The biggest players in private credit have been waiting years for a moment like this.
After years of circling opportunities in investment-grade lending and the booming artificial intelligence sector, top private credit firms have finally landed a headline-grabbing prize a $29 billion financing package for Meta Platforms Inc.’s massive AI data center project in Louisiana.
Led by Pacific Investment Management Co. (Pimco) and Blue Owl Capital Inc., the deal checks every box: it involves a high-quality borrower in one of the world’s fastest-growing industries, sidesteps the traditional bank-led funding route, and comes with an eye-popping price tag.
“
Private credit has been itching to get into this space,” noted John Medina, senior vice president on Moody’s global project and infrastructure finance team. “This is one of the first deals of its kind for private credit. If it works, we’ll likely see more.”
Tech giants are in an all-out arms race to dominate AI, and the costs are staggering. Elon Musk’s xAI Corp. recently told investors it plans to spend $18 billion on data centers, considering project-backed financing instead of corporate-level borrowing. Other major players like Amazon.com Inc. and OpenAI Inc. are building similar facilities across the U.S.
Morgan Stanley projects that AI-related capital expenditures could top $3 trillion over the next three years.
In Meta’s case, Pimco is expected to arrange $26 billion in debt, while Blue Owl contributes $3 billion in equity. The debt will likely take the form of investment-grade bonds secured by the data center’s assets, though the final structure is still being finalized.
The race to secure Meta’s financing was intense and months-long. Private credit firms have been eager to break deeper into the investment-grade market — a space long dominated by banks.
Apollo Global Management Inc. and KKR & Co. reached the final bidding round, with Brookfield Asset Management Ltd., Blackstone Inc., and Ares Management Corp. also in the mix, according to people familiar with the matter. While Morgan Stanley advised Meta, it’s not leading the financing.
By scale, this is the largest AI-specific data center funding on record. Comparable deals such as those for xAI or CoreWeave Inc. have all been well under $10 billion.
For context, the closest debt transaction in recent memory was a $26 billion bond issuance in March for Mars Inc.’s acquisition of food rival Kellanova, arranged by a consortium of banks for traditional syndication.
The industry is flush with cash. Private credit managers currently hold $450 billion in dry powder, according to Preqin, and are seeking new avenues to deploy it.
M&A activity traditionally a key driver for private credit deals has slowed to a crawl. That’s pushing these firms to expand into territory typically reserved for Wall Street’s largest banks: advising on deals, structuring debt packages, and supplying a portion of the financing themselves.
Some see this as the pathway to turning private credit into a $40 trillion market, a projection floated by Apollo.
“This private investment-grade market has massive potential and strong tailwinds,” said Michael Zawadzki, global CIO at Blackstone’s credit and insurance division.
KKR and Energy Capital Partners already committed to a $50 billion partnership last year to accelerate AI infrastructure development.
For Marc Lipschultz, Blue Owl’s CEO, the AI boom is reminiscent of the Gold Rush except the modern “gold” is computational power, and the “picks and shovels” are data centers.
“In this case, it’s the modern version the data centers,” Lipschultz said on a July 31 earnings call. “And we believe we’re the best-positioned firm to help build and finance them.”
With Meta’s groundbreaking Louisiana project, private credit has officially planted its flag in the AI infrastructure race. If the deal succeeds, it could mark the start of a new era where private lenders rival and even replace banks in some of the largest, most strategic financings in the world.
As a leading independent research provider, TradeAlgo keeps you connected from anywhere.