The Alphabet Century Bond and the Growing AI Bubble Concern

This week, Alphabet issued a striking financial maneuver that has analysts sounding alarm bells. The company sold ultra-long-term bonds maturing in 2126—a century from now—and the investor response was almost too enthusiastic: bids came in at nearly £9.5 billion against the £1 billion the company actually needed. That’s 10 times the required amount. While such overwhelming demand might seem positive at first glance, market watchers are interpreting this as a warning sign that technology companies are spending recklessly on artificial intelligence, raising uncomfortable questions about whether a bubble is forming.

A £1 Billion Offering Draws Extraordinary Investor Appetite

Alphabet raised £1 billion across five different sterling-denominated notes, with the century bond paying approximately 6% annually. The overwhelming response—nearly ten-fold oversubscription—underscores intense appetite among UK pension funds and insurance companies seeking long-term investment vehicles. This offering is just one piece of Alphabet’s broader $20 billion borrowing program spanning multiple currencies. Combined with comparable initiatives from Amazon, Microsoft, Oracle, and Meta, industry analysts project these major technology firms will collectively borrow approximately $3 trillion over the next five years to maintain competitiveness in AI infrastructure development.

The choice to issue in sterling and tap British institutional investors also served another purpose: avoiding further strain on US capital markets where Alphabet and peers have already accumulated substantial debt. Yet this strategy itself raises questions about the scale of capital deployment across the tech industry.

History’s Bubble Pattern: The 1990s Dotcom and Telecom Lessons

The century bond itself is remarkable because no major technology company has borrowed this far into the future since the 1990s. Then, Motorola and IBM executed similar long-dated bond issuances before the dotcom crash devastated the sector. The parallels are instructive and sobering.

Motorola, once a top-25 US company, now ranks 232nd with just $11 billion in annual sales. IBM and Coca-Cola also participated in the 1990s century bond trend and subsequently lost their dominant market positions as newer competitors surged ahead. But the historical warning extends beyond individual company decline. During the telecom bubble of that era, companies raised $1.6 trillion and issued $600 billion in bonds to build out internet infrastructure—infrastructure for which demand never materialized. The result: widespread bankruptcies and bond investor losses so severe that some recovered only 20 percent of their initial capital.

“If you’re looking for a signal of a top, it does look a bit like a signal of a top,” Bill Blain from Wind Shift Capital told CNBC when discussing current AI spending patterns. He characterized present tech borrowing as “off-the-historical scale” and drew explicit parallels to past speculative cycles where excitement outpaced critical risk assessment.

AI Infrastructure Spending at an Unprecedented Scale

Alphabet needs its $185 billion in capital expenditure this year, with the vast majority directed toward data centers and artificial intelligence equipment. Building and operating these facilities requires continuous electricity supply, advanced cooling systems, and constant hardware refreshes. The marginal cost structure is unforgiving: even minor demand fluctuations can render expensive infrastructure economically unviable.

This is the core concern animating expert skepticism. Should AI adoption plateau, or should technological shifts render current infrastructure architecture obsolete, these multibillion-dollar facilities could transform into money-losing assets. As Phoenix Group, a major UK pension manager, observed to CityAM, other technology companies will “undoubtedly take notice” of Alphabet’s successful century bond offering and pursue similar financing strategies. Should competitive herding materialize, it would effectively validate the bubble hypothesis that market participants themselves recognize exists.

Why Markets Are Drawing Bubble Signals Today

The precedents are haunting. People who purchased Motorola’s century bonds in 1997 believed they were investing in an unstoppable technology leader. History rendered that thesis catastrophically wrong. Nobody genuinely knows whether Alphabet will maintain dominance through 2126. Betting on any corporation for a hundred-year horizon represents a considerable wager on both company longevity and sector continuity.

Meta has already raised $30 billion through private credit markets, while Oracle’s debt has climbed past $100 billion. The concentration of mega-scale capital deployment in narrow technology sectors, all directed at similar infrastructure objectives, resembles historical bubble dynamics more than measured corporate financing.

The concern isn’t that AI infrastructure won’t generate returns—it’s that the magnitude of capital deployment may systematically outpace actual demand, creating lasting excess capacity. When that recognition arrives, bondholders holding century bonds face an uncomfortable reality: they’ve committed capital to an organization they’re betting will thrive for a hundred years in an industry that changes every decade. History suggests that’s an exceptionally risky proposition.