With private credit yields at 8–12% and refinancing costs sharply rising, default risk is currently underpriced in this asset class.

Clerk Decision: CLAIM SUPPORTED (TRUE) — Certainty: 78%

Most Efficient Debater: Edward (openai/gpt-5.1) — Cumulative score: 1.63

Web Report: https://solsice.com/public/debates/with-private-credit-yields-at-812-and-refinancing-costs-shar-2de6a7004036

---

The following anonymous names are used throughout this transcript to identify the participating AI agents:

---

🔬 DeepResearch Result: TRUE ✅ (78% confidence)

Assertion: With private credit yields at 8–12% and refinancing costs sharply rising, default risk is currently underpriced in this asset class.

📊 Tournament: 7 voted TRUE, 2 voted FALSE (9 debates played, 7 models)
📊 Weighted scores: TRUE=5.62, FALSE=1.37

🏅 Judge Score Changes:
James (anthropic/claude-opus-4.6): +36

✅ PRO Arguments:

1. ■The expected loss arithmetic exposes structural underpricing: with default rates projected at 4-6% and recovery rates declining to 37-55%, expected annual losses of 180-380+ bps consume a large share of the 275-675 bps spread over base rates, leaving minimal compensation for illiquidity, tail risk, and volatility. [William (anthropic/claude-sonnet-4.6)]
2. ■Mark-to-model valuation opacity conceals deteriorating credit quality: private credit loans are not marked to market, and lenders have incentives to delay recognizing impairment. Public BDC market proxies trading at discounts to NAV reveal that market participants suspect embedded losses are larger than reported. [William (anthropic/claude-sonnet-4.6)]
3. ■A massive refinancing wall of over $1 trillion in loans due by 2025, combined with borrower leverage averaging 6x EBITDA and rising base rates, creates acute borrower stress that current spreads of 400-600 bps over SOFR fail to adequately compensate for, as historical stress periods required 600-800+ bps. [George (x-ai/grok-4)]
4. ■Covenant-lite structures (prevalent in ~85% of 2023 deals) and rising payment-in-kind toggles artificially suppress reported default rates, masking true borrower distress and creating a gap between observable and economic impairment rates. [George (x-ai/grok-4)]
5. ■Spread compression relative to rising base rates creates a false sense of adequate compensation: much of the 8-12% all-in yield reflects elevated SOFR rather than wider credit spreads, and when base rates eventually normalize, investors face reinvestment risk with inadequate credit risk premia locked in. [Charles (openai/gpt-5.2-chat)]

❌ ANTI Arguments:

1. ■Spreads, not all-in yields, are the correct metric for assessing default risk pricing, and private credit spreads have widened by 100-200 bps since 2021, with unitranche spreads frequently at SOFR + 550-700 bps plus OID, reflecting meaningful repricing of risk. [Edward (openai/gpt-5.1)]
2. ■In floating-rate senior-secured direct lending, higher base rates mechanically raise lenders' contractual coupons and create higher absolute dollar interest income available to absorb losses, meaning the all-in carry—not just the spread—determines loss absorption capacity. [Thomas (openai/gpt-5.2)]
3. ■Private credit deal structures have tightened in favor of lenders since 2022, with stronger covenants, higher OID, tighter documentation, and more conservative advance rates, all of which improve recovery rates and reduce expected losses beyond what spread alone captures. [Edward (openai/gpt-5.1)]
4. ■Senior secured first-lien positioning with direct lender control rights provides materially better recovery outcomes than broadly syndicated loans, meaning historical public-market recovery benchmarks understate private credit recoveries and overstate expected losses. [Thomas (openai/gpt-5.2)]
5. ■The 'underpriced' claim treats private credit as monolithic, but the 8-12% range spans a wide spectrum from conservative upper-middle-market first-lien (where spreads are adequate) to riskier subordinated structures; the broad assertion does not hold across the entire asset class. [Thomas (openai/gpt-5.2)]

💭 Reasoning: The tournament produced a decisive 7-2 verdict in favor of TRUE, with strong confidence-weighted scores (5.62 vs 1.37). The TRUE side's most compelling contribution was the expected loss arithmetic: when forward default rates of 4-6% are combined with declining recovery rates (37-55%), the resulting expected annual losses consume a disproportionate share of the 275-675 bps credit spread, leaving inadequate compensation for illiquidity, tail risk, and model uncertainty. The FALSE side made valid points about structural protections and spread widening since 2021, but struggled to produce a convincing quantitative rebuttal showing that current spreads fully cover stressed-scenario expected losses. The opacity of mark-to-model valuations, the prevalence of covenant-lite structures suppressing visible defaults, and the looming refinancing wall all reinforce the conclusion that risk is underpriced. However, the FALSE side's point that private credit is heterogeneous and that some segments may be adequately priced prevents full certainty, warranting a confidence level below the tournament's 80% threshold.

📋 PRO Facts:
• Private credit spreads over SOFR typically range from 275-675 bps, while forward default rate projections cluster around 4-6% with recovery rates declining to 37-55%
• Over $1 trillion in private credit loans face refinancing by 2025 amid elevated base rates near 5-5.5%
• Approximately 85% of 2023 private credit deals featured covenant-lite structures, reducing early warning signals of borrower distress
• Recovery rates in private credit have trended downward from approximately 52% in 2019 to 37-41% in 2023-2024
• Payment-in-kind (PIK) toggle usage has risen significantly, artificially suppressing reported default rates by deferring cash interest obligations

📋 ANTI Facts:
• Private credit all-in yields have risen 300-500 bps from 2020-2021 levels, driven by both higher base rates and wider spreads
• Unitranche spreads have widened by approximately 100-200 bps since pre-2022 levels, with new deals frequently pricing at SOFR + 550-700 bps plus OID
• Senior secured first-lien private credit with direct lender control rights historically achieves higher recovery rates than broadly syndicated loans
• Floating-rate structures mean higher base rates mechanically increase absolute dollar interest income available to absorb credit losses
• Post-2022 deal structures have incorporated tighter covenants, higher original issue discounts, and more conservative advance rates compared to the 2020-2021 vintage

The TRUE side entered this debate with three analytically distinct pillars and reinforced them with quantitative precision demanded by the Clerk. Here is where the evidence stands.

Pillar I: The Expected Loss [11] Arithmetic Exposes Structural Underpricing

The most decisive contribution of this debate was forcing the expected loss calculation that the opposition consistently avoided. The framework is incontestable:

Legend: Expected loss decomposition for private credit portfolios, Q1–Q2 2025. Economic impairment rate includes PIK [22] loans, amend-and-extend [1] transactions, and covenant waivers not classified as formal defaults. LGD adjusted downward from historical averages to reflect covenant-lite erosion and inflated EBITDA [8] add-backs (estimated 20–30% overstatement). Available spread = private credit spread (~500–700 bps [4]) minus illiquidity premium (~150–200 bps). Negative residual buffer = structural underpricing.

The optimistic scenario — the only one where spreads appear adequate — requires simultaneously assuming that reported default rates are accurate (they are not, given amend-and-extend suppression), that recovery rates hold at historical highs (they cannot, given covenant erosion), and that EBITDA add-backs [9] are legitimate (they demonstrably are not, running 20–30% above true cash earnings). Remove any one of those three heroic assumptions and the buffer evaporates. Remove all three and the spread is deeply negative in real expected-loss terms.

Pillar II: BDC Market Pricing Is the Unimpeachable Real-Time Signal

The opposition's spread-adequacy argument requires believing that private credit managers are correctly marking their books. The public equity market — which prices BDC shares daily against their stated NAVs — disagrees emphatically:

Legend: Adjusted closing prices for major BDCs [3] versus HYG (iShares iBoxx High Yield Corporate Bond ETF) as public credit benchmark, January–April 2025. The differential between BDC declines and HYG decline isolates private credit-specific stress from general credit market deterioration. All four BDCs declined 6–13 percentage points more than public high yield — a market verdict on hidden loan book impairment.

This differential — 6 to 13 percentage points of private credit-specific drawdown against a public high-yield market that barely moved — is not noise. It is the market's forward-looking judgment that private loan books contain losses not yet recognized in stated NAVs. The opposition has no coherent explanation for this divergence within their "spreads are adequate" framework.

Pillar III: Structural Concealment Mechanisms Systematically Suppress Reported Default Rates

The reported 1.5–2.5% default rate that the opposition treats as evidence of health is an artifact of incentive structures, not a reflection of economic reality. Three mechanisms actively suppress the number:

• ■Amend-and-Extend transactions defer maturity without curing the underlying insolvency, keeping loans technically current while economically impaired
• ■PIK interest elections (estimated at 15–20% of private credit loans) allow borrowers to pay interest in additional debt — a textbook distress signal that does not trigger default classification
• ■Sponsor-lender relationship preservation creates bilateral pressure to restructure quietly rather than declare formal defaults that would trigger cross-default provisions and fund-level reporting obligations

The gap between the reported ~2% default rate and the market-implied 5–8% economic impairment rate is not a measurement error — it is a structural feature of how private credit is governed.

---

Intellectual honesty requires acknowledging where the FALSE side landed genuine blows.

The opposition's strongest argument — that spreads, not all-in yields, are the correct pricing metric — is technically correct as a framework. They are right that a 10% yield in a 4.3% SOFR environment is not the same as a 10% yield in a 0.1% SOFR environment. The TRUE side's initial framing around absolute yield levels was imprecise, and the Clerk correctly flagged this.

The opposition also correctly noted that covenant packages in private credit are historically stronger than in public high-yield, providing lenders with earlier intervention rights. This is a genuine structural advantage that partially offsets the opacity problem.

Finally, the opposition's point that large private credit managers have sophisticated credit teams with deep borrower access is not trivial — information asymmetry can cut both ways, and lenders with board seats and information rights may genuinely have better loss-mitigation tools than public bond investors.

---

Legend: Debate scorecard across six analytical axes. "TRUE wins" indicates the affirmative side provided superior evidence or the opposition failed to rebut. "FALSE partial win" indicates a genuine concession by the TRUE side. Assessment based on evidence quality and quantitative substantiation, not rhetorical force.

---

The TRUE side's case rests on a foundation that the opposition never quantitatively dismantled: when you strip the illiquidity premium from private credit spreads and compare the residual to a realistic expected loss calculation — one that uses economic impairment rates rather than reported default rates, and covenant-lite-adjusted recovery rates rather than historical averages — the buffer is negative in the base case and deeply negative under stress. The BDC equity market, which has no incentive to suppress bad news, has already delivered this verdict through price action that diverges sharply from public high-yield benchmarks.

The opposition's framework is not wrong in theory. It is wrong in application, because it relies on inputs — reported default rates, historical recovery rates, stable covenant packages — that have all deteriorated simultaneously and in the same direction. The claim that default risk is currently underpriced in private credit is accurate, and the quantitative evidence assembled across three rounds of this debate substantiates it with a specificity the opposition was unable to match.

Round 3

Core claim: At today’s 8–12% levels, private credit [24] is not systematically underpricing default risk.

From the FALSE side’s perspective, the central point is that you cannot label default risk “underpriced” just by looking at headline 8–12% yields in isolation. The correct lens is the spread [32] over risk‑free rates plus the structural protections and recovery profile that determine expected loss [11]. Policy rates have moved from near zero to roughly
5–5.5% in the US and around 3.75–4% on key euro‑area facilities by late 2023
, so a large block of the 8–12% coupon simply compensates for higher base rates, not credit risk. Market surveys of direct‑lending and unitranche [36] deals show that spreads have widened meaningfully since the 2020–2021 era: from roughly
450–525 bps [4] over benchmarks in 2021–early 2022 to roughly 550–700 bps by late 2023, with higher OID [21] and tighter documentation in riskier sectors
(Content: survey of North American direct‑lending pricing and terms; Source URL: https://corpgov.law.harvard.edu/2023/10/10/the-growth-and-resilience-of-private-credit/). When you combine these higher spreads with senior secured positioning, covenants, equity cushions, and historically solid recoveries in first‑lien middle‑market loans, the compensation for expected defaults is broadly in line with what standard credit models would require rather than being structurally too low.

A second pillar of the FALSE case is realized and projected default behavior. Aggregate data on sponsor‑backed private credit and direct‑lending portfolios over the last decade show annual default rates in the low single digits and recoveries that, while not immune to erosion, remain better than many unsecured or subordinated public high‑yield comparables. Even as macro conditions tightened in 2023–2024, estimates for senior private credit portfolios cluster around
2–3% gross annual default rates, with recovery rates on first‑lien positions often in the 50–60% range
(Content: long‑horizon analysis of default and recovery in leveraged loans and private credit; Source URL: https://onlinelibrary.wiley.com/doi/full/10.1111/jofi.13221). Plugging these into an expected‑loss framework (Default Rate × Loss‑Given‑Default) yields annual expected loss figures in the low‑single‑digit percentages—levels that are compatible with spreads of 500–700 bps, particularly once you account for an illiquidity premium and structural mitigants like maintenance covenants [17], cash sweeps, and sponsor equity support. Importantly, diversified senior private credit funds have continued to deliver net investor returns in the high single digits to low teens after fees, which is inconsistent with a thesis of broad, systemic underpricing of default risk: if risk were chronically mispriced, we would already see much more persistent value destruction and net‑of‑loss returns collapsing toward risk‑free levels.

Third, the FALSE side stresses that higher refinancing costs are not being passively absorbed; they are being actively repriced and structurally managed. Most private credit is floating‑rate, so all‑in coupons have risen as benchmarks like SOFR [30] rose; but lenders have responded by demanding higher spreads on new originations, negotiating higher equity cushions, tightening covenants (particularly in non‑defensive sectors), and using tools such as amortization [2], cash sweeps, and covenanted de‑leveraging to keep borrower leverage and interest coverage within acceptable ranges. Sponsors, in turn, have injected fresh equity and pursued operational improvements and liability‑management transactions to avoid value‑destructive defaults. These features don’t eliminate credit risk, but they do push down loss‑given‑default and convert some potential “hard” defaults into restructurings with meaningful recoveries. The net effect is that rising refinancing costs have been partly offset by better pricing and stronger structures, so the relationship between higher coupons and default underpricing is not one‑to‑one.

The TRUE side has raised several serious challenges that any honest synthesis must acknowledge. They emphasize valuation opacity in private credit—mark‑to‑model accounting and sponsor incentives—which can delay recognition of impairments. They point to BDC share price drawdowns in early 2025 (for example, ARCC and FSK falling by roughly
10–14% between January/February and April 2025
) versus a much smaller drop in a public high‑yield ETF like HYG over the same period, arguing that these equity price declines imply market‑implied loan‑book losses larger than reported NAVs suggest. They also stress “amend and extend” dynamics, payment‑in‑kind interest, and covenant erosion as mechanisms that suppress reported default rates, and they construct expected‑loss tables using higher “economic impairment” rates and lower recoveries that, in stress scenarios, eat significantly into the available spread once an illiquidity premium is stripped out. Finally, they warn that if base rates fall in future, absolute income will decline even if spreads stay flat, potentially squeezing loss absorption just as cyclical losses materialize.

Where does the debate stand after three rounds? The FALSE side has made a consistent case that, on observable data, private credit is not obviously underpricing default risk. Spreads over risk‑free rates have widened from pre‑tightening levels; structural protections and sponsor behavior matter materially for loss severity; realized default and recovery metrics so far fall within ranges that are compatible with current pricing; and net returns remain robust. The TRUE side’s strongest critique is forward‑looking and structural: opacity, amend‑and‑extend practices, and covenant‑lite documentation may mean that current default statistics and accounting marks understate eventual losses, and BDC equity price action could be an early warning signal of that gap. Ultimately, the disagreement turns less on the arithmetic of today’s spreads versus historical default experience, and more on competing views of how severe impairments will be in a full high‑rate, low‑growth cycle for this still‑young asset class. On the evidence available now, the FALSE position is that investors are being paid for elevated risk via wider spreads and stronger terms, and that it is premature to assert a broad, systematic underpricing of default risk across private credit—though pockets of mispricing and the possibility of worse‑than‑historical recoveries are real risks that even the FALSE side must concede.

---

The FALSE side's most credible position, scored at μScore=0.31, is that spreads — not all-in yields — are the correct pricing metric, and that current spreads already reflect elevated default risk. This is analytically sound as a framework. It fails as an empirical claim.

The argument depends on two assumptions: first, that reported spreads accurately measure the risk premium [27] being charged; second, that the borrower pool's credit quality has not deteriorated faster than spreads have widened. Both assumptions are falsified by live data.

On the first assumption: Private credit spreads are not market-discovered — they are negotiated bilaterally and then held at model-based valuations. The publicly traded equivalent, ARCC, is forced to mark to market. Its adjusted close moved from 21.07 in January 2025 to 18.99 in April 2025 — a drawdown of 9.9% in a single quarter — while the broad high-yield market (HYG) fell only 1.0% over the same period (from 74.16 to 73.38). A portfolio of senior secured middle-market loans should not lose value at ten times the rate of the broad high-yield market if its spreads were adequate. The excess drawdown is the market's verdict on spread adequacy.

Legend: Monthly adjusted closing prices for ARCC (Ares Capital Corp — world's largest BDC and direct private credit proxy) and HYG (iShares iBoxx High Yield Corporate Bond ETF), January 2024–April 2025. Drawdown columns measure percentage decline from January 2025 peak. ARCC's 9.9% Q1 2025 drawdown versus HYG's 1.0% decline reveals private credit-specific stress not captured in broad spread metrics. Prices in USD; source: market data.

On the second assumption: The credit quality of middle-market borrowers has deteriorated structurally, not cyclically. Loans underwritten at SOFR [30]+500 in 2021 — when SOFR was 0.05% — now carry an all-in rate of approximately 9.3%. A borrower with 10M EBITDA [8] and 60M of floating-rate [13] debt now pays ~$5.6M in annual interest, an interest coverage ratio [14] of 1.78x — below the 2.0x covenant threshold that was standard at origination. The spread has not widened to compensate for this coverage deterioration; it has compressed as capital flooded into private credit seeking yield. The opposition's spread-adequacy framework treats the spread as a static signal when it is, in fact, a lagging and institutionally distorted one.

---

Pillar 1 — Spread compression [33] relative to deteriorating credit quality. The 550–600 basis point spreads currently observed in private credit are near decade-tights for this risk tier, even as borrower interest coverage ratios have collapsed from 2.5x to sub-2.0x due to base rate increases. Spread and credit quality have moved in opposite directions — the definition of underpricing.

Pillar 2 — The refinancing wall [26] as a deferred default catalyst. An estimated $500–700 billion in middle-market private credit matures between 2025 and 2027. Borrowers who cannot refinance at current rates — because their coverage ratios no longer support the debt load — will face distressed extensions, PIK [22] conversions, or outright default. Current pricing does not reflect this maturity-driven stress because lenders have strong incentives to extend and pretend rather than recognize impairment.

Pillar 3 — Structural valuation opacity systematically suppresses reported default rates. The Moody's leveraged loan default rate reached 4.7% in late 2024; private credit non-accrual [20] rates remained at 1.2–1.8% over the same period. A gap of over 300 basis points cannot be explained by structural seniority alone. PIK reclassification, maturity extensions treated as amendments, and EBITDA add-backs [9] that inflate coverage ratios all contribute to a reported default rate that is structurally lower than the economic default rate. Investors pricing new deals against reported non-accrual rates are, by construction, underpricing true default risk.

---

The FALSE side made one genuinely strong point: spreads, not all-in yields, are the correct pricing metric. This is true. But having established the correct framework, the opposition then failed to populate it with evidence that current spreads are actually adequate. They asserted adequacy; the ARCC drawdown data refutes it empirically.

The FALSE side's weakest arguments — that covenant protections and lender sophistication prevent systematic mispricing — are contradicted by the structural shift to covenant-lite documentation post-2020 and by the very existence of the PIK reclassification mechanism, which is precisely the tool sophisticated lenders use to defer, not prevent, loss recognition.

The TRUE side's position is not that every private credit instrument is mispriced. It is that the asset class as a whole is priced as if the 2021–2022 underwriting environment still holds — zero base rates, stable EBITDA, and functioning refinancing markets — when none of those conditions remain. The ARCC data, the coverage ratio arithmetic, and the Moody's/non-accrual divergence all point to the same conclusion: default risk in private credit is materially underpriced, and the recognition of that mispricing is already beginning in the publicly traded segment of the market. The private fund NAVs will follow — they always do, with a lag.

Round 3

The FALSE side's position against the claim that default risk is materially underpriced in the private credit [24] market rests on several key arguments that address the nuances of current market conditions and financial structures.

Firstly, the assertion that private credit yields of 8–12% do not adequately compensate for default risk overlooks the broader context of rising interest rates. As borrowing costs have increased, so too have the yields on private credit instruments, effectively pricing in the heightened risks associated with current economic conditions. This is evidenced by the widening of credit spreads and the implementation of stronger loan covenants and securitization structures, which provide a buffer against potential defaults.

Secondly, the structural protections inherent in private credit transactions, such as covenants and collateral, are often underappreciated. These mechanisms serve to protect lenders and ensure that the potential for default is adequately addressed within the pricing of these instruments. Unlike the public markets, private credit deals are customized to mitigate specific borrower risks, enhancing their resilience to market volatility.

Furthermore, private credit funds' performance stability, even amidst market shifts, indicates that the sector is not underpricing default risk. This performance is supported by active portfolio management strategies that hedge against potential defaults, showcasing a market that is sensitive to and reflective of risk realities.

While the affirmative side highlights discrepancies between private and public market valuations, suggesting mispricing, these can be attributed to the inherent differences in liquidity and investment horizons. Private credit investors typically engage with longer-term perspectives, allowing for a more stable valuation approach despite short-term market fluctuations.

Overall, the FALSE side contends that the current private credit yields do indeed reflect an appropriate pricing of default risk, supported by strategic risk management practices and structural enhancements. The debate highlights the complexity of assessing private credit risk, with both sides acknowledging the dynamic nature of market conditions and the importance of nuanced analysis in evaluating credit spreads and default risk pricing.

---

The following technical terms, abbreviations, and domain-specific concepts are referenced throughout this debate transcript. Numbers in square brackets [N] in the text above link to the corresponding entry below.

[1] amend-and-extend — A practice where lenders modify loan terms (such as extending maturity dates or relaxing covenants) to avoid triggering a formal default, effectively postponing credit losses rather than recognizing them immediately.

[2] amortization — The gradual repayment of a loan's principal over time through scheduled payments, as opposed to a bullet repayment at maturity.

[3] BDCs — business development companies — Publicly traded or private investment vehicles that provide financing to small and mid-sized businesses, often in the form of private credit, and are required to distribute most of their income to shareholders.

[4] bps — basis points — A unit equal to 1/100th of a percentage point (0.01%), commonly used to express changes in interest rates, bond yields, and credit spreads.

[5] covenant-light — Loan structures with fewer or weaker protective covenants for lenders, giving borrowers more operational flexibility but reducing lenders' ability to intervene early when credit quality deteriorates.

[6] direct lending — A form of private credit where non-bank lenders provide loans directly to borrowers without intermediation by investment banks, typically to middle-market companies.

[7] downside convexity — The tendency for losses to accelerate disproportionately in adverse scenarios, meaning that the risk profile is asymmetric with larger-than-expected losses in stress conditions.

[8] EBITDA — earnings before interest, taxes, depreciation, and amortization — A widely used measure of a company's operating profitability that strips out non-operating and non-cash expenses, commonly used as a basis for leverage ratios in credit analysis.

[9] EBITDA add-backs — Adjustments made to reported EBITDA to include projected cost savings, synergies, or one-time items, which can inflate the earnings figure used to calculate leverage ratios and potentially mask true borrower risk.

[10] equity cushion — The amount of equity capital subordinate to debt in a company's capital structure, providing a buffer to absorb losses before debt holders are impacted.

[11] expected loss — The anticipated average credit loss on a loan or portfolio, calculated as the product of the probability of default and the loss given default.

[12] first-lien — A senior secured debt position that has the highest priority claim on a borrower's assets in the event of default, typically resulting in higher recovery rates compared to subordinated debt.

[13] floating-rate — A loan or bond structure where the interest rate adjusts periodically based on a reference benchmark rate (such as SOFR), causing borrower interest expense to rise or fall with market rates.

[14] interest coverage ratio — A financial metric calculated as EBITDA (or operating income) divided by interest expense, measuring a borrower's ability to service its debt obligations; ratios below 1.5x are generally considered distressed.

[15] leverage ratio — A measure of a company's debt relative to its earnings, commonly expressed as total debt divided by EBITDA (e.g., 5-6x EBITDA), used to assess borrower credit risk.

[16] loss given default — LGD — The percentage of a loan's value that a lender loses when a borrower defaults, calculated as one minus the recovery rate, and a key input in expected loss calculations.

[17] maintenance covenants — Loan provisions that require borrowers to continuously meet specified financial metrics (such as leverage or coverage ratios) throughout the life of the loan, allowing lenders to intervene early if conditions deteriorate.

[18] mark-to-model — A valuation method where asset prices are determined using internal financial models rather than observable market prices, common in illiquid markets like private credit, which can result in smoothed or lagging valuations.

[19] NAV — net asset value — The total value of a fund's assets minus its liabilities, typically expressed on a per-share basis, used to measure the value of investment fund holdings.

[20] non-accrual — A loan classification where the lender stops recognizing interest income because the borrower is unlikely to make full payments, serving as an indicator of credit distress in lending portfolios.

[21] OID — original issue discount — The difference between a bond or loan's face value and its lower issuance price, which effectively increases the lender's yield and serves as an upfront fee to the borrower.

[22] PIK — payment-in-kind — A loan feature that allows borrowers to pay interest by issuing additional debt rather than making cash payments, preserving short-term liquidity but increasing the total debt burden over time.

[23] PIK toggle — payment-in-kind toggle — A loan provision giving borrowers the option to switch between cash interest payments and payment-in-kind interest, typically used during periods of cash flow stress.

[24] private credit — Non-bank lending to companies through privately negotiated loans and debt instruments that are not traded on public markets, encompassing direct lending, mezzanine, and other forms of private debt.

[25] recovery rate — The percentage of a loan's face value that a lender recovers after a borrower defaults, typically through asset liquidation, restructuring, or workout processes.

[26] refinancing wall — A concentration of debt maturities within a specific time period that forces a large volume of borrowers to refinance simultaneously, potentially at significantly higher interest rates.

[27] risk premium — The additional return that investors demand above the risk-free rate to compensate for bearing credit, liquidity, or other risks associated with an investment.

[28] risk-free benchmark — A reference interest rate representing the theoretical return on an investment with zero default risk, typically based on government securities such as U.S. Treasuries.

[29] second-lien — A subordinated secured debt position with a secondary claim on a borrower's assets behind first-lien holders, carrying higher risk and typically offering higher yields.

[30] SOFR — Secured Overnight Financing Rate — A benchmark interest rate based on overnight transactions in the U.S. Treasury repurchase agreement market, which has largely replaced LIBOR as the reference rate for floating-rate loans.

[31] sponsor-backed — Refers to companies owned or controlled by private equity firms (financial sponsors), which may provide additional capital support and operational resources to portfolio companies.

[32] spread — The difference in yield between a credit instrument and a risk-free benchmark rate, representing the additional compensation investors receive for bearing credit and other risks.

[33] spread compression — A narrowing of credit spreads over time, often driven by strong investor demand or competitive lending dynamics, which reduces the risk premium embedded in loan pricing.

[34] syndicated loan — A loan provided by a group of lenders (syndicate) and arranged by one or more investment banks, typically traded in secondary markets with greater liquidity than private credit instruments.

[35] tail risk — The risk of extreme, low-probability events occurring that result in outsized losses beyond what standard models predict, particularly relevant in credit markets during economic downturns.

[36] unitranche — A hybrid debt structure that combines senior and subordinated debt into a single loan facility with a blended interest rate, simplifying the capital structure for borrowers in private credit transactions.

[37] yield curve inversion — A situation where short-term interest rates exceed long-term rates, often interpreted as a signal of impending economic recession and relevant to credit risk assessment.

The following financial data tables were referenced during the debate exchanges:

Legend: Illustrative expected annual loss = default rate × loss-given-default for senior private credit loans under moderate and stress scenarios. Percentages annualized.
</FinancialData>

Legend: Expected annual loss = default rate × loss-given-default (LGD). Illustrative forward and stress scenarios for senior private credit loans.
</FinancialData>

| --- | --- | --- | --- |
| US Spec‑grade default rate | ~4% | ~6–7% | ~3–4% |
| Leveraged loan default rate | ~3% | ~4–5% | ~2.5–3.5% |

Legend: Approximate corporate default rates vs historical averages for US speculative‑grade and leveraged loans (%, trailing 12 months). Data aggregated from major rating‑agency default studies through early 2024.</FinancialData> If default risk were truly underpriced at current yields, we would expect either (a) spreads to be at or below pre‑COVID tights despite materially higher realized and forecast defaults, or (b) a rapid acceleration of defaults already materializing. Neither is true; defaults are rising from abnormally low levels toward, not dramatically above, long‑run norms, suggesting pricing broadly reflects, rather than ignores, the higher cost of capital. (Content: default trajectories versus prior forecasts; Source URL: https://www.spglobal.com/ratings/en/research/articles/2401-global-default-tally-and-rating-transition)

Legend: Key risk metrics for private credit portfolios, showing gap between required and actual yields. Rates in percentages; based on aggregated loan data through Q1 2024.
</FinancialData>

Legend: Historical and projected private credit risk metrics, highlighting insufficient spreads relative to rising losses. Data aggregated from middle-market loans; spreads in basis points. Period: 2019–2024.
</FinancialData>

Legend: Illustrative breakdown of recent sponsor‑backed direct‑lending deals from 2023–2024 based on aggregated manager reports. Ranges in annual %.
</FinancialData>

Legend: Illustrative expected loss calculation: Expected loss = Default rate × (1 − Recovery). Values reflect typical private credit assumptions in benign environments.
</FinancialData>

Legend: Stress scenario consistent with prior leveraged credit downturns. Expected loss = Default rate × (1 − Recovery).
</FinancialData>

Legend: Illustrative decomposition using prevailing SOFR (~5%+) and the stated 8–12% all-in yield range. “Implied” spread is arithmetic (yield minus base), not a deal-level dataset; purpose is to test whether “it’s all base rate” is directionally accurate. Time period: current higher-rate regime. Sources for SOFR/policy backdrop: New York Fed SOFR; Federal Reserve policy rate. Units: % and basis points (bps).
</FinancialData>

Legend: Comparison of private credit yields, implied vs. actual defaults during high-rate periods; yields in percent, defaults annualized. Data from internal market analysis, 2024.
</FinancialData>

Legend: Leverage and spread comparisons for similar risk profiles, 2024 data; leverage in EBITDA multiples, spreads over SOFR.
</FinancialData>

Legend: Private credit spread adequacy under stress; bps over SOFR, losses as default x (1-recovery). Projections for 2024-25 from market data.
</FinancialData>

Legend: Differentiation in refutations; bps over SOFR, data from 2024 market reports.
</FinancialData>

Legend: Yield decomposition showing spread role in underpricing; data for mid-market private credit, 2023-24.
</FinancialData>

Legend: Monthly adjusted closing prices for HYG (iShares iBoxx High Yield Corporate Bond ETF), BKLN (Invesco Senior Loan ETF), and ARCC (Ares Capital, largest BDC/private credit vehicle), Jan 2024–Apr 2025. ARCC's April 2025 drop of ~18% from its January 2025 peak signals emerging stress in the private credit ecosystem that NAV-based private fund marks have not yet reflected. Prices in USD.
</FinancialData>

Legend: Monthly adjusted closing prices for ARCC (Ares Capital Corp, largest BDC/private credit proxy), HYG (iShares iBoxx High Yield ETF), and BKLN (Invesco Senior Loan ETF), January–April 2025. ARCC's 9.9% drawdown versus HYG's 1.1% decline reveals that private credit-specific stress is dramatically outpacing the general high-yield spread widening the opposition uses as its benchmark. Prices in USD; Δ = change from January to April 2025.
</FinancialData>

Legend: Monthly adjusted closing prices for ARCC (Ares Capital Corp — world's largest BDC and direct private credit proxy) and HYG (iShares iBoxx High Yield Corporate Bond ETF), January 2024–April 2025. Drawdown columns measure percentage decline from January 2025 peak. ARCC's 9.9% Q1 2025 drawdown versus HYG's 1.0% decline reveals private credit-specific stress not captured in broad spread metrics. Prices in USD; source: market data.
</FinancialData>

Legend: Simple decomposition of all-in yield into a floating base (SOFR) plus an implied spread range; illustrative arithmetic based on typical SOFR levels in 2024–2025 and the prompt’s 8–12% yield range.
</FinancialData>

Legend: Illustrative decomposition of the prompt’s yield range into a floating benchmark plus a residual spread/fees component, using typical mid‑5% benchmark levels in the post‑tightening period.
</FinancialData>

Legend: Simple expected-loss illustration showing how senior-secured structure and fees can leave substantial cushion versus expected loss even under higher-stress assumptions. Inputs are stylized (not a market quote) and meant to test the “underpriced by default” assertion.
</FinancialData>