ByteDance has closed a $29.6 billion syndicated loan, the second-largest in Asia this year and roughly three times what the TikTok owner raised offshore in 2024. The facility started life at $20 billion. Banks put in more than $30 billion of orders, and it was upsized.

The number that matters is the spread

Size is the headline; price is the signal. The margin came in at 68 basis points over SOFR, down from 85 basis points on the company's previous deal. A borrower that increases a facility by half and simultaneously cuts its spread by 17 basis points is being competed over, not accommodated. That is a statement about how the syndicated loan market currently views large private Chinese technology credit — and it is a harder data point than any valuation mark.

What the common framing gets wrong

Headlines are attaching this money to AI data centres. Treat that as inference rather than disclosure. A general syndicated facility of this size is corporate borrowing, not a capital-expenditure announcement: it does not commit ByteDance to a build, a site, a megawatt figure or a delivery date, and a loan agreement is not a capex plan. Second correction, and it is the one most often skipped: this is debt. Coverage that files it alongside equity rounds as "ByteDance raises $29.6bn" is comparing an obligation that must be repaid with capital that need not be.

Why a private company borrows at this scale

ByteDance has no listing and therefore no equity market to tap, while facing the same compute bill as its listed peers. Bank debt is the instrument left, and at 68 over SOFR it is currently cheap enough to prefer. The scale of the facility relative to 2024 tells you how quickly the funding requirement has grown; the spread tells you the lenders are comfortable with it.

The regional read

Asia's loan market has spent two years being described as risk-averse toward Chinese technology borrowers. A $20 billion ask that draws $30 billion of orders and prices tighter than the last deal is inconsistent with that description, at least at the top of the credit spectrum.