Meta Raises Interest Costs by $48 Million on $12 Billion Texas Data Center Bond as Yields Exceed 7%
Meta’s $12 billion financing for a 1GW data center in Texas will be priced at yields above 7%, adding approximately $48 million to annual interest costs, according to details of a BlackRock-led bond deal. The current pricing is about 40 basis points higher than the $27 billion Hyperion offering Meta completed nine months ago.
The wider yield adjustment highlights rising investor caution on artificial intelligence infrastructure debt. Wellington Management noted that data-center borrowing is becoming costlier amid valuation uncertainty and shrinking returns, while Goldman Sachs and JPMorgan Chase launched products this week to allow investors to quickly adjust their exposure to hyperscale technology credit.
From the sources (7 posts)
@junkbondinvestAI data centers need so much money they're now borrowing from the market that finances Norwegian salmon farms and ferry operators. The first deal set a record. The next two got pulled. That was the whole cycle. Ten weeks.
@businessData-center debt risks are rising, there’s uncertainty over what the properties are worth and returns are falling, according to Wellington Management
@ftMeta faces higher borrowing costs in latest $12bn data centre financing
@firstsquawkMeta faces higher borrowing costs in latest $12bn data centre financing - FT
@wallstengineMeta new $12B, nearly 1GW Texas data center deal is being discussed at yields above 7%, around 40bps higher than $META’s $27B Hyperion financing just 9 months ago. That difference adds $48M in annual interest. Hyperion bonds are currently
@businessGoldman Sachsand JPMorgan Chase launched products this week allowing investors to quickly cut — or ramp up — exposure to tech industry debt, as concerns mount over hyperscalers’ huge future bond sales for artificial intelligence investments
@ftFT Exclusive: Bond investors are looking for significantly higher yields on a new BlackRock-led deal compared with the terms secured just nine months ago, as markets price in higher risks around AI exposure. https://