Perspective on Risk - June 26, 2025 (SLR Reform & Private Credit)
SLR Reform, Private Credit, and who really sets the capital standards
Kudos (I Think) On SLR Reform
As you know, I expected the Fed to exempt Treasury securities from the SLR as a way to help deal with the flood of Treasuries coming to the market. Instead, the Agencies have decided to lower in absolute terms the SLR ratio. This is better for banking, as it loosens a “binding constraint” that discouraged banks from conducting some business that otherwise would make sense from a simple risk-return framework. Maybe not as good for improving Treasury market function or absorbing the coming flood of issuance.
Current U.S. GSIB holding company SLR requirements are 5% (3% SLR +2% eSLR). Proposal would change the eSLR from a flat 2% to a firm-specific amount equal to one-half of the firm's Method 1 GSIB surcharge.
This move doesn’t expand the Fed’s balance sheet directly but it creates the conditions for systemic leverage to quietly rise beneath the surface. By reducing the Enhanced Supplementary Leverage Ratio, the Fed is effectively loosening the regulatory constraints that limit bank
Private Credit Update
The Plumbing Has Re-Routed: Banks → NBFIs, Not Companies
There has been a six-sigma jump in ‘lending-to-lenders’. US banks’ on-balance-sheet loans to non-bank financial institutions (NBFIs) quintupled in the past decade and topped $1 trn by Q1-25, now representing >10 % of total loans—up sharply since 2023. Bank loans to NBFIs reached $1.2 trn in Mar‑25 (Fitch), up 20 % YoY; Moody’s tallies $525 bn of commitments after an 18 % CAGR since 2021—still < 4 % of total lending but the fastest‑growing line, especially at mid‑tier banks.
The Boston Fed has analyzed Y-14 data, which shows that “jumbo” private-credit warehouse lines (> $2 bn) exploded from 5 deals in 2019 to 42 in 2023.1
Banks have discovered low risk-weighted-asset (RWA) capital treatments (as low as 20 %) for senior NBFI exposures and began scaling the trade; in the last 18 months mid-tier banks have joined the game (quite reminiscent of how hedge fund lending evolved in the 1990s).2 As Huw van Steenis of Oliver Wyman puts it, banks “keep the senior risk and shed the tail,” leveraging 20 % RWAs and synthetic risk transfers.3
Divergent Business Models: Insurance Power vs. Asset-Light Distribution
Post‑pandemic, the “big three” have forked. Apollo’s balance‑sheet assets stand at $395 bn against $32 bn of equity — a 12.2× leverage ratio comparable to JPMorgan., funded by its insurer Athene. Blackstone, by contrast, trumpets an asset‑light model that “sells products better than anyone else on Wall Street.” The split matters: insurer‑backed models carry bank‑like mark‑to‑market risk, while fee‑based giants rely on uninterrupted fundraising. Either way, model heterogeneity means future stress may hit platforms differently — a nuance absent from the pre‑2024 narrative.4
Macro Transmission: The Non-Bank Lending Channel
Recent micro- and cross-country work reshapes how we think about the non-bank credit channel. Fresh micro‑data confirm substitution. Weekly Fed H.8 figures show loans to NDFIs racing 22 % YoY versus just 2 % for all other bank lending.5 Using Denmark’s matched loan-register and balance-sheet data show that when policy rates rise, long-maturity funding actually flows into finance companies, enabling them to expand lending even as banks pull back; that extra credit almost fully offsets the usual decline in firms’ investment and markedly softens the hit to household consumption.6
A complementary 33-nation panel from the New York Fed finds that the type of lender financing a boom shapes its macro fallout: bank-driven credit surges predict multi-year GDP weakness and a higher probability of -2 percent growth “tail” events, whereas non-bank booms fade quickly and are associated with lower downside risk two to three years ahead.7
The cushioning, however, has limits. Acharya, Gopal & Steffen (2025) document that firms heavily reliant on shadow-bank term loans receive smaller, more expensive bank credit-line commitments and, in stress episodes such as the 2014-16 oil-price collapse or the March 2020 cash dash, suffer sharper contractions in liquidity and real activity than otherwise comparable borrowers . Taken together, the 2024-25 research confirms that non-banks now dampen the immediate impact of rate tightening, but that this comes with heightened rollover and liquidity fragility when systemic stress hits.8
Take-away: Since late ’24 we have micro-data that both confirms substitution and maps the trade-offs (rate insulation vs. liquidity fragility).
Intentional Opacity As A Feature
Over the last year the information architecture of private-credit markets has hardened into intentional opacity, and three new strands of evidence show why that matters.
Recent theory formalises this logic. Gorton, Li & Ordoñez (2025) show that information-insensitive debt achieves its highest capacity when price discovery is costly and collateral ownership is probabilistic; adding intermediaries further dilutes any one lender’s incentive to value the asset, making deliberate opacity a feature, not a bug.9
First, direct-lending managers are actively blocking attempts to build a secondary-trading venue. FT Alphaville reports that JPMorgan’s weekly “loan runs” now land in a “penalty-box”: sponsors threaten to cut off any bank that dares quote their paper, because regular marks would kill the low-volatility story they sell to investors. The article illustrates the stakes with Pluralsight’s loan, which JPMorgan valued at “over 90 cents on the dollar” while creditors quietly marked it near 50 cents, a gap that would be impossible to hide if the debt traded freely.10
Second, scrutiny is mounting over how headline returns are manufactured. A June 2025 Covenant Lite bulletin dissects fund-level track records and shows that survivorship bias alone can add several percentage points to reported IRRs, making the asset class look sturdier than it is once dead deals are restored to the sample.11 This critique of “performance smoothing” was largely absent from pre-2024 discourse and directly undermines the marketing claim that direct lending offers equity-like returns with bond-like risk. It similarly echoes the private equity mismarking comments of Cliff Askness that we’ve previously discussed.
A National Association of Insurance Commissioners study found private‑letter grades from smaller agencies run three notches higher on average; such ratings now cover 86 % of insurers’ private‑credit holdings. JPMorgan analysts warn this “inherently more risky” blind spot leaves outside observers unable to gauge credit quality. Ann Rutledge is blunter: “There’s a build‑up of risk in the insurance industry… the opacity and the risk are both attributable to cracks in the rating industry.”12
Finally, the distribution engine has pivoted from institutions to retail wealth, amplifying the pressures described above. Michael Batnick’s “Relentless Ask” chronicles how financial advisers are now receiving “twenty emails a week” from alternative-asset sales desks as fund-raising from endowments and pensions stalls, turning opacity into a feature that can be sold as “vol-free diversification” to high-net-worth portfolios.13 Together these developments suggest that private credit’s illiquidity premium increasingly depends on keeping price discovery at bay, even as the investor base broadens to channels that are less equipped to interrogate unmarked risk.
Liquidity Back-stops
Banks have become the core liquidity back-stop for the private-credit complex, and the latest data reveal both the strength and fragility of that arrangement. Boston Fed analysis of Y-14 supervisory filings shows that nearly every publicly traded business-development company (BDC) funds itself with revolving bank credit rather than term loans—about 97 percent of the dollar volume is first-lien, senior-secured revolvers—and roughly 30 percent of those lines were still undrawn at end-2023, giving lenders substantial “dry-powder” capacity.14
Banks remain the core shock absorber. Moody’s estimates 58 % of bank facilities are first‑lien and fully collateralized, and Fitch stresses that a downturn is “unlikely to have widescale financial‑stability implications for the largest banks” — but admits second‑order effects are hard to quantify.1516 Theory suggests there is a real trade-off: collapsing chains or mandating frequent marks could curb capacity just when corporates need it most.17
That liquidity can cushion a downturn, but it also means a correlated draw could hit bank balance-sheets all at once; the paper warns that such tail-risk may be “under-appreciated,” especially if borrower defaults prove more correlated than models assume. The plumbing is getting bigger, too: jumbo warehouse facilities (single lines above $2 billion) exploded from five in 2019 to forty-two in 2023, with average size climbing to $2.4 billion.1819
Frankly, as a supervisor, this is at least a transparent risk, unlike the “liquidity puts” that resulted in large subprime problems for some banks during the GFC.
At the borrower level, this safety net is decidedly conditional. Acharya, Gopal & Steffen (2025) find that firms heavily financed by non-bank term loans receive smaller and more expensive bank credit-line commitments; a full shift from bank to non-bank funding cuts credit-line availability by roughly 18–25 percentage points and raises drawn spreads by about 50 basis points, leaving these companies especially vulnerable when markets seize up .
Finally, headline numbers may still understate the bank–private-credit nexus. Barclays estimates that rising use of synthetic risk-transfer trades (SRTs)—banks buying credit protection on corporate loans from investment funds—replicates the same exposure off balance-sheet and is invisible in Call-Report data, suggesting the true volume of “lending to lenders” is materially larger than reported.20 The financial stability policy establishment should address this.
In short, the liquidity back-stop that underpins today’s private-credit boom is wide, senior, and cheap—but it is also concentrated, increasingly sizeable, and potentially pro-cyclical when many borrowers need cash at the same time.
Financial Stability Implications
The recent plumbing changes carry both shock-absorbing and shock-amplifying elements.
The BIS paper shows that non-banks funded with long-maturity debt expand lending after a hike, dampening the usual fall in investment and consumption. The Boston Fed notes that private-credit funds employ markedly lower leverage than banks and rely on capital locked up for years, not demand deposits, which “could reduce overall financial-stability risk” if growth mainly reflects a substitution of bank loans. Revolving lines to BDCs are first-lien, senior-secured and priced ~200 bp over reference rates—well inside the 600 bp spreads BDCs charge their own borrowers—so banks would eat credit losses only in a deep, correlated default cycle.
The $1 trn surge in “lending to lenders” is capital-relief-driven; banks book senior NBFI exposures at just 20 % risk-weight, and synthetic-risk-transfer trades hide even more off balance sheet. Barclays warns that as NBFI leverage grows, historical loss data become “less useful,” echoing pre-2008 mis-calibration worries.
The Boston Fed stresses that default correlations could be “higher than anticipated.” Acharya et al. add that firms financed by non-bank term loans already face tighter, costlier bank credit lines, so a market-wide shock could deliver a double-whammy—non-banks step back just as banks ration liquidity. Retail-wealth distribution and off-balance-sheet SRTs widen the circle of stakeholders exposed to the same underlying credits, potentially magnifying a sentiment or valuation shock, and echoes the later stages of the subprime boom. Finally, blocking secondary trading to preserve “price stability,” making mark-to-model valuations and smooth IRRs central to their business will delay stress signals and concentrate losses.
SLR Reform and the Private‑Credit Feedback Loop
As we discussed above, the Fed and OCC proposed cutting the enhanced SLR add‑on to a 3.5‑4.25 % sliding scale, down from today’s 5 % (holding‑company) and 6 % (bank‑subsidiary) thresholds, so the ratio once again acts as a back‑stop rather than a day‑to‑day binding cap. (federalreserve.gov, occ.gov)
The move frees scarce leverage capacity. With risk‑weighted capital already light (20 % RWAs on senior NBFI exposures), the SLR has been the real bottleneck. A lower denominator charge makes warehouse lines, BDC revolvers and SRT hedges even more attractive relative to riskier C&I lending or even long‑duration Treasuries. Expect faster growth in “lending‑to‑lenders” and thinner spreads as banks compete harder.
Cheaper, more abundant bank credit lets private‑credit funds lever up—or pass the savings to sponsors in lower coupons. Either way, it further entrenches the bank–NBFI loop and may blunt the pricing discipline that scarcity once imposed.
Proponents argue that extra balance‑sheet headroom will support Treasury‑market functioning and absorb deposit inflows during stress. Critics counter that SLR discipline was the last brake on correlated lines and that lower capital against off‑balance exposures raises tail‑risk if many funds draw at once.
The prototypical 4 %-capital commercial loan is a senior, unsecured or well-secured credit facility to a large corporate borrower carrying an A- to A rating under the standardized approach. That is not an unreasonable level. The problem in the Global Financial Crisis was the Aaa and “Super-Senior” ratings given to senior tranches of subprime securitizations that allowed 50-100x leverage. Here, they would be allowing 25x leverage.
Lowering the supplemental leverage ratio (SLR) removes one important speed-limit for large banks, but the relative economics and non-capital frictions that drove origination into private-credit funds are still in place. The “old ways” of bank financing don’t seem likely to suddenly return. At most, the proposal will slow the flow of new warehouse-type exposures; it is unlikely to send banks racing back into plain-vanilla, on-balance-sheet C&I lending.
Senior loans to NBFIs carry 20 % RW (≈1.6 % total-capital charge), while typical unrated middle-market corporates sit at 100 % RW (8 % capital). Providing short-term warehouse lines provides better pricing and doesn’t tie up liquidity (needed for Liquidity Coverage Ratio constraints).
SLR reform doesn’t change the capital-arbitrage equation.
And Just A Reminder - Rating Agencies Really Set Capital Levels
Request for Comment: Banks - Proposed Methodology Update (EMEA / Americas)
… we propose to recalibrate how we arrive at the Capital sub-factor score within the Solvency factor to incorporate a more consistent approach to leverage, and to replace the Funding Structure and Liquid Resources sub-factor ratios within the Liquidity factor to provide a more nuanced approach to assessing funding risks and the most effective mitigants.
The proposed revisions would also enhance the adjustments we may make to sub-factor scores, including those for interest rate risk, credit concentration, and the degree of balance sheet growth. Our updated approach would also provide greater clarity on how we apply Qualitative Adjustments notching for outsized risks that we might not otherwise capture in the Financial Profile, and it would reflect our view that business models and risk appetite tend to be among the key causes of bank failures. Other proposals include changes to the Macro Profile, instrument notching and our approach to stress testing.
Moody’s Seeks to Impose Capital Increase on American Economy, Harming Lending and Economic Growth (BPI)
In broad terms, the proposed changes relating to the capital sub-factor would require banks to hold additional capital in order to maintain their current ratings under the existing Moody’s framework.
The higher capital calibration would result, in part, from the proposed recalibration of the tangible common equity (“TCE”) to risk-weighted assets (“RWA”) ratio, setting the lowest category at below 7%, in contrast to “around 5% in the current methodology.” Moody’s states that this increase reflects “our view that a higher point of failure more closely aligns with the capital levels at which banks have failed, or at which regulators have stepped in over the last decade.”
In addition, under the proposal, Moody’s would be “unlikely to assign a subfactor score above that indicated by the financial ratio category of Moderate where a bank holds an excess of capital less than 1.5% of RWA above its minimum regulatory requirements
Banks Continue to Lend to the Financial Economy, Not the Real One (As The Consumer Turns)
Non-bank lending and the transmission of monetary policy (BIS: Cucic and Gorea, 2024)
The Disparate Outcomes of Bank‑ and Nonbank‑Financed Private Credit Expansions (Libery Street: Boyarchenko and Elias, 2024)
Private credit thrives in darkness (FT Alphaville)
Covenant Lite #23: The Private Credit Performance Illusion (Covenantlite)
The explosion of jumbo facilities squares with Gorton et al.’s prediction that each extra intermediary widens the funding valve by lowering any one lender’s monitoring incentive.