BIS Maps $1.7 Trillion Private Credit Push Into Digital Infrastructure Finance

The Bank for International Settlements published a working paper this week quantifying private credit's displacement of sovereign and multilateral capital in digital economy buildouts, naming the channel as a structural feature rather than a cycle-dependent substitute. The research arrives as AllianzGI closes a €1.1 billion impact direct lending vehicle and second-quarter GP filings show $47 billion in fresh digital infrastructure commitments across twelve domiciles.

The BIS paper isolates three financing gaps where private credit now commands allocation primacy: subsea fiber networks requiring $280-$340 million per route with fifteen-to-twenty-year revenue visibility, data center campus construction averaging $1.8 billion per hyperscale facility, and last-mile 5G densification in jurisdictions where sovereign balance sheets cannot accommodate the $14-$19 billion required per ten million covered lives. The authors note loan-to-value ratios in these structures range from 61% to 68%, tighter than commercial real estate but wider than regulated utility debt, with spread premiums of 240 to 310 basis points over comparable sovereign curves.

What separates this cycle from prior infrastructure dealflow is counterparty composition. The BIS identifies nineteen private credit platforms now writing checks above $500 million for digital projects that five years ago required Export-Import Bank or European Investment Bank participation. The paper cites contractual revenue certainty—hyperscaler take-or-pay agreements, telecom carrier IRUs with minimum-volume clauses—as the structural feature allowing non-bank lenders to price without sovereign guarantee wraps. AllianzGI's €1.1 billion fund, the largest impact direct lending close on record, allocates 68% to digital infrastructure with return targets of gross IRR 9.2% to 11.7%, achievable because cash flows begin within eleven months of financial close and run for twelve to eighteen years.

The second-order effect is capital markets plumbing. Private credit's dominance in this vertical means fewer benchmark sovereign or quasi-sovereign digital infrastructure bonds, which in turn reduces public market price discovery and increases reliance on GP-reported NAVs for portfolio marks. The BIS paper does not advocate for policy intervention but notes that $1.7 trillion in private credit exposure to digital economy projects—aggregated across the twenty-four jurisdictions surveyed—now exceeds the outstanding stock of rated digital infrastructure bonds by a factor of 3.8x. For allocators, this creates a valuation dependency on manager-reported data in a sector where comparable public instruments are thinning.

Operators and allocators should monitor three developments over the next six months. First, whether the European Investment Bank's proposed €4.2 billion digital infrastructure envelope, expected to launch in Q2 2025, prices inside or outside the 240-310 basis point private credit premium—if inside, it signals crowding-in; if outside, it confirms private capital as the marginal pricer. Second, the outcome of IOSCO's consultation on private asset valuation, closing March 2025, which may impose quarterly independent appraisals for funds holding more than 40% in illiquid infrastructure. Third, redemption queues at the seven largest semi-liquid private credit vehicles with digital infrastructure sleeves, where NAV marks will face first real-world tests if rate cuts stall and public comps remain scarce.

The BIS named the market structure. The next twelve months will prove whether $1.7 trillion in non-bank digital loans can weather a mark without public bond anchors.