Infrastructure Credit: The Hard Asset Hedge Against Direct Lending Concentration
- 1 hour ago
- 3 min read
What's New
Large direct lending portfolios carry just 2 to 3% exposure to infrastructure deals, leaving allocators with heavy concentration in software and healthcare at the exact moment those sectors face AI disruption risk and macro uncertainty. Stonepeak Credit Partner Ryan Roberge, speaking on the Alt Goes Mainstream podcast, argues that infrastructure credit offers comparable yields to core direct lending but with fundamentally different risk drivers: hard asset collateral, contracted cash flows, and exposure to capital intensive megatrends like power buildout and digitalization that are largely absent from traditional private credit portfolios.
Why It Matters
The private credit industry has scaled by financing the LBO flywheel, predominantly in asset light sectors where unlevered free cash flow supports high leverage multiples. Infrastructure borrowers operate on a completely different logic. They consume capital to build physical assets, may burn cash for years during construction phases, and require lenders who can underwrite return on capital deployed into the ground rather than recurring software revenue. That specialization requirement is precisely what creates the opportunity: most generalist direct lending platforms lack the sector expertise to originate and underwrite these deals, leaving infrastructure credit as a structurally under penetrated segment of the private credit market.
Big Picture Drivers
Concentration risk: Core direct lending has become heavily weighted toward software and healthcare, with large BDC portfolios allocating minimal exposure to infrastructure assets, creating a diversification gap for allocators seeking uncorrelated return streams.
Specialization moat: Underwriting infrastructure credit requires asset level expertise in power markets, generation dispatch economics, equipment lifecycles, and construction risk that generalist credit teams do not possess, limiting competition and supporting spread stability.
Capital intensity tailwind: A massive infrastructure funding gap driven by AI compute buildout, energy transition, and grid modernization is generating sustained borrower demand for non dilutive structured credit from sponsors growing middle market platforms.
Spread stability: Unlike core direct lending, which saw spreads compress from 650 to 750 over SOFR down to much tighter levels as capital flooded in, infrastructure credit spreads have remained more stable because the bank market provides a persistent competitive floor for hard asset borrowers.
Maturation lifecycle: Infrastructure companies access different capital at different stages of growth, from early development equity through middle market credit (Stonepeak's target) to investment grade bonds at scale, creating a natural deployment window for specialized lenders.
By The Numbers
2 to 3%: Typical infrastructure deal exposure in large public BDC portfolios, highlighting the asset class's absence from mainstream private credit allocations.
25 to 100M EBITDA: Target borrower range for Stonepeak's infrastructure credit strategy, focused on middle market and lower middle market platforms backed by financial sponsors.
50 to 60%: Typical loan to value basis at which Stonepeak lends against hard asset collateral, providing substantial equity cushion below their position.
4 to 7 years: Standard loan duration for infrastructure credit deals, underwriting against underlying assets with 20+ year useful lives.
~500bp: Target blended spread for Stonepeak's infrastructure credit portfolio, designed to match core direct lending yield profiles with what they view as better risk adjusted returns.
Key Trends to Watch
Data center underwriting divergence: Roberge draws a sharp distinction between hyperscaler backed facilities with bulletproof contracts (bankable at 80 to 90% loan to cost with tight spreads) and more speculative builds for neo cloud tenants requiring significantly more equity cushion and higher spreads.
AI circularity risk: The most contrarian insight in the conversation centers on circular capital flows in the AI ecosystem, where hardware makers invest in customers who lease back that same hardware, masking whether true end user demand exists to support the infrastructure buildout.
Releasing risk as the real threat: With technology cycles moving far faster than infrastructure asset lives, the terminal value of purpose built AI training facilities in remote locations is uncertain. Roberge's team focuses on facilities that can be reconfigured for general enterprise or colocation use with minimal capex.
Displacement opportunity: Stonepeak sees its primary growth path not in competing with large infrastructure equity funds but in displacing regional banks, term loan B lenders, and generalist direct lending platforms that lack the sector expertise to structure creative, asset specific credit solutions.
The Wrap
Infrastructure credit's pitch to allocators is straightforward: comparable yields to core direct lending, backed by physical assets with contracted cash flows, in sectors that have almost zero overlap with the software and healthcare concentration dominating most private credit portfolios. The catch is that it requires genuine specialization to underwrite. Roberge's most valuable insight may be his contrarian warning on data centers: the asset everyone is rushing to finance is also the one with the most uncertain terminal value if technology cycles render purpose built facilities obsolete before their debt matures. The lenders who will win in this space are the ones underwriting what happens when the lease expires, not just the ones underwriting whether the tenant pays the bills today.



Comments