The Funding Gap That Debt Alone Cannot Close
Building the infrastructure for artificial intelligence — data centers, custom chips, high-voltage power connections, fiber — costs money at a scale that is straining even the balance sheets of the world's most cash-generative companies. The projected bill across the industry runs to $820 billion, according to reporting from MarketWatch, and the financing question is no longer theoretical.
Google moved first. Now Meta may follow. The mechanism under discussion is equity issuance: selling new shares to raise capital directly from stock markets. It is a tool that has been largely dormant for mature mega-cap technology companies, which have spent the past decade buying back shares rather than issuing them. That the calculus may be reversing is a meaningful signal about the capital intensity of the AI moment.
Bond Markets Got There First
Fixed-income markets — where companies borrow by issuing bonds that pay interest over time — have already been doing heavy lifting on AI financing. Investment-grade technology debt has found ready buyers, in part because the underlying companies carry strong credit ratings and predictable cash flows from their core advertising and cloud businesses.
Bond investors, per MarketWatch, are reportedly pleased with the prospect of equity issuance entering the mix. The logic is straightforward: if companies raise equity capital, they reduce their need to issue additional debt, which protects the seniority and coverage ratios that bond investors care about. More equity on the balance sheet is, from a creditor's perspective, a cushion.
What Equity Investors See Differently
Shareholders face a different set of concerns. Equity issuance — unless offset by proportional earnings growth — is dilutive. Each new share issued means existing shareholders own a slightly smaller fraction of the company. For a company like Google or Meta, where earnings per share (EPS) is a closely watched metric, dilution is not a trivial consideration.
The deeper question is return on invested capital, or ROIC: the ratio of profit generated to the capital deployed to generate it. AI infrastructure spending is, at this stage, a bet on future monetization. Data centers built today may not generate meaningful revenue for years. Equity investors are being asked to absorb dilution now in exchange for returns that remain speculative in their timing and magnitude.
A Structural Shift in Capital Strategy
For most of the past decade, large technology companies were net buyers of their own stock. Share buybacks reduced share counts, boosted EPS mechanically, and returned capital to shareholders without requiring a specific investment thesis. The pivot toward issuance — even if selective and measured — represents a reversal of that posture.
It also reflects a broader truth about the AI buildout: the capital requirements are large enough that companies cannot rely on operating cash flow and debt alone. Google's move, and Meta's potential follow, may mark the beginning of a period in which Big Tech treats its own equity as a legitimate financing instrument rather than something to be retired.
Whether that is good capital allocation depends entirely on what the AI infrastructure ultimately produces. Bond investors have already placed their bet. Equity investors are still doing the math.