and high-valuation AI software stocks have pulled back, and investors are starting to favor investments with clearer order flow and more stable cash flows. Meanwhile, Alphabet announced a large-scale equity financing and revised its 2026 capital expenditure guidance upward in its Q1 earnings report.

These two events can't be simply described as "financing causes decline." A more accurate context is that the market is re-pricing AI from a software growth story to a heavy-asset infrastructure cycle.

The key term here is capital expenditure. AI isn't a business that can expand with just a few lines of code; it requires chips, data centers, networks, electricity, and land. The larger the capital expenditure, the more investors will ask three questions: Where does the money come from? How expensive is it? How long until they break even?

Alphabet's financing prompts the market to recalculate funding

Alphabet's financing itself isn't a crisis signal, but it serves as a strong reminder: AI development is already a massive capital project.

According to SEC filings and reports from Reuters and Investing, Alphabet announced in June 2026 plans to raise about $80 billion in equity financing, later increased to $84.75 billion. The funds are intended for AI infrastructure and computing capacity expansion, but not all directly allocated to AI capital expenditure. SEC documents show that of the $40 billion ATM plan, about $30 billion is expected to cover administrative obligations related to employee equity vesting taxes.

This distinction is important. Listing the entire $84.75 billion as "AI infrastructure funding" would exaggerate the direct scope, but it still influences investor sentiment. Even cash cows like Alphabet need to expand financing in the public markets, prompting questions: if they need to bolster financial flexibility, who will provide the funds for OpenAI, Anthropic, xAI, data center REITs, and power companies next?

Capital expenditure and operating expenses are not the same. Spending on hiring and marketing is operating expense; buying servers, building data centers, and powering them are capital expenditure. The latter is more like building a factory, with high upfront cash flow pressure, reflected gradually through depreciation on the books, but the market will immediately assess the payback period.

In its Q1 2026 earnings call, Alphabet raised its full-year capital expenditure guidance from $175 billion–$185 billion to $180 billion–$190 billion. The reasons include investments related to the Intersect acquisition and AI compute demand. The company emphasizes maintaining a healthy balance sheet and financial flexibility, and management did not describe the financing as a survival pressure.

Investors are doing another calculation. As capital expenditure guidance continues to rise, the denominator in valuation models also changes: increased depreciation, pressure on free cash flow, higher financing costs, and potential equity dilution are factored in. As AI enters the next phase, the previous focus on imagination shifts to capital efficiency.

AI's money isn't only burning on big tech's books

The capital needs for AI infrastructure aren't limited to giants like Alphabet, Microsoft, Amazon, and Meta. What's truly concerning the market is that multiple entities might compete simultaneously for the same pool of capital.

The first category is frontier model companies. OpenAI, Anthropic, xAI grow rapidly, but training and inference models require continuous purchase of compute power, leading to high cash consumption. Unlike mature cloud providers with advertising, cloud, and software cash flows, they rely more on external financing, strategic investments, and may eventually depend on IPOs or debt markets.

The second category is data center companies. AI demands are not for ordinary office servers but high-density, energy-intensive data centers. Data center REITs raise capital to build server farms and lease compute infrastructure to cloud providers or AI firms. Assets like Digital Realty and Equinix will benefit from demand growth, but expansion also requires ongoing financing.

The third category is power and utilities. One of the biggest bottlenecks for AI data centers isn't chips but electricity. Large data centers transfer pressure to power grids, substations, transmission lines, and long-term power purchase agreements. AI companies' spending won't stop at GPUs; it flows along the supply chain into land, server rooms, cooling, power grids, and energy projects.

According to Axios on June 10, Alphabet, Amazon, Meta, Microsoft, and Oracle have raised $255.34 billion through equity and debt financing by 2026, and these companies expect AI data center spending to reach about $750 billion within the year. While this figure isn't an exact causal proof, it provides the market with a scale: AI's capital demand is shifting from individual companies to a financing cycle that the entire financial market needs to absorb.

Historically, the market viewed AI as a software revolution: low marginal costs, rapid growth, high profit margins. Now, frontier AI resembles infrastructure revolutions like railways, electricity, and fiber optics: early-stage concentrated construction, huge investments, potential value creation, but also tests of financing capacity, capital costs, and capacity utilization.

Valuation logic shifts to payback speed

When re-evaluations occur, prices usually reflect not deteriorating fundamentals first, but investors switching to a different set of questions.

In the past, the questions were: Who has the strongest AI narrative? Who's growing fastest? Who's closest to the next-generation platform? Now, the questions are: Who can turn invested capital into cash flow? Who has sufficiently certain orders? Who can finance at low cost? Who will be diluted or pressured on profits during high-capex cycles?

This explains recent divergence within the AI sector. High-valuation AI software and companies with long-term stories are more vulnerable because their valuations depend on future growth. When market conditions raise the cost of capital, the discounted value of future cash flows declines. Some semiconductor companies are also affected, as investors worry whether orders can continue to grow faster than expected.

But this doesn't mean all AI assets are abandoned. Hardware, storage, networking equipment, data centers, and power assets with clearer order flow may be relatively supported in the revaluation. The reason is straightforward: when the market begins to focus on construction cycles, sellers of tools still have demand; but investors will be more selective, asking whose orders are real and visible, and who is just riding narratives to inflate valuations.

This is also where the disagreement between Alphabet management and cautious investors lies. Management emphasizes AI investment as a strategic necessity, and financing as a way to stay competitive long-term. Cautious investors worry that AI monetization may lag behind capital expenditure, especially if multiple giants and model companies expand financing simultaneously, leading the capital markets to demand higher returns and lower valuations.

Both perspectives can be valid. AI can be a long-term infrastructure investment, but it can also temporarily suppress free cash flow and valuation multiples. For investors, "bullish on AI" and "cautious on some AI valuations" are not mutually exclusive.

Next steps: capital expenditure and revenue realization

It's too early to interpret recent declines as driven primarily by AI financing pressures, nor to say AI is experiencing liquidity issues. Macro rates, profit-taking, cooling of crowded trades, and employment data disruptions could all cause sector volatility. Financing news is more likely being incorporated into the broader explanation framework rather than being the sole price driver.

However, this framework itself warrants attention. Once the market begins pricing AI based on "capital expenditure, financing costs, payback cycles," the ranking of assets will change.

For cash-rich giants like Alphabet, the issue isn't whether they can raise money, but whether AI investments will continue to squeeze free cash flow and whether new investments can translate into cloud revenue, advertising efficiency, subscription income, or enterprise services. As long as revenue growth can cover depreciation and financing costs, the market can accept higher capital expenditure; if guidance continues upward without timely returns, valuation pressures will intensify.

For pure AI companies, the question is more direct: can high revenue growth keep pace with compute costs? If OpenAI, Anthropic, xAI can demonstrate that enterprise clients are willing to pay continuously, and that unit economics improve, external capital will still flow in; if revenue growth is mainly eaten up by higher training and inference costs, the next funding round or IPO valuation will become more selective.

For data center and power assets, the market will focus on long-term contracts, utilization rates, financing structures, and energy constraints. The more genuine the AI demand, the more these "foundation" assets matter; but if financing costs rise or data center expansion outpaces actual demand, they may shift from beneficiaries to burdened heavy assets.

The most important validation points ahead are not daily semiconductor index fluctuations, but whether future earnings reports continue to raise capital expenditure guidance, whether AI revenue can be realized faster, and whether the public market can still smoothly absorb large-scale equity and debt issuance. As long as these variables remain positive, AI trading won't end; but the valuation language for AI has already shifted from just imagining potential to assessing real progress.

Click to learn about Rhythm BlockBeats job openings

Join the official Rhythm BlockBeats community:

Telegram Subscription Group: https://t.me/theblockbeats

Telegram Discussion Group: https://t.me/BlockBeats_App

Twitter Official Account: https://twitter.com/BlockBeatsAsia