How Meta's financing structure for its $27B Louisiana data center, which helps it keep debt off its books, hinges on some convenient accounting assumptions
“... a marvel of financial engineering: Meta Platforms is building a $27 billion data center in Louisiana, financed with debt, and neither the data center nor the debt will be on its own balance sheet.” Dave Lee / @davelee.me : Nvidia saying “Unlike Enron, NVIDIA does not use Special Purpose Entities to hide debt...” ... is an interesting point to make, given how Meta has set up its data center financing www.wsj.com/tech/meta-ai... [embedded post] George Pearkes / @peark.es : Some very good arguments here but there's also a KISS question here: “Why should a company consolidate an entity where it only owns a 20% stake?” — Seems like the operating versus financing lease distinction is more important than the consolidation of the entity, they're on much weaker ground there. Kara Swisher / @karaswisher : “Artificial intelligence, meet artificial accounting.” The guy who first pointed out Enron's sketchy accounting is now scrutinizing AI accounting and it's the same nonsense: www.wsj.com/tech/meta-ai... Paris Martineau / @paris.nyc : it's not a good sign when the guy who popped the enron bubble starts questioning your accounting practices ! [embedded post] Justin Hendrix / @justinhendrix : “It seems like a marvel of financial engineering: Meta is building a $27 billion data center in Louisiana, financed with debt, and neither the data center nor the debt will be on its own balance sheet.”
Context & Ripple Effects
Meta's Louisiana buildout was previously tied to a planned mix of outside equity and debt, as outlined in earlier reporting on its capital raise. This story explains the accounting structure that determines where that financing and the underlying asset appear.
The arrangement sits within a broader move by AI infrastructure builders to use special-purpose vehicles for data-center funding, documented in reporting on SPV-backed AI data centers. It matters because the treatment depends on assumptions about control and consolidation, not simply on who benefits from the capacity.
First-order effects
- Meta can finance the Louisiana project without reporting the data center or its associated debt on its own balance sheet under the structure described.
- The accounting judgment becomes a focal point for investors, auditors and commentators assessing whether Meta's stake and obligations warrant consolidation; EY's later red flag underscores that scrutiny.
Second-order effects
- Private-capital lenders and infrastructure partners gain a larger role in funding AI capacity, while investors must evaluate exposure through contractual commitments and project entities rather than headline corporate debt alone.
- Other large AI infrastructure buyers have a clearer template for separating project financing from corporate balance sheets, increasing pressure on peers' disclosure of control, guarantees and lease-like obligations.
Third-order effects
- If SPV financing becomes standard for AI data centers, reported leverage may become a less complete measure of companies' infrastructure commitments, shifting analysis toward entity-level terms and contingent obligations.
- Sustained scrutiny could push auditors, standard setters and regulators toward more consistent disclosure of off-balance-sheet AI infrastructure financing, though the eventual response depends on how these structures perform.
The trend: AI compute expansion is increasingly being funded as project finance, blending hyperscaler demand with private capital and making accounting treatment central to how infrastructure risk is assessed.