Banks used to be the only place a company could borrow real money. That’s no longer true. A parallel lending system bigger than most national economies, mostly invisible to regulators, and now reaching into ordinary people’s retirement accounts has quietly become one of the most important stories in global finance. Here’s exactly how it works, why it exists, and why some of the smartest people in finance are nervous about it.

What Is Shadow Banking, Actually?

Formally, this sector is called Non-Bank Financial Intermediation (NBFI) credit provided outside the traditional deposit-taking banking system, largely by alternative asset managers, insurers, pension funds, and specialty lenders rather than regulated commercial banks. The Financial Stability Board (FSB) tracks NBFI globally and puts it at over half of all global financial assets well over $250 trillion.

Within that broader universe, “private credit” specifically refers to privately negotiated corporate lending direct loans, mezzanine finance, distressed debt, and increasingly asset-based finance that never trades on a public exchange. Here’s the number that actually needs unpacking, because sources cite wildly different figures depending on what they’re counting: the current global private credit market (direct lending, broadly defined) sits at roughly $1.7-2.1 trillion as of 2025-2026, per Preqin, Global Market Insights, and Bloomberg data. The $40 trillion figure your draft referenced is a different number, it’s the total addressable market Apollo and other major managers project private credit could eventually capture if it expands into asset-based finance (auto loans, aircraft leases, mortgages, royalties) at scale, not the market’s current size. Both numbers are real and worth knowing; conflating them overstates today’s footprint by roughly 20x.

How a Loan Actually Happens in the Shadow System

Three mechanics drive this expansion, and they rarely make headlines individually:

Asset-Based Finance (ABF). This is the largest growth engine behind projections of the market someday approaching that $40 trillion addressable figure. Private lenders write debt secured by hard, cash-flowing collateral aircraft leases, consumer auto loans, residential mortgages, franchise royalties, solar arrays effectively privatizing lending that used to sit on bank balance sheets or in securitized public bond markets.

PIK toggles (Payment-in-Kind). When a portfolio company can’t service its debt with cash in a higher-rate environment, lenders let it pay interest by issuing more debt instead of cash. This keeps the loan looking “performing” on paper while the company’s actual liability load keeps growing critics call the result “zombie corporations”: companies that aren’t technically in default but are accumulating debt they may never organically repay.

NAV financing. Private equity funds that can’t exit portfolio companies (via IPO or sale) borrow against the existing value of their portfolio itself, a second mortgage on a private equity fund, essentially to generate liquidity without selling anything. It’s a increasingly common tool for aging funds, at a real cost in fees and subordination.

Why Did This Even Happen? (Hint: Blame the Rules)

Post-crisis capital and risk-weighted asset requirements under Basel III/IV specifically penalize banks for holding non-investment-grade, illiquid corporate debt on their balance sheets. Banks responded rationally by retreating from middle-market lending, leveraged buyouts, and long-duration infrastructure financing. Alternative asset managers Blackstone, Apollo, Ares, and others stepped directly into that vacuum, deploying capital raised from pensions, insurers, and increasingly retail/wealth channels.

The retreat has been reinforced by a structural demand shift: the AI infrastructure buildout (data centers, energy grid upgrades, semiconductor fabrication) requires bespoke, multi-billion-dollar debt packages with decade-long horizons, a maturity profile banks’ short-duration deposit funding was never built to match. Private credit funds, which can lock up investor capital for years, are structurally better suited to that financing profile, which is a real and durable reason for the sector’s growth independent of regulatory arbitrage.

Who’s Actually Running This?

Current private-credit AUM figures (Q1 2026, varying by source and definition direct lending vs. total credit platform):

The AUM figures above vary meaningfully between sources some count total credit platform assets, others isolate direct lending, and some include committed-but-uncalled capital. Treat any single number as directional rather than precise; cross-referencing multiple Q1 2026 earnings releases is the only way to get a fully reconciled figure, and even analysts disagree on methodology.

Beyond the lenders themselves, the “plumbing” matters: State Street, holding over $54 trillion in assets under custody/administration, provides fund administration, NAV lending infrastructure, and cross-border tracking for private credit funds and special purpose vehicles profiting from the system’s growth without originating the debt itself. Vanguard, by contrast, has stayed almost entirely on the sidelines, anchored in low-cost public-market index investing and publishing investor-education material warning about private credit’s illiquidity risk rather than participating in it directly.

The Part That Should Worry You (And Why)

Three structural vulnerabilities compound each other:

• No mark-to-market pricing. Private credit assets are valued via internal models and periodic (often quarterly) subjective appraisals rather than continuous market pricing, meaning deteriorating credits can sit on a fund’s books looking healthy for multiple quarters before a forced sale or default event reveals actual impairment.
• Negative free cash flow at scale. IMF data has historically flagged roughly 40% of private credit borrowers as running negative free cash flow meaning they aren’t servicing debt from organic profits, but through PIK toggles and refinancing.
• Retail-liquidity mismatch. Semi-liquid retail-facing vehicles (interval funds, non-traded BDCs) promise periodic redemption windows, but the underlying assets are genuinely illiquid. Non-traded BDC NAV climbed roughly 47.6% year-over-year through mid-2025 before decelerating into Q3 2025, a pace that itself is being watched as a possible leading indicator of an inflection point. If redemption requests spike simultaneously (a “run”), funds hit gate limits that legally allow them to delay or cap withdrawals locking retail investors’ capital precisely when they want liquidity most.
• The insurer transmission channel. Because major private credit managers (Apollo/Athene, Blackstone/Corebridge-AIG partnerships) are deeply intertwined with insurance and annuity balance sheets, a private-credit default wave doesn’t just hit a hedge fund it flows into the retirement and insurance backstops of ordinary policyholders, a channel regulators have flagged as harder to see coming than a 2008-style bank run because it doesn’t show up on traditional bank stress tests.

The FSB and IMF have both flagged NBFI/private credit opacity as a systemic risk repeatedly since 2023, but it’s worth being precise about what that means: “flagged as a risk to monitor” is not the same as “predicted to cause a crisis.” Private credit has not yet been tested through a full, sustained default cycle since its scale reached current levels that’s a genuine unknown, not a foregone conclusion of collapse. Reasonable analysts disagree on whether this is a slow-building systemic risk or a well-collateralized asset class that simply looks scarier because it’s new and opaque.

The Global Picture

Non-bank financial institutions in Europe manage over €50 trillion in assets roughly 42% of the region’s total financial system driven by the same Basel III/IV-style constraints pushing European banks away from mid-sized corporate lending. The Bank of England has run repeated stress tests specifically on private equity and private credit exposure given how deeply embedded the shadow system is in UK and eurozone corporate financing. Cross-border capital flows compound the interconnection risk: global pension funds allocate billions into US-managed private credit funds, while those same US funds lend directly to European and Asian corporates meaning a shock originating in any single jurisdiction’s private credit book has plausible transmission channels into institutional portfolios worldwide.

The Speculative Frontier Algorithms Lending to Algorithms

As financial institutions scale semi-autonomous AI agents into core operations automated trade execution, compliance monitoring, cash-flow management there’s an emerging thesis that these systems will increasingly manage private credit allocation and corporate liquidity pools programmatically, effectively creating machine-to-machine lending relationships outside traditional central bank deposit-tracking metrics. This is a genuinely early-stage, thinly-evidenced trend as of mid-2026, it’s a reasonable extrapolation from current AI deployment patterns in trading and compliance functions, not something with robust public data behind it yet. We’re flagging it because it’s the direction several institutional strategists are pointing, not because it’s a documented, measurable phenomenon today.

Strip away the jargon and the core story is simple enough for anyone to follow: banks got more cautious after 2008, private investment giants stepped into the gap, and that gap has grown into a multi-trillion-dollar parallel lending system that increasingly touches ordinary people’s retirement savings while remaining largely invisible to the regulators who’d normally be watching for trouble.

The professional version of that same story is more nuanced: today’s private credit market is real but far smaller ($1.7–2.1 trillion) than the headline $40 trillion figure suggests that number is a projected total addressable market, not today’s footprint. The genuine risk isn’t that private credit is inherently fraudulent or doomed; it’s that it has never been tested through a full default cycle at its current scale, its valuations are self-reported rather than market-priced, and its growing retail distribution channel creates a liquidity mismatch that hasn’t yet been stress-tested by a real redemption panic. Whether that turns into 2008-style contagion or simply an asset class that matures through its first hard cycle is the open question every credit desk and every regulator is currently trying to answer, and nobody actually knows yet.

This analysis reflects publicly available data from Preqin, the FSB, the IMF, Global Market Insights, and Q1 2026 earnings releases from major private credit managers, current as of August 2026. It is not investment advice.