Private Credit Report 2024 — Systemic Scale, Untested Claims

By 2024, private credit had grown from a niche allocation into a substantial part of corporate credit provision. That growth changes the nature of the analytical question.

At a small scale, the relevant question is whether an allocation is attractive — does the return compensate the risk, does it diversify a portfolio. At systemic scale, a further question arises: what happens to the financial system if this asset class experiences stress simultaneously?

That question had not been answered by 2024, and the reason is straightforward: the asset class had never experienced a conventional default cycle. As the 2020 report describes, the one moment that might have provided a test was removed by policy support within weeks. The record covers a period of unusually low defaults, one brief and heavily-supported shock, and a rate rise whose effects — per the 2022 report — were being deferred rather than resolved.

Three developments in 2024 changed the risk profile, and all three ran in the same direction:

Competition compressed spreads. Capital continued flowing in faster than loan supply grew. As in any market where capital outpaces assets, the difference appeared in price and terms — partially reversing the favourable conditions of the 2022–2023 vintage.

Bank relationships deepened. Rather than banks retreating further, they built partnerships with private credit managers — originating loans for distribution, providing leverage facilities, and taking positions in fund structures. This is efficient and it also means the two systems are connected in ways the "banks left, we filled the gap" account does not capture.

Retail access expanded. Semi-liquid and evergreen structures brought private credit to individual investors. These vehicles offer periodic redemption, which introduces a liability that can demand liquidity from assets that cannot readily provide it.

Why scale changes the question

An asset class small relative to the financial system can experience severe losses without systemic consequence. The losses are borne by holders and the effects remain contained.

Beyond a certain scale, three additional channels open:

• Correlated behaviour. If many holders face similar pressures simultaneously — redemptions, mark-driven limits, covenant triggers on their own leverage — they act in the same direction at the same time. That is what turns individual losses into a market event.
• Funding channel effects. If private credit became a primary source of financing for mid-market companies, a contraction in its lending capacity contracts credit availability to the real economy — regardless of whether any lender fails.
• Interconnection. Losses transmit to counterparties: banks providing leverage facilities, insurers holding fund positions, funds-of-funds and retail vehicles holding interests.

The third channel is the one that has grown fastest and is least well documented, because the connections are private. A bank providing a subscription line or a leverage facility to a credit fund has exposure to that fund's assets, at a remove, in a way that does not appear as a leveraged loan on its balance sheet.

The honest position on all three is that they are plausible and unquantified. Data on private credit's interconnections is poor precisely because the market is private. Supervisors have repeatedly identified this as a data gap rather than as a resolved question, and it remains one.

The question is not whether private credit is a good asset class. It is what happens when a large, opaque, illiquid, model-valued market experiences its first genuine default cycle while connected to the banking system in ways nobody has mapped.

Competition reverses the terms improvement

The 2022 report notes that loans originated in 2022 and 2023 carried the best terms in the asset class's history — wider spreads, lower leverage, better documentation — because lenders had pricing power when syndicated markets were closed.

That advantage eroded in 2024, through the mechanism the 2017 private equity report describes in a different context: when capital grows faster than the supply of assets, the difference appears in price.

Several forces converged:

• Fundraising continued strongly, on the track record generated in the benign period.
• Syndicated markets reopened, restoring competition for large financings — precisely the segment private credit had won in 2022.
• Transaction volume did not keep pace. Private equity deal activity remained below its peak, so the pipeline of new financings was smaller than the capital seeking to fund them.

The consequences were the ones this pattern always produces: spreads compressed, leverage on new loans rose, documentation loosened, and lenders competed on speed and flexibility rather than on price.

This creates a vintage problem that matters for anyone assessing the asset class. Loans from 2021 and earlier carry tight spreads and high leverage originated in a low-rate environment. Loans from 2022–2023 are the best in the asset class's history. Loans from 2024 onward are moving back toward the 2021 profile.

Aggregate portfolio statistics blend all three, and the blend describes none of them. Vintage analysis is not optional in this asset class — a manager's aggregate mark tells you far less than the distribution of when its loans were written.

Bank relationships complicate the origin story

The asset class's foundational narrative, set out in the 2016 report, is that regulatory capital requirements made leveraged lending uneconomic for banks, non-banks filled the gap, and the shift is therefore structural.

That account remains substantially correct about origination. It is increasingly incomplete about the system as it now stands.

By 2024, banks and private credit managers were connected through several channels:

• Origination partnerships, where a bank sources a loan using its client relationships and distributes it to a private credit partner, keeping the relationship without the balance sheet.
• Leverage facilities. Private credit funds use borrowed money to enhance returns, and that borrowing frequently comes from banks. The bank's exposure is to the fund rather than to the underlying loans, but the underlying loans are what ultimately support it.
• Subscription lines, short-term borrowing secured against LP commitments, near-universal in fund finance.
• Direct investment, with some banks holding positions in credit funds or managing private credit strategies themselves.

The implication is worth stating carefully, because it cuts against a claim the asset class makes about itself:

The risk did not leave the banking system entirely. Some of it changed form. A bank that would previously have held a leveraged loan directly may now hold a facility to a fund that holds leveraged loans. The exposure is more remote, structurally subordinated to the fund's other creditors in some configurations, and senior in others — and it is materially harder to observe from outside.

This is not necessarily worse. A fund with committed long-dated capital is a more stable holder of illiquid assets than a bank with demandable deposits — that is the genuine structural improvement, and it is real. But it is a different risk configuration than either the pre-crisis system or the "banks left" narrative describes, and it has not been through a cycle.

Retail access and liability mismatch

The expansion of private credit to individual investors through semi-liquid structures introduces a specific risk that is structural rather than speculative.

The mismatch. The underlying assets are illiquid — loans that cannot be sold quickly at a reliable price. The vehicles offer periodic redemption, typically quarterly and typically capped at a percentage of net asset value.

The management tools. Vehicles hold liquidity buffers in cash and traded credit, and impose redemption limits — "gates" — that cap total redemptions in a period.

Why the tools have a limit. In normal conditions, redemptions are modest and buffers absorb them. In stress, three things happen together:

• Redemption requests rise, because investors seek liquidity precisely when they most need it.
• The buffer is consumed, and replenishing it requires selling — the liquid assets first, which raises the illiquid share of what remains.
• Gates are imposed, which limits outflow but also signals that the vehicle is under pressure, which can raise redemption requests further.

That last dynamic is the important one and it is well documented in other illiquid-asset vehicles. A gate protects the fund and can accelerate the run, because an investor who cannot redeem now has reason to file a request immediately rather than wait.

There is a further consideration specific to valuation. These vehicles strike a net asset value from model-based marks. An investor redeeming receives that NAV. If the marks are above realisable value — which is exactly the concern in stress — then early redeemers receive more than the assets are worth, and the cost falls on those who remain. The valuation question and the liquidity question are the same question in a vehicle that offers redemption at NAV.

None of this makes semi-liquid structures inappropriate. It means their behaviour in stress depends on a combination — retail investor behaviour, gate mechanics, and mark accuracy — that has not been observed together in this asset class.

What "systemic" requires and what it does not

Describing an asset class as systemically significant is a strong claim and is frequently made loosely. It is worth specifying what would make it true, because the specification is what distinguishes analysis from alarm.

Size alone is not sufficient. A large asset class whose losses are absorbed by long-dated institutional capital that can hold through a cycle is not systemic — it is a large asset class, and the losses are borne by parties equipped to bear them. That is arguably the system working as designed, and it is the strongest defence of private credit's structure.

Three conditions convert size into systemic significance:

Correlated forced action. Holders must be compelled to act in the same direction at the same time. This requires a mechanism — redemptions, leverage covenants triggered by marks, regulatory capital limits. Closed-end funds with committed capital have none of these, which is why the asset class's original structure was genuinely more stable than bank balance sheets. Semi-liquid vehicles and leveraged fund structures reintroduce them.

A funding channel to the real economy. If private credit is a primary source of financing for mid-market companies, a contraction in its lending capacity contracts credit availability regardless of whether any lender fails. This is the least discussed channel and possibly the most consequential, because it operates through reduced new lending rather than through losses.

Interconnection with the regulated system. Losses that transmit to banks, insurers or other supervised institutions. This is the channel that has grown fastest — through leverage facilities, subscription lines, origination partnerships and direct positions — and it is the least observable, because the connections are private.

On the evidence available, the first condition is partially met and growing, the second is plausible and unquantified, and the third is real and poorly measured. That is not a prediction of failure. It is a statement that the asset class has moved from a structure where none of the three applied to one where all three partially do, without that transition having been through a cycle.

The original structural argument — that long-dated committed capital is a better holder of illiquid assets than demandable deposits — is sound and survives everything else. It applies to closed-end committed capital and applies less as the funding mix shifts toward vehicles offering redemption.

What an allocator could act on

Read the supervisory literature, not only the industry literature. The IMF Global Financial Stability Report and the Federal Reserve Financial Stability Report both cover private credit from a supervisory perspective, are free, and are the best independent counterweight to industry-sourced figures. They are also the source for the honest characterisation that interconnection data is a gap rather than a resolved question.

Track bank lending to non-bank financial institutions. The Federal Reserve's H.8 release publishes this weekly and free. It is the most direct public evidence for the interconnection channel and it is watched by almost nobody outside supervisory circles.

Analyse by origination vintage. The three-way split — 2021 and earlier, 2022–2023, 2024 onward — describes materially different books with different terms and different prospects. Aggregate portfolio statistics blend them and describe none, and the split is a reasonable thing to request.

Assess semi-liquid vehicles on the interaction, not the components. Redemption terms, gate mechanics, buffer size and mark accuracy are individually manageable and jointly untested. The critical point is that the valuation question and the liquidity question are the same question in a vehicle that redeems at a modelled NAV: if marks exceed realisable value, early redeemers are paid by those who remain.

Note that a gate protects the fund and can accelerate the run. An investor who cannot redeem now has reason to file a request immediately rather than wait, which is well documented in other illiquid-asset vehicles and has not been observed in this asset class at scale.

Separate the structural case from the performance record — again. Bank retrenchment from leveraged lending is a rule change and it makes the supply of loans durable. That establishes the asset class will exist. It says nothing about whether the loans are priced correctly, and the two arguments are routinely presented as one.

What 2024 established for private credit

• Scale converted an allocation question into a systemic one, with three transmission channels that are plausible and unquantified.
• Competition reversed the 2022–2023 terms improvement, creating a three-way vintage split that aggregate statistics conceal.
• Bank interconnection deepened, complicating the structural origin narrative and reducing observability.
• Retail liability structures introduced a mismatch in which the valuation question and the liquidity question become the same question.
• The central claims remained untested into a ninth year of benign conditions.