Will Private Credit Lead To Another Financial Crisis: Goldman Answers

Amid the constant barrage of negative private credit news, which most recently saw some of the marquee names in alternative asset management such as Apollo, Ares and Blackrock all gate their investors following a historic surge in redemptions requests by retail and high net worth clients, many are wondering whether the private credit crisis will remain contained to a (still) relatively obscure corner of the global financial universe of will they spill over? 

In a note published by Goldman's Manuel Abecasis, the economist puts private credit risks in a macroeconomic context to answer that question. Not surprisingly, he reaches a rather optimistic conclusion, finding that "private credit stress is unlikely to generate large macroeconomic spillovers on its own" claiming that while "lending by private credit firms will likely tighten in coming months, bank lending to businesses has accelerated recently, corporate sector balance sheets are healthy, and increased AI-related investment demand will likely be a tailwind to credit growth." Still, the bigger risk is that while private credit is unlikely to generate large growth spillovers on its own (and we will counter this optimistic view with a much more pessimistic read on the topic tomorrow), he still cautions that higher overall credit spreads as a result of AI-related uncertainty or a broader tightening in financial conditions would pose larger risks.

We excerpt from his post below.

Concerns about the outlook for private credit have increased recently and driven a pickup in retail investor redemptions from several funds. In this report, Goldman frames the potential macroeconomic impact of private credit losses on economywide lending and GDP growth.

The private credit industry has grown quickly in recent years, and it currently holds about $1.7tn in levered loans to the corporate sector, or 4% of all credit to the private non-financial sector. While the sector has grown, it is still small relative to the rest of the financial system - for context, residential mortgages amounted to about 45% of credit to the private non-financial sector in the runup to the 2008 financial crisis (this is a clear counter to the flurry of comparisons such as this one from BofA's Michael Hartnett to the 2008 financial crisis).

There is limited visibility into these loans, but available indicators suggest that loan performance as of 2025 Q4 has remained broadly in line with its average since 2023 (although that is changing for the worse as a result of the prepondrance of software-linked loans). The share of underperforming loans in private credit companies’ investment portfolios has increased slightly in 2025 H2 but remained below 2023 levels (left side of Exhibit 2). And while the share of loans with payment-in-kind (PIK) options — the ability to issue additional securities to pay interest and a potential indicator of borrower stress short of outright default — is higher now than it was before 2023, this mostly reflects the fact that more recent loans have been originated with a PIK option, rather than migrating into PIK because of borrower stress. Indeed, the share of borrowers who transition into PIK has remained stable recently (right side of Exhibit 2).

Nevertheless, instances of fraud on a few large loans and the rapid growth in private lending in recent years have sparked concerns that credit performance could deteriorate, and investors in these loans are locked into lower rates of return than are now available.

Additionally, the possibility of AI-related disruption in the software sector - which catalyzed the recent revulsion across the sector - has also increased concerns about credit losses at private credit companies with software exposure. Goldman's equity analysts estimate that the software sector accounts for a little under 25% of business development company loans (left side of Exhibit 3). The right side of Exhibit 3 shows that tech company borrowers are more levered than other borrowers in private credit. In addition, software loans might have lower loan recovery rates than other sectors because they have fewer tangible assets that can be used as collateral in loans

The private credit sector has become increasingly interconnected with other financial institutions in recent years, with insurance companies in particular boosting their allocation to the sector while increasing leverage and relying more on short-term wholesale funding (see "Insurance Companies Crushed As Private Credit Contagion Spills Over"). Banks have also become more closely interconnected with private credit by supplying loans and credit lines.

To gauge the potential effects of higher losses in the private credit sector on economy-wide lending, Goldman combined its estimates of the relationship between private credit and other financial institutions based on observations from the bank's equity analysts and the Fed’s “from-whom-to-whom” dataset, conservative estimates of recovery rates from the bank's credit strategists, and estimates of the extent to which different types of financial institutions would pull back on lending to preserve capital in response to private credit defaults.

Goldman estimates that a scenario in which private credit default rates rise from about 1% in 2025 to 3-4% (the lower end of the range for leveraged loans in prior credit cycles) would result in around $45bn in additional defaults, which translates to about $25bn in losses given a conservative 40% recovery rate. This would result in a small drag of 0.2% or less on the stock of loans (or roughly 1.5% or less of the gross flow of new lending to the private sector) even after accounting for a pullback in lending from other financial institutions.

A more severe risk scenario in which private credit default rates rise to 10% (the top of the historical range for leveraged loans) would result in about $150bn in defaults, which translates to about $90bn in losses under a 40% recovery rate. If the recovery rate were 30%—say because defaulted software loans see lower recoveries than even the bank's conservative assumption—the losses would be about $105bn.

This, in turn, could result in a $350-400bn drag on overall credit to the non-financial private sector, corresponding to 5-6% of the gross flow of new lending to the private sector (or 0.7-0.9% of the stock of loans) if the shock is concentrated in a few funds and affects lenders to private credit, like banks. If the shock is spread out across many private credit firms, the lending pullback would be smaller because institutions like banks would be less affected. Recent estimates of counterparty exposure to private credit by the Treasury’s Office of Financial Research using confidential data suggest the ultimate impact could be even smaller. To put these estimates in context, there was a 30% pullback in the flow of lending to the private sector in 1990 around the recession and savings and loans crisis and a 55% pullback in the wake of the 2008 financial crisis.

This lending pullback will probably not translate one-for-one into lower output, in part because unaffected lenders can fill some of the gap. To gauge the impact of a slowdown in lending on output, Ggoldman estimates a vector autoregression model with its financial conditions index, an indicator of credit supply shocks developed by Fed researchers based on the Senior Loan Officer Opinion Survey (SLOOS), real GDP growth, and bank lending growth. As shown in Exhibit 4, the bank estimates that a credit supply shock that results in a 1% slowdown in lending as a share of GDP results in a 0.3-0.4% hit to real GDP. While it is difficult to cleanly identify the impact of credit supply shocks on growth, these estimates are broadly similar to other estimates from economic studies

Combined with these estimates, Goldman's 3-4% default rate scenario would generate a small drag of about 0.1% on GDP. The more extreme 10% default scenario would generate a drag of 0.2-0.5% on GDP. Smaller or less concentrated shocks would generate smaller growth effects. Goldman's estimates of the macroeconomic impact of even severe losses are only moderate in part because the private credit sector is relatively small and its liabilities are less runnable than those of other financial institutions. This, on the surface, can be a naive assumption if, in retrospect, there are indirect effects across the lending space, and the loan issuance channel is shuttered not due to direct exposure of regulatory limits but purely psychologically due to the market stress that would accompany a historic credit event that a 10% mass default event (with 20-40% recoveries) would engender. Also, keep in mind that UBS recently raised its base case forecast for 15% defaults across private credit, which is far worse than Goldman's worst case scenario, and where the downstream consequences, both direct and indirect, would be far more dire. For those who missed it, we suggest rereading it as the UBS conclusion is much more alarming than Goldman's envisioning record "cascading defaults" and widespread contagion. 

Here Goldman notes that while it is framing its analysis in terms of default scenarios for private credit, the bank stresses that 1) private credit loan defaults don’t translate into monetary losses to the same extent as other loans because there are more covenants in private credit loans that can be breached before a company misses interest payments; and 2) private credit loans are more senior in borrowers’ capital structures now than in the early days of the industry, so higher private credit defaults would likely overlap with losses in other asset classes (which is negative). For the purpose of this analysis, the bank focuses on scenarios where stress originates in private credit loans in order to isolate the specific risks posed by the sector.

Goldman concludes that according to its analysis, "private credit stress is unlikely to generate large macroeconomic spillovers on its own" although a growing chorus of pessimists would disagree. And while lending by private credit firms will likely tighten in coming months, corporate sector balance sheets are healthy - assuming, of course, that the Iran war leads to a prompt resolution that does not spark a global stagflationary recession, and the AI bubble does not burst - bank lending to businesses has picked up, and increased AI-related investment demand will likely be a tailwind to credit growth.

Still, Goldman cautions that while private credit is unlikely to generate large growth spillovers on its own, higher overall credit spreads as a result of AI-related uncertainty about companies’ future prospects or a broader tightening in financial conditions would pose larger risks.

While the Goldman conclusion may appear somewhat disconnected with reality, tomorrow we will follow up with a far more pessimistic assessment of what a private credit crisis would mean for the US economy. 

More in the full note available to pro subscribers.