Private debt as a parallel banking system, and what it asks of the people who govern it

After the financial crisis, Basel III made banks hold more capital against the leveraged corporate lending they had done freely before. Capital that used to sit comfortably on a bank balance sheet became expensive, and mid-market corporate credit was among the first things to be repriced or pulled. Private lenders moved into the space the banks left.

They stayed for two reasons: regulatory arbitrage and, more durably, execution. A private credit fund can offer a borrower speed and certainty that a syndicated bank process struggles to match. One lender with one credit committee can close in weeks rather than months, and for a private equity sponsor running an auction on a tight timetable that certainty is worth paying for. Private credit now funds most leveraged buyouts, a reversal from a decade ago when banks and broadly syndicated loans dominated. The asset class and the private equity industry have grown dependent on each other.

The instruments grew with the market. The unitranche facility, which blends senior and subordinated debt into a single tranche at one blended rate, began as a tool for deals in the $50 to $100 million range. It now routinely exceeds a billion. Two or three large managers can club together to underwrite financings that once needed a full bank syndicate. The investor base institutionalised alongside it. Pension funds, insurers, sovereign wealth funds and endowments supply most of the capital, with retail still a small single-digit share of most managers’ books. That last share is the boundary the FCA now watches most closely.

That shift is what matters for governance. When credit intermediation moves from banks to funds, the prudential safeguards that applied to banks do not follow it. The funds are regulated, but under a regime built for asset managers rather than deposit-takers. The Bank of England’s Financial Policy Committee has said repeatedly that it is monitoring leverage and interconnectedness in non-bank finance for that reason. Anyone running a private credit fund in the UK should assume supervisory attention will grow.

It helps to be precise about which rules apply to a UK credit manager. The framework a private credit manager operates under is more demanding than “lightly regulated” suggests.

A UK alternative investment fund manager running a credit fund operates under the onshored AIFMD framework, which lives in the FUND sourcebook and SYSC. The core duties aren’t exotic. You need a permanent risk management function that is functionally and hierarchically separate from portfolio management, under SYSC and FUND 3.7. You set, document and monitor leverage limits and report leverage to the FCA under FUND 3.7.5R. You run liquidity management systems consistent with the fund’s liquidity profile under FUND 3.6, and the redemption terms you offer investors have to be supportable by the liquidity of the underlying loan book. For an open-ended credit fund that mismatch is the central tension of the whole structure, and I have written about it at length elsewhere.

Valuation is where firms most often struggle to show compliance. Under FUND 3.9 an AIFM must have written valuation policies and procedures that ensure a sound, independent valuation of assets, and the valuation function has to be independent of portfolio management or subject to safeguards that manage the conflict. For a fund holding private loans with no observable market price, that means a documented, repeatable methodology, evidenced inputs, and a governance trail showing the price was set independently of the people whose performance fees depend on it. A loan book marked by the deal team with no independent challenge gets picked apart on review.

Then there is the change reshaping the origination side. AIFMD II, Directive (EU) 2024/927, applies from 16 April 2026 and creates a harmonised regime for loan-originating funds across the EU. It brings risk retention requirements, concentration limits where a fund lends a large share of its capital to one borrower, a ban on originate-to-distribute strategies, and a structural steer toward closed-ended structures for funds that originate to a significant degree. The UK is not bound to copy it, and the FCA has signalled a more outcomes-based approach rather than wholesale onshoring. But any UK manager marketing into the EU, or running parallel structures, has to plan around it now. Leaving the work until close to April 2026 tends to force a rushed implementation.

In the funds I see, the failures cluster in a few predictable places. These failures are evidential, not credit failures.

The first is valuation independence that exists on paper and not in practice. The policy names an independent valuation committee. The minutes show the deal partner presenting the marks and nobody challenging them. When a loan deteriorates the mark stays flat a quarter or two longer than it should, because the people who originated it are effectively setting the price. What fixes it is a valuation process where the challenge is real, recorded, and visible to someone with no stake in the fund’s performance fees.

The second is liquidity terms that don’t match the assets. A fund offers quarterly redemptions on a book of five-to-seven-year private loans and relies on new subscriptions to fund the outflows. That holds until subscriptions slow. The FCA’s work on liquidity management tools, and the wider supervisory focus on liquidity mismatch in open-ended funds, is aimed squarely here. Redemption terms that only work in benign conditions become a liquidity problem under stress.

The third is leverage that is real but not fully seen. Fund-level subscription lines, asset-level leverage, and structural leverage in the way facilities are arranged can combine so the look-through exposure runs materially above the headline. Under FUND 3.7 you calculate and monitor leverage on both the gross and commitment methods, and you have to be able to explain the total. The board needs every source of leverage, including structural and subscription-line leverage.

The fourth is conflicts in allocation and cross-fund activity. Where a manager runs several funds or vehicles that could each hold the same loan, SYSC 10 requires you to identify, manage and where necessary disclose the conflict, and you need an allocation policy that produces fair, evidenced outcomes rather than an after-the-fact rationalisation. Where one fund’s capital touches another’s, the conflict is sharp and the documentation has to be airtight.

For most of its growth private debt has been an institutional asset class. Pensions, insurers and sovereign funds supplied the capital, understood the illiquidity, and were paid for it. The base is still overwhelmingly institutional, with retail a small single-digit share. But the pressure to widen access is real, and it’s where the regulatory risk concentrates.

In the UK the vehicle built to bridge that gap is the Long-Term Asset Fund, governed by COLL 15.8. The LTAF is an authorised open-ended structure designed to hold illiquid assets, private credit among them, with a redemption frequency and notice period matched to the assets rather than to investor convenience. It’s the FCA’s attempt to let a broader pool of investors — certain defined contribution pension savers and, within limits, retail — reach private markets without recreating the liquidity mismatch that has caused so much trouble elsewhere. Retail access requires a coherent liquidity framework, redemption terms that reflect the assets, and disclosure a less sophisticated investor can understand.

The bar rises with the breadth of the investor base. The Consumer Duty under PRIN 2A applies to retail products and asks whether the product delivers fair value and whether investors understand what they hold. A private credit strategy sold to institutions on the basis that they accept the illiquidity is one thing. The same strategy packaged for a wider audience invites questions about whether the redemption terms, the fees and the risk disclosures meet a standard written to protect people who can’t price illiquidity for themselves. A firm that wants retail or DC capital should build the governance for it before launch, not after.

The supervisory concern isn’t individual credit losses — it’s interconnectedness. The concern is interconnectedness: the web of relationships linking private credit funds to banks, to insurers, to the private equity sponsors they lend to, and to each other.

Banks didn’t leave leveraged lending. They moved up the structure, providing subscription lines, leverage facilities and warehouse financing to the very funds that replaced them in direct lending. A stress in private credit can travel back into the banking system through those facilities, and that is the linkage the Bank of England’s Financial Policy Committee has flagged. Insurers have become large allocators to private credit, tying the asset class to long-term savings and annuity books. And because a handful of large managers now club together on the biggest financings, the distress of one large lender is not a contained event.

For a fund’s governance, “our book is fine” is not a complete answer. The board has to know where the fund’s financing comes from, what happens if a leverage provider pulls a line, how correlated the fund’s exposures are with those of its peers and sponsors, and how a downturn would move through the structure rather than through individual credits alone. The FCA and the Bank of England have run, and keep developing, system-wide stress exercises aimed at non-bank finance for that reason. A manager that can describe its own place in that web is ready for the supervisory conversation those exercises are rehearsing.

Take a single mid-market loan. The borrower’s trailing earnings have softened, covenant headroom has narrowed, and the sponsor has flagged a slower year. The deal team proposes holding the mark flat, on the view that the business is sound and the dip is temporary.

In a fund that fails the evidence test, that recommendation goes through. The mark holds, the quarter looks stable, and the problem surfaces later, larger and harder to explain.

In a fund that passes, the valuation committee, chaired by someone independent of origination, asks for comparable spreads on similar credits, the revised base case, and a downside scenario. It records that it considered widening the discount rate, took the sponsor’s projections with appropriate scepticism, and settled on a mark with a written rationale and a named owner. The dataroom holds the inputs. If the loan deteriorates further, the trail shows the committee was awake to it, and if the regulator or the depositary asks, the answer is already on file.

Managers sometimes treat the depositary as a pure compliance cost. Under FUND 3.11 an authorised AIF must appoint a depositary responsible for cash flow monitoring, safekeeping of assets, and oversight of the AIFM’s compliance with the fund rules and applicable law. For a fund holding loans rather than listed securities, safekeeping means verification of ownership and record-keeping rather than custody in the traditional sense, and the oversight duty makes the depositary a standing independent check on whether the AIFM is doing what it says.

Used well, the depositary is an independent check on the evidence trail. When the depositary asks how a loan was valued, or whether subscription monies were applied correctly, a well-run fund already has the answer written down, and the exchange becomes a rehearsal for the questions a regulator or an LP would ask. The relationship works best when the fund’s governance is good enough that the depositary rarely has to push.

This insight is provided for general informational purposes only and doesn’t constitute legal, investment, or regulatory advice.