Meta Platforms finished pricing $15 billion of dollar-denominated bonds on Nov. 4 and immediately began marketing its first euro notes, a six-part sale with maturities running from four years to 38. The company had not borrowed in euros before. It chose to start in the middle of the largest debt-funded building program in its history.
The dollar deal was split into five tranches. Demand was heavy enough that the order book reached $58 billion at one point, nearly four times the amount sold. The bonds carry a coupon noticeably higher than the one Meta offered in April, a difference that says more about the market than about the company.
That difference is the story. Meta’s credit has not deteriorated in any way that a rating agency would flag. What has changed is the supply of paper. Every large technology company now funds data centers with debt, and the buyers of that debt have spent the past six months watching the borrowing totals rise. A company that could sell 30-year paper at one price in the spring sells it at a higher price in the autumn, because the marginal buyer has more alternatives and less appetite.
Bloomberg reported that with the euro tranches included, the total raised in this round will be at least $20 billion. That figure puts the sale among the larger corporate borrowings of the year, and it lands in a market that has been asked to absorb a great deal of similar paper from similar issuers.
The structure explains the strategy. Five dollar tranches let Meta match maturities to the years in which it expects to be spending, rather than paying for long money it does not need yet. A 38-year euro tranche does the opposite: it locks in funding past the depreciation life of almost every asset the money will buy, which is expensive and useful in equal measure. Long bonds are how a company with a 20-year construction plan avoids refinancing risk in year eight.
The euro debut is a separate decision with its own logic. European investors hold large pools of long-dated money and have historically paid up for high-quality corporate credit. A first-time issuer in that market usually receives a concession, a slightly higher yield than its existing dollar curve would imply, because buyers need a reason to add an unfamiliar name. Meta is large enough that the concession should be small, and the benefit is a second investor base that does not vanish when the dollar market tightens.
The timing is not accidental. A dollar book that is four times oversubscribed is the best possible advertisement for the euro sale that follows it. Buyers in Frankfurt and Paris can see that the dollar tranches cleared easily, which reduces the work of explaining the credit. Issuers with weak demand do not usually follow a dollar deal with a debut in a new currency within hours.
What the coupon increase reflects is a repricing of the entire cohort. For most of the past decade, the largest technology companies borrowed at rates close to those of utilities, because their cash flows were treated as bond-like. Data center spending has changed the comparison. A company that will spend a defined sum over five years to build assets with a defined useful life looks less like a utility and more like a telecom operator, and telecom operators pay a spread for the privilege of borrowing.
Investors are not refusing the paper. A $58 billion book is not a protest. They are demanding to be paid for a risk they can now measure: the gap between what these companies have promised to spend and what they have disclosed about the returns. When that gap was small, the bonds priced on the strength of the balance sheet. When it is large, they price on the spending plan.
Meta’s own disclosure has made the calculation easier. The company has told investors that capital expenditure will rise sharply, and the bond market has responded by requiring a higher coupon for each incremental dollar. That is the ordinary functioning of credit markets, applied to a borrower that until recently did not need them.
The second-order effect is what other issuers will notice. A first euro sale from a company of Meta’s size establishes a benchmark, and benchmarks are what the next issuer is measured against. If the euro tranches price tightly, more American technology companies will follow into the currency. If they price wide, the queue will slow, and the dollar market will absorb the difference at a higher cost.
None of this suggests Meta’s borrowing is in trouble. The company can service the debt from operating cash flow many times over, and the bonds will trade. What the sale shows is that the price of financing the buildout is no longer a footnote. It is a line item that moves every quarter, and the companies that locked in long money early are now the ones with an advantage.
