Site navigation

Alphabet Expands Debt Strategy With Rare 100-year Bond

Graham Turner

,

Alphabet 100-year bond
The century-long bond is highly unusual in tech and aims to attract a wider pool of investors amid strong demand.

Alphabet is lining up a rare 100-year bond as part of a fresh fundraising push, with the world’s biggest technology companies turning increasingly to debt markets to pay for massive investments in AI and the data centres that power it.

The so-called “century bond” is expected to form part of Google parent Alphabet’s debut sterling bond sale this week, according a report from the Financial Times which cites ‘people familiar with the matter.’

The move comes alongside a broader multi-currency borrowing drive, with Alphabet also selling $20bn of dollar bonds on Monday and preparing a Swiss franc bond offering. The dollar portion of the deal was increased from $15bn after strong investor demand.

Century bonds, which lock in borrowing for a full 100 years, are highly unusual. While a number were issued during the ultra-low interest days following the financial crisis, only a handful of organisations have ever sold these kinds of bonds in sterling.

They are even rarer in the technology sector, where most large companies typically issue debt with maturities of up to 40 years, although it’s worth noting that IBM sold a 100-year bond in 1996.

According to the Financial Times, a banker familiar with Alphabet’s transaction said the company’s multi-currency bond offering is an effort to expand the investor pool given the massive amount of capital needed by Big Tech companies as the jostle for position in the AI race. The banker added that issuing a century bond in sterling is more cost-effective than in dollars, where interest rates are higher.

Big Tech firms and their suppliers are expected to invest almost $700bn in AI infrastructure this year, with many increasingly relying on debt to finance the expansion.

In November, Alphabet sold $17.5bn of bonds in the US, including a 50-year bond – the longest-dated dollar bond issued by a tech company last year – and raised €6.5bn in Europe. Last week, Oracle raised $25bn in a bond sale that attracted more than $125bn of orders.

Alphabet, Amazon and Meta have all increased their capital spending plans in recent earnings reports, prompting questions over whether such unprecedented investment can be funded through cash flows alone.

Alphabet reported annual sales topping $400bn for the first time, beating investor expectations in its most recent quarter, and said it plans to spend as much as $185bn on capital expenditure this year, roughly double last year’s total, to capitalise on demand for its Gemini AI assistant.


Recommended


Alphabet’s long-term debt rose to $46.5bn in 2025, more than four times the previous year, although the company ended the year with $126.8bn in cash and equivalents.

The fundraising comes amid a broader boom in borrowing by AI-focused companies. Oracle’s $25bn sale earlier this month followed a series of large deals in late 2025, including Meta’s $30bn bond issue – the largest-ever individual non-M&A high-grade bond sale — alongside offerings from Alphabet and Amazon.

The five major AI hyperscalers – Amazon, Google, Meta, Microsoft and Oracle – issued $121bn in US corporate bonds last year, according to BofA Securities. Moody’s estimates the largest US hyperscalers are on track to spend $500bn this year, while Morgan Stanley forecasts as much as $400bn in bond issuance to help fund that spending, potentially driving $2.3 trillion in total debt issuance across markets.

Barclays analysts expect overall US corporate bond issuance to reach $2.46 trillion this year, up nearly 12% from 2025.

Graham Turner

Sub Editor

Latest News

AI

Nvidia Launches Open Secure AI Alliance for AI Safety and Security

AI Business Recruitment

Nearly a Quarter of Orgs Reducing Entry-level Hiring Due to AI Automation

Business

Scottish Businesses Turn to Self-funding as Growth Confidence Dips in H2

Data Finance

Payment Leaders are Struggling to Get Real-time Data