The $3 Trillion Market Whose Regulator Just Used the Word "Circular" — And Every Equity Investor Is Ignoring It
Private credit deep dive. The NAIC is now asking whether Apollo- and KKR-sponsored investment vehicles indirectly own pieces of themselves. Fitch's private credit default rate is pinned at an all-time high. Ares Capital just wrote down 3% of book value in one quarter — and still trades at 0.99x NAV, while Blue Owl's BDC trades at 0.76x. That gap is the trade.
Published Monday, August 3, 2026 · Prices verified intraday 12:07 p.m. ET · @dailyanalysts · Standalone private credit piece #1, week of Aug 3–9
HIGHEST-CONVICTION FINDING: The equity market has stopped pricing private credit risk as a sector and started pricing it as a brand. Ares Capital (ARCC) trades at 0.99x its just-reported June 30 NAV after a quarter in which book value per share fell 3.0%, five borrowers went on non-accrual, GAAP EPS came in at $0.24 against $0.50 of net investment income, and five brokers cut targets. Blue Owl Capital Corp (OBDC) trades at 0.76x a NAV that is 4 months stale, having already cut its dividend and authorized a $300M buyback. The Cliffwater BDC index sits at a 17.6% discount to NAV. ARCC's ~17-point premium to its own sector is not a quality premium anymore — it is a stale reflex. Long OBDC / short ARCC is the cleanest risk-adjusted expression of the private credit cycle available today.
1. What actually happened — the news most equity investors skipped
Three items landed in the last five sessions. None of them moved the S&P 500. All three change the private credit picture.
(a) The NAIC put the word "circular" in writing
US insurance regulators opened a review of structured investment vehicles sponsored by Apollo and KKR — specifically Apollo's Multi-Asset Prime Securities (AMAPS) and KKR-sponsored securitisation vehicles reported as Thunderbird and Lightning. Per the July 31 report, the National Association of Insurance Commissioners is examining the possibility of "circular ownership, where investment structures could indirectly own interests in themselves or repeatedly invest in overlapping assets, creating greater concentrations of exposure than insurers realize." Regulators warned that "as the market expands, separate investment vehicles could increasingly invest in one another," and are weighing additional look-through disclosure requirements on the underlying assets. Apollo defended the programme, saying its multi-asset securities hold diversified assets.
Why this matters, mechanically: insurance balance sheets are the single largest incremental buyer of private structured credit. Rated-note and multi-asset structures are how illiquid loans get converted into NAIC-designated paper an insurer can hold at favourable capital. If look-through disclosure is imposed, nothing defaults — but the speed at which new private credit can be placed into insurance balance sheets slows. That channel is the fastest-growing fee engine at Apollo and KKR. Disclosure is not a capital charge, and this is a slow-moving committee process; treat it as a 2027 multiple risk, not a 2026 earnings event.
(b) Fitch's private credit default rate is stuck at a record high
Fitch's US Private Credit Default Rate (PCDR) remained at an all-time-high 6.0% for the trailing twelve months through May 2026 — 14 defaults in the month, with 55% of unique defaulters carrying EBITDA of $25M or less. The critical detail from the report: the $0–25M EBITDA cohort's default rate rose to 11.5% from 11.0% in April. Default mechanics were maturity extensions under stress, PIK introduction, and uncured payment defaults — i.e. the "amend-and-extend" pipe is full. Context: Fitch put full-year 2025 private credit defaults at a record 9.2%.
Consequence: "the headline is flat at 6.0%" reads like stabilisation. It isn't. The small-borrower tail is still deteriorating, and PIK conversions mean reported yields overstate cash income. That flows directly into BDC dividend coverage below.
(c) The retail money is still trying to leave
Semi-liquid, non-traded private credit vehicles remain gated. Apollo Debt Solutions ($26B) received redemption requests equal to 16.8% of shares against a 5% quarterly repurchase cap — meaning roughly two-thirds of exiting investors were told to wait. Blue Owl capped redemptions at two private credit funds again in July, while telling investors pressure "eased" in Q2 and that the vast majority of investors in its $34B Blue Owl Credit Income Corp stayed put.
2. The number nobody is talking about: US private credit is now more expensive than Europe
This is the most under-discussed data point of the past week. Per Houlihan Lokey's Private Credit DataBank, European direct lending loans now price about 4bps tighter than comparable US loans — versus a historical European premium of roughly 22bps. A 26bp relative swing sounds trivial. It is not: it is the price signature of a liquidity constraint. US lenders sitting on gated retail vehicles are hoarding cash rather than deploying it, so US borrowers pay up. Concrete comparison in the same report: a €425M unitranche for GBA Group (CVC / Goldman Sachs / Apollo) priced at ~E+475, while Blue Owl and Blackstone's US financing for Caris Life Sciences priced at ~S+500.
MY OPINION: Consensus is reading this as "Europe is hot." That is the wrong causal arrow. Europe is not hot; the US direct lending market is capital-constrained because its own funding vehicle is gated. The evidence is Blue Owl's own quarter: net deployment of just $600M on $3.6B of originations. Lenders are recycling, not growing. A lender that cannot grow its book cannot grow fee-paying AUM — and that is the timing bomb inside the alternative asset manager equities that ripped higher today.
3. Blue Owl's quarter: FRE is a lagging indicator, and the market bought it anyway
Blue Owl (OWL) reported Q2 on July 31 and the stock rose ~6% on the print. It is up another 5.5% today to $10.87. Here is what was actually in the release, per the Q2 summary:
Metric
Q2 2026
Comparison
Read
Fee-related earnings
$392.2M
+9% y/y, above consensus
Lagging — priced off AUM raised 4–8 quarters ago
AUM
$319B
+12% y/y
Lagging
Total fundraising
$7.6B
vs $12.1B a year ago (−37%)
Leading — the tell
Credit fundraising
$1.8B
Lowest quarterly total in 3 years
Leading — the tell
Direct lending originations / net deployment
$3.6B / $0.6B
Well below prior year
Balance sheet in run-off mode
Direct lending as % of AUM
~35%
vs ~50% two years ago
Mix shift is real but dilutive to margin
The mechanism: fee-related earnings are earned on fee-paying AUM, which is a function of capital raised in prior periods. A 37% year-over-year collapse in gross fundraising and a three-year low in credit fundraising shows up in FRE with a two-to-four-quarter lag. Buying OWL today because FRE beat by a few million dollars is buying the rear-view mirror. Co-CEO Marc Lipschultz's argument — that direct lending is now only ~35% of assets and wealth-channel direct lending only ~11% of fee-paying AUM — is true and is also an admission that the highest-margin franchise is shrinking as a share of the whole.
4. Ares Capital's Q2: read the release, not the headline
ARCC reported June 30 results on July 29. The headline was "core EPS in line." The release says something else.
Line item
Q2 2026
Prior / comparison
Core EPS
$0.47
In line with consensus
Net investment income / share
$0.50
Covers the $0.48 dividend by 104%
GAAP EPS
$0.24
Half of NII — the gap is marks
NAV per share
$19.35
from $19.94 at 3/31/26 → −3.0% in one quarter
Portfolio at fair value
$29.35B
New commitments $2.59B
Weighted avg yield on debt investments
10.3%
Compressing
Non-accruals
2.4% at cost / 1.4% at fair value
Five borrowers moved to non-accrual in Q2
Debt-to-equity
1.15x
Near the top of its historical band
Q3 2026 dividend
$0.48
Unchanged
Three things in that table with consequences
The non-accrual mark is 58 cents on the dollar. 1.4% at fair value against 2.4% at cost implies the troubled loans are carried at roughly 58% of cost. On a $29.35B book that is roughly $300M of write-downs already taken, about $0.44 per share on my estimate of ~705M shares. If ultimate severity is 70% rather than 42%, another ~$200M (≈$0.28/share, 1.5% of NAV) comes out. That is the quantified downside, not a vibe.
The NAV arithmetic does not close. $19.94 + $0.24 GAAP EPS − $0.48 dividend = $19.70. Reported NAV is $19.35. The missing ~$0.35 (1.8% of book) came from capital-account items rather than current-quarter operating results. I do not know the full attribution from the press release alone and I am flagging it as the first question I want answered on the Q3 call. Either way, book value per share fell 3.0% in a quarter in which management declared an unchanged dividend.
Coverage is thinner than it looks. Core EPS $0.47 against a $0.48 dividend is 98% coverage. Each incremental 1% of the portfolio going non-accrual removes roughly $30M of annual interest (10.3% yield on ~$293M), about $0.043/share/year. Two more points of non-accruals takes coverage to ~94%. Layer on rate cuts: on my arithmetic, 100bp of SOFR easing costs ARCC roughly $0.25–0.30/share of annual NII — 13–16% of the $1.92 dividend. ARCC is a 10.0% yielder paying out ~100% of core earnings into a rising-default, falling-rate environment, and it trades at 0.99x book.
The sell side has begun to move: target cuts from Citizens JMP ($23→$22), RBC ($22→$21), Truist ($22→$21), KBW ($21→$20) and JPMorgan ($19.00→$18.50), plus a downgrade to Sell on August 2 citing the five new non-accruals. Note that JPMorgan's $18.50 target is below today's $19.15 price.
Do not mistake today's move for a credit re-rating. Every one of those names is up because President Trump called off strikes on Iran: WTI −7% to $78.59, Brent −6% to $83.03, the 10-year yield −7bp to ~4.67%, Dow +546. Nothing about private credit fundamentals changed between Friday and Monday. Beta rallies are where you sell credit risk you don't want, not where you buy it.
What the 0.76x on OBDC is actually pricing
OBDC's last reported book (Q1 2026, May 6): NAV $14.41, adjusted NII $0.31, base dividend rebased to $0.31, non-accruals 2.0% at cost / 1.0% at fair value, net leverage 1.13x, weighted average yield 10.0%, and a $300M buyback authorised. Portfolio: $15.3B fair value, 230 companies, 78% senior secured, 96% floating rate.
A 24% discount on ~500M shares (my estimate from $7.2B of implied net assets) prices roughly $1.7B of future credit losses — about 11% of the portfolio. To lose 11% of a senior-secured book with 50–60% severities, non-accruals have to run to 18–22% of the portfolio. Non-accruals are 2.0% at cost. The market is pricing a 2009-magnitude middle-market credit event. I think this cycle is a grinding four-to-six-quarter workout with a record-but-plateauing default rate — painful for dividends, not existential for senior-secured book value.
6. Second- and third-order effects most people are missing
Fee streams and credit risk have decoupled — temporarily. Managers (OWL, ARES, APO, KKR, BX) are up 2–5.5% today; the vehicles that actually hold the loans (ARCC, OBDC, FSK, TCPC) trade at 0.76–0.99x book with 10–12% yields. That is rational only if fee streams are insulated from origination volumes. They aren't — they're lagged. Watch Q3/Q4 fee-paying AUM growth, not FRE.
Small-cap credit transmission. US private credit is now the expensive market versus Europe, and the $0–25M EBITDA borrower cohort is defaulting at 11.5%. Those borrowers are the private cousins of the leveraged Russell 2000. IWM below $285 would confirm the credit channel is transmitting into small-cap equity; above $300 says the risk-on bid is winning.
The bank channel is quietly tightening. A quantitative forensic working paper on private credit systemic risk (Djouad, 2026 — a working paper, not peer-reviewed, so weight it accordingly) documents a $648M reduction in an FS KKR credit facility in May 2026 with the remainder repriced higher, and frames private credit's growth from ~$400B in 2009 to ~$3T by mid-2026 across five coupled channels, with a stress convergence window in Q1–Q2 2027. Independent of the paper's conclusions, the observable fact pattern — bank facilities shrinking while asset yields compress to ~10.3% — squeezes BDC net spread from both ends.
Software is now a hated private credit sector. Lenders are re-underwriting software LBO debt for AI disruption risk; a $5B Thoma Bravo-backed Proofpoint refinancing met AI-related resistance in late July. Higher private debt costs for sponsor-owned software raise the discount rate on every legacy software equity, public or private, and shrink the sponsor exit pipeline.
The M&A-recovery trade has a financing problem. If the marginal US financing bid is gated and hoarding liquidity, sponsor exits get pushed into the IPO window at soft prices — visible in a run of Blackstone- and Permira-backed listings debuting muted in the past week. Consensus is long the capital-markets recovery; the private credit plumbing says the recovery is being rationed.
7. Positioning — specific, with levels
TRADE 1 — LONG OBDC / SHORT ARCC (dollar-neutral pair) · HIGH CONVICTION
Entry zone: long OBDC $10.55–11.10; short ARCC $18.85–19.60. Current pair: 0.76x vs 0.99x P/NAV — a 23-point gap.
Sizing/execution: establish the ARCC short leg in full now; put on half the OBDC leg before Wednesday's print, half after. Event risk on Aug 5 is real and I will not pretend otherwise.
Target: gap compresses to ≤12 points inside 1–3 months, i.e. OBDC ~0.85x (≈$12.10 on a $14.20 NAV) and ARCC ~0.90x (≈$17.40 on a $19.30 NAV). Expected pair return +8% to +14%.
Carry: long OBDC yields 11.3%, short ARCC costs 10.0% — the pair is roughly carry-neutral (+130bp). This is precisely why you express this view as a pair rather than an outright short: naked-shorting a 10% yielder bleeds 2.5% per quarter in dividends.
Invalidation (any one): OBDC reports Q2 NAV below $13.80 or non-accruals above 3.5% at cost on Aug 5 (then the discount is informative, not excessive — cut the long leg, keep the short); or ARCC closes above $20.50; or ARCC's Q3 print shows NAV flat-to-up with non-accruals back below 2.0% at cost.
Timeframe: 1–3 months. Two independent signals: (1) ARCC's own Q2 disclosure — NAV −3.0%, five new non-accruals, GAAP EPS half of NII; (2) the sector benchmark at 0.82x P/NAV plus five broker target cuts and a Sell downgrade, one target ($18.50) already below spot.
TRADE 2 — LONG OBDC outright (existing position, MAINTAINED) · HIGH CONVICTION
Status callback: opened July 31 at $10.30–10.90 with a $12.75 target. Spot $10.98 — marginally above the zone, thesis intact, nothing invalidated (NAV $14.41 vs $13.50 trigger; non-accruals 2.0% vs 4% trigger).
My pre-print estimates: NAV $14.05–14.25 (−1% to −2.5% q/q), adjusted NII $0.30–0.32, non-accruals 2.0–2.8% at cost. The single most important disclosure is how much of the $300M buyback was executed — at 0.76x book, every $100M repurchased adds roughly $0.06–0.07 to NAV per share and signals the board disagrees with the market's loss estimate.
Target $12.75 (0.88x NAV). Invalidation: NAV <$13.50, another dividend cut, non-accruals >4% at cost, or a weekly close below $10.20. Timeframe 6–12 months.
TRADE 3 — SHORT / UNDERWEIGHT OWL · HIGH CONVICTION (upgraded from Speculative)
Entry zone: $10.60–11.40 (spot $10.87 after a +5.5% beta pop — this is the gift).
Thesis: FRE +9% is a lagging print; gross fundraising −37% y/y and credit fundraising at a three-year low ($1.8B) are the leading prints, and net deployment of $600M means the origination engine is idling. Redemption caps were re-imposed in July even as management called a turning point.
Target $8.75 (−20%). Invalidation: weekly close above $11.90, or Q3 credit fundraising re-accelerating above $3.5B, or a quarter with net deployment above $2B. Timeframe 1–3 months.
Honest risk: this is a consensus-crowded short and management is actively marketing a "turning point." If redemption requests fall again in Q3, the multiple re-rates before the FRE lag ever shows up. Size accordingly — half the size of Trade 1.
TRADE 4 — APO: hold, but the NAIC review is now a live overhang · WATCH
Spot $127.92. Existing long (entry $118–126) is above zone and intact. I am not adding here.
AMAPS is named in the NAIC circular-ownership review. Disclosure requirements do not hit 2026 earnings, but they slow the insurance-placement channel that underwrites Apollo's growth narrative — and they invite headline risk at every NAIC working-group call this autumn.
Levels: a close below $112 invalidates the long. A break below $120 on NAIC headlines is the trigger to buy 3-month puts as a hedge rather than sell the stock. Above $135, the market has dismissed the review and the overhang is off the table.
TRADE 5 — Continue to AVOID FSK and TCPC · HIGH CONVICTION avoid
FSK $10.98 (+3.9%) and TCPC $3.36 (+3.4%) are the highest-beta expressions of exactly the risk I want to be short, with weaker balance-sheet flexibility. A $3 handle on TCPC is not value; it is a market telling you the equity is an option on the workout. If you must own the discount, own it in OBDC, where the dividend has already been rebased and the buyback is authorised.
8. Three scenarios into Wednesday's OBDC print
Scenario
Probability
Trigger conditions
Consequence
Bull
25%
NAV ≥ $14.30; non-accruals ≤2.0% at cost; more than $75M of the $300M buyback executed; NII ≥ $0.32
OBDC to $11.80–12.40; discount to 0.83–0.86x; the whole BDC complex re-rates 5–8% and the pair pays out through the long leg
Base
55%
NAV $14.00–14.25; adjusted NII $0.30–0.32; non-accruals 2.0–3.0% at cost; modest buyback
OBDC $10.75–11.50; discount grinds in slowly; the pair pays out mainly through ARCC de-rating toward $17.50–18.00 as its premium to the sector closes
Bear
20%
NAV <$13.80 (>4% q/q decline), or non-accruals >3.5% at cost, or further dividend action
OBDC $9.75–10.25; cut the long leg on the invalidation, keep the ARCC short — in this scenario ARCC's 0.99x becomes indefensible and does the work alone. Sector P/NAV goes to 0.75x
9. Timeline: how this evolves
Wed Aug 5 (after close) / Thu Aug 6 10:00 ET: OBDC Q2 — NAV, non-accruals, buyback execution. The single most important private credit data point of the month.
Fri Aug 7: July payrolls (consensus +87.5k, unemployment 4.3% from 4.2%). A weak print pulls cuts forward — good for BDC discounts, bad for BDC net investment income. That tension is why I want the sector expressed as a pair, not a direction.
Mid-August: next monthly Fitch PCDR. Above 6.0% and the "plateau" narrative dies; the $0–25M EBITDA cohort above 12% is the number to watch.
September–October: NAIC working-group sessions on structured/multi-asset vehicles. Any move from "disclosure" toward "capital treatment" is the moment to get properly defensive on APO and KKR.
Late October: Q3 BDC prints. Two consecutive quarters of NAV declines with unchanged dividends is how dividend cuts get pre-announced. ARCC is the one to watch.
Q1–Q2 2027: the refinancing wall. Both the working-paper analysis and Fitch's amend-and-extend evidence point to the same window.
10. Bottom line
Private credit is a ~$2–3 trillion asset class with a record 6.0% default rate, gated retail funding, a regulator publicly asking about circular ownership, and public vehicles trading at an 18% average discount to book. Within that, the market is charging Blue Owl's BDC a 24% discount for risk it has already recognised and charging Ares Capital nothing for risk it is still recognising. I do not think this is a systemic 2009 repeat; I think it is a grinding workout in which dividends get cut and book values bleed 5–10% over four quarters — and in that world relative value beats direction.
Actionable summary:
Put on long OBDC / short ARCC now (ARCC leg in full; OBDC half now, half after Wednesday). Target +8–14% over 1–3 months. Invalidation: OBDC NAV <$13.80 or non-accruals >3.5%; ARCC above $20.50.
Sell the beta rally in OWL ($10.60–11.40 entry, $8.75 target, invalidation weekly close >$11.90) — FRE is the rear-view mirror.
Hold APO, don't add. Hedge with puts on a break of $120; invalidation below $112.
Keep avoiding FSK and TCPC. Same risk, worse balance sheets, no rebased dividend.
Watch IWM $285 as the tell for whether private credit stress is transmitting into public small-cap equity.
Prior call resolved this session
ARCC "hold / quality anchor" (opened Jul 31) is INVALIDATED by its own trigger. The stated invalidation was "NAV <$19 or non-accruals >2.0%." NAV held at $19.35, but non-accruals printed 2.4% at cost. I am not going to fudge that: the hold thesis is dead and ARCC moves to the short side of the pair above. This is exactly what pre-committed invalidation levels are for.