A friend of mine in Toronto sold his dental practice a few years ago. A PE-backed rollup chain offered him a number that seemed too good to be true. He took it. The chain promised scale, back-office support, better purchasing power. He’d still practice dentistry. They’d handle the business.
Within two years, the staff was cut. Patient volume targets went up. The purchasing “savings” went to service the debt the PE firm loaded onto the chain to fund the acquisition. My friend doesn’t work there anymore.
I wanted to understand how many dentists, veterinarians, and urgent care doctors had the same experience. So I went back to the data.
Private Credit Stress
In Pebbles and Pebbles II, I parsed 10,210 holdings from 18 BDC SEC filings and built the first cross-fund contagion map of the private credit system. Seven platforms gated over $100 billion in redemptions. Apollo capped at 5% when investors wanted 11.2%. I published that analysis on March 26.
Since then, it’s gotten worse.
On April 2, Blue Owl disclosed that investors in its technology-focused BDC asked to pull 40.7% of their money back. They got 5%. Its $36 billion credit fund saw redemption requests jump from 5.2% to 21.9% in one quarter, a 4x acceleration. $3.2 billion in investor capital is now trapped.
Hours before the Blue Owl disclosure, Apollo’s president Jim Zelter went on Bloomberg Television and called the wave of redemption requests “growing pains” and a “skirmish on the sidelines.” He said the 5% cap was clearly disclosed, “on page one, in black and white,” and that gating was working as planned. “It’s actually quite an easy conversation,” he said.
It is not an easy conversation. The Wall Street Journal reported that investors asked to pull nearly $14 billion from BDC funds in Q1. Up from $5.7 billion the prior quarter and $3.7 billion in all of 2024. The acceleration is exponential: 4x in one year, 4x again in one quarter. Investors were able to redeem roughly half. The rest is trapped.
Blue Owl’s tech BDC backlog could take two years to clear at the 5% quarterly cap. Blue Owl’s co-CEO Doug Ostrover admitted the quiet part out loud: “Between us, and the advisers who sell our products, I don’t think we made it clear enough.” Blue Owl’s co-president Craig Packer called the 41% redemption request a “small minority” from “certain wealth channels and regions.” Forty-one percent is not a small minority. Blue Owl Credit Income posted its worst month since 2022 in February. Blue Owl Technology Income posted its worst month ever. OWL stock is down 43% year-to-date. ARES is down 37%.
The PIK deferrals I flagged in Pebbles are confirmed by Fitch’s data across the full BDC industry: deferred interest payments now account for 8% of BDC interest and dividend income, up from 4% in 2019. Doubled in six years. Morgan Stanley published a report the same day: “In a nutshell, we think risks in private credit are significant.” They expect loan defaults to increase, fundraising to be sluggish, and returns to disappoint.
The game theory is now working against the funds. Brian Jacobs, portfolio manager at Aptus Capital: “There’s this huge game theory incentive to be the first one to leave versus stick around, even if you think the fundamentals are fine.” Investors are over-redeeming because they know they’ll be prorated to 5%. If you want $100 back and you know you’ll get $25, you ask for $400. That’s how 22% becomes 41%. The redemption numbers aren’t just measuring fear. They’re measuring the rational response to gating.
The fear has spread beyond the asset class. Phil Blancato, chief market strategist at Osaic, which advises on $700 billion: “I got a request from a client today saying, ‘get me out of all my private credit investments.’ They weren’t in any.”
And the structural question nobody is answering. John Cocke, deputy CIO of credit at Corbin Capital: “If outflows stay at 5% a quarter for the next year, how do you ultimately get inflows to start coming back in? It becomes an existential question of the vehicle.”
Lloyd Blankfein, former CEO of Goldman Sachs, compared unsold private assets to “tinder on the floor of the forest” just waiting for a spark. “At some point there needs to be a forcing function or a reckoning that causes you to come to grips with what your balance sheet really is worth,” he said on Bloomberg Television.
Blackstone’s chief legal officer responded with an op-ed accusing critics of a “hyperbolic cliché festival.” He wrote this while $4.6 billion in investor capital sits trapped behind withdrawal limits across the industry.
Apollo’s co-president called the defaults in Pebbles I “pebbles.” Apollo’s president called the gating a “skirmish.” Blue Owl’s co-CEO says they didn’t explain the risks. Blue Owl’s co-president calls 41% a “small minority.” Blackstone’s lawyer calls the criticism “hyperbolic.” Morgan Stanley says the risks are “significant.” Lloyd Blankfein says the reckoning is coming. The language from the firms is getting smaller. The language from everyone else is getting louder.
I obtained Blue Owl’s shareholder letter to OTIC investors. It deserves a close read because it shows the two theses converging.
OTIC is Blue Owl’s $6.2 billion technology-focused BDC. 190 companies, 91% backed by PE sponsors, concentrated in software. The letter describes portfolio companies as “mission-critical” businesses “actively adapting to, or already benefitting from, AI-driven innovation rather than being disrupted by it.”
That is the exact language I flagged in SaaSpocalypse. The companies that talk about AI the most on their earnings calls have the worst forward stock returns, a measured effect across 716 SaaS calls (p = 0.043). “Adapting to AI” is the phrase every software CEO uses when the business model is breaking. It’s not a defense. It’s a diagnosis.
The numbers in the letter confirm the stress. OTIC’s leverage is 0.82x against a target range of 0.90-1.25x. That sounds conservative until you realize why: they’re hoarding cash to meet redemptions instead of deploying it into new loans. The fund is shrinking, not lending. They sold $400 million in assets in February, “heavily oversubscribed at 99.6% of par,” which sounds like validation until you realize they’re liquidating performing assets to fund the exit queue. When you’re selling your best loans to pay investors who want out, the remaining portfolio gets worse by selection.
And then there’s the risk disclosure, printed on the same page as the reassurances: “You should not expect to be able to sell your shares regardless of how OTIC performs and you should consider that you may not have access to the money you invest for an indefinite period of time.” That was on page one when investors bought in. It’s on page four now, after 41% asked to leave.
Blue Owl told investors that fresh capital exceeded redemptions. The math relies on counting automatic dividend reinvestment, your own money recycled back into the fund you’re trying to leave , as an “inflow.” Strip out DRIP and the picture is different. 41% asked to leave. 5% were allowed to. The rest had their dividends reinvested into the position they requested to exit. That’s not fresh capital. That’s captive capital with a marketing label.
Two of my published theses, private credit gating (Pebbles) and AI disrupting SaaS (SaaSpocalypse), are colliding in one fund. The investors fleeing OTIC aren’t panicking about credit quality. They’re panicking about whether AI will destroy the revenue of the software companies their fund lends to. The SaaSpocalypse thesis is causing the private credit thesis to accelerate.
That analysis was sector-agnostic: every holding, every company, every PIK loan. This time I filtered for one sector.
Healthcare.
582 holdings. 294 unique companies. Dental chains, veterinary rollups, urgent care networks, behavioral health platforms, dermatology groups, imaging centers, home health agencies. All PE-backed. All leveraged. All sitting inside the same BDC funds that gated redemptions in March.
59 of those holdings are on payment-in-kind. They can’t pay cash interest. They’re paying with IOUs.
A note on PIK: not all PIK is bad. If you’re building a data center or developing a drug, PIK makes sense, because the asset doesn’t generate revenue yet. The PIK I’m describing here is different. These are operating businesses (dental chains, vet clinics, urgent care centers) that were paying cash interest and got amended to PIK because they can no longer cover debt service. That is distressed PIK. It is the leading indicator of default.
8 are in non-accrual. They’ve stopped pretending the money is coming back.
And the defaults have names.
Pathway Vet Alliance. A veterinary rollup acquired by TSG Consumer Partners at 21 times EBITDA during COVID. By March 2025, the company completed a debt exchange: extending maturities, converting cash interest to PIK, stripping the balance sheet to survive. Fair value in Ares Capital’s portfolio: $54.3 million. Status: non-accrual.
Your vet is probably owned by a company that can’t pay interest on its debt.
Absolute Dental Group. A dental rollup. Two tranches in ARCC totaling $52.5 million. Both in non-accrual. The dentist who sold his practice got paid. The PE firm extracted its dividend. The BDC investors holding ARCC shares are absorbing the loss.
Canadian Orthodontic Partners. A dental rollup in BBDC. Payment-in-kind AND non-accrual simultaneously. Paying interest it can’t afford on debt the lender has already written off. This is the financial equivalent of a patient on life support receiving a bill.
These aren’t hypothetical stress scenarios. These are companies that have already defaulted. In funds you can buy on your brokerage account today.
And behind every defaulted rollup is a dentist who sold a practice they spent 20 years building, a veterinarian who joined a network that promised administrative support, a behavioral health therapist who took a salaried position at a PE-backed clinic because private practice was too hard to sustain alone. They exchanged independence for stability. What they got was a platform that cut their staff, increased their patient load, deferred their employer’s interest payments, and, in the case of Steward, couldn’t pay them at all. The extraction starts at the top of the capital structure with the PE dividend. It ends at the bottom with the provider who can’t afford to leave and the patient who can’t afford to go somewhere else.
The playbook.
Private equity did to healthcare services exactly what it did to software companies. The playbook is identical. Buy a fragmented industry (dental practices, veterinary clinics, urgent care centers) at 12 to 15 times EBITDA. Load the platform with 6 to 8 times leverage. Extract a dividend recapitalization within two years. Cut costs by centralizing operations, reducing staff, and increasing patient volume per provider. Sell the debt to BDC funds. Collect management fees regardless of outcome.
The software version of this playbook produced Pluralsight, where Vista Equity Partners lost an estimated $2.2 billion in equity and lenders took 100% ownership through a debt-for-equity swap. It produced the PIK pipeline I documented in Pebbles II: 1,020 holdings paying interest with IOUs, with Fitch reporting that PIK deferrals drove 60% of all private credit defaults in the twelve months through January 2026.
The healthcare version is the same structure with one critical difference: the revenue is government-controlled.
Software companies can raise prices. Dental chains cannot bill Medicare more than Medicare allows. Urgent care networks cannot negotiate higher Medicaid rates when Medicaid is cutting reimbursement. The No Surprises Act cut physician-billed emergency revenue by 20.1% between 2019 and 2023. Medicare reimburses hospitals at 83 cents on the dollar, over $100 billion in annual underpayment. Healthcare labor costs increased $42.5 billion between 2021 and 2023. Contract labor surged 258%.
Revenue is capped by the government. Costs are set by the labor market. And the debt was structured during a zero-rate environment that no longer exists.
When software rollups couldn’t pay interest, they went on PIK. When healthcare rollups can’t pay interest, they go on PIK, but they can’t raise prices to grow out of it. The ceiling is regulatory. The floor is leverage. The space between them is compressing.
The vertical integration problem.
This is where the healthcare thesis diverges from Pebbles. Pebbles was about horizontal contagion: the same company appearing in multiple funds, creating synchronized markdowns. Healthcare has that too. But it also has something worse.
Ares Management owns approximately 40% of Aspen Dental, the largest dental services organization in the United States, operating over 1,000 offices across 43 states. Ares co-owns Aspen with Leonard Green & Partners, with American Securities holding the remaining 20%.
Ares Management also manages Ares Capital Corporation (ARCC), the largest publicly traded BDC in the world, with $28.7 billion in invested assets.
This means Ares is simultaneously the PE owner of Aspen Dental and the manager of the largest BDC in the world. Different pockets, same hand. The same management team that extracted $1.1 billion from a dental chain is making lending decisions for $28.7 billion in healthcare-exposed assets. Three roles. Three fee streams. One company.
Since 2012, the PE sponsors have extracted at least $1.1 billion in debt-funded dividends from Aspen Dental. In June 2021, they took $835 million in a single dividend recapitalization, more than nine years of the company’s annual free cash flow in one transaction. The leverage jumped to 7.4 times EBITDA. Moody’s downgraded the outlook to negative.
Let me make the $835 million concrete. Aspen Dental’s annual free cash flow was approximately $90 million. The PE sponsors took $835 million in one transaction. That is 9.3 years of the company’s free cash flow extracted in a single dividend. The money came from new debt. The debt was structured at rates that assumed a low-rate environment that ended in 2022. The company now services that debt at higher rates, out of the same $90 million in free cash flow, while Medicare cuts reimbursement, while labor costs rise 5.5% annually, while Moody’s watches with a negative outlook.
The dividend went to the PE sponsors. The leverage stayed with the company. The risk transferred to the dental chain’s lenders. Meanwhile, the same firm manages $28.7 billion of other people’s healthcare debt through ARCC.
This is not cross-fund contagion. This is vertical integration contagion. The same firm sits at every layer of the capital structure (owner, lender, fund manager) earning fees at each layer while the downside concentrates in the BDC shares held by retail investors and retirement accounts.
FSOC noticed. On March 25, 2026, the Financial Stability Oversight Council published guidance targeting nonbank financial company structures where the same entity originates, manages, and values the same assets. The Ares/Aspen structure is a textbook example. The regulators are building the framework. The enforcement hasn’t started.
The ARCC healthcare portfolio.
ARCC, the BDC that Ares manages, has 78 healthcare holdings. 16 are on PIK. 4 are in non-accrual.
That is a 20.5% PIK rate for healthcare specifically, double the system-wide average of 10%.
The names:
AthenaHealth. A healthcare IT company acquired by Bain Capital and Hellman & Friedman for $17 billion. Fair value in ARCC: $301.2 million. Status: PIK. The largest single healthcare PIK holding in any BDC fund I analyzed.
Global Medical Response. The largest emergency medical services company in the United States, air and ground ambulances. KKR-backed. $5.3 billion in total debt. S&P downgraded to CCC+. Fair value in ARCC: $80.3 million. Status: PIK. Currently restructuring $4 billion in debt.
AmeriVet Partners. A veterinary rollup. Fair value in ARCC: $64.7 million. Status: PIK. Also held in FSK, cross-fund overlap. When ARCC marks it down, FSK follows.
Modernizing Medicine. Healthcare IT. Fair value in ARCC: $97.2 million across two tranches. Status: PIK. Also held in OBDC, another cross-fund node.
NMC Skincare. A dermatology rollup. $30.7 million across two tranches in ARCC. Status: PIK. Your dermatologist might be owned by a company paying interest with IOUs.
The pattern extends beyond ARCC. GBDC and GSBD each have 13 healthcare PIK holdings, the same overlap pattern from Pebbles II, where those two funds shared 175 identical companies. NMFC has 10 healthcare PIK holdings, including both Aspen Dental and Heartland Dental, the two largest dental rollups in America, in the same fund.
The numbers that matter.
Ninety-three percent.
That is the percentage of the most distressed healthcare companies, those rated B3 negative or lower by Moody’s, that are owned by private equity. Not some. Not most. Ninety-three percent. Out of 45 companies in the bottom-rated healthcare cohort, 42 are PE-backed.
Private equity is not investing in healthcare. Private equity is leveraging healthcare. And when the leverage breaks, the losses cascade through the same BDC funds, the same pension allocations, the same cross-fund contagion network I mapped in Pebbles.
The defaults are not theoretical. They are already here.
Envision Healthcare was the largest physician staffing company in the United States. KKR bought it in 2018 for $9.9 billion. By May 2023, it filed for bankruptcy, the largest healthcare PE failure in history. KKR’s equity was wiped out. Creditors took a 70% haircut. $5.6 billion in debt was extinguished. The company was split in two: AmSurg (surgery centers) went to PIMCO and Blackstone with $2.4 billion in new debt. Envision Physician Services emerged with $250 million. The emergency rooms didn’t close. But the physicians staffing them watched their employer go through bankruptcy while the PE firm that loaded the debt walked away.
The No Surprises Act accelerated Envision’s collapse. Before the law, emergency physicians could balance-bill patients, charging the difference between what the insurer paid and what the provider billed. That practice was predatory. But it was also the revenue line that serviced KKR’s debt. When Congress eliminated surprise billing, physician-billed emergency revenue dropped 20.1%. The debt didn’t drop with it.
Steward Health Care was 31 hospitals across seven states. Cerberus Capital Management bought it in 2010 and spent the next decade extracting value. $1.3 billion left the building during Cerberus’s ownership. When Steward filed bankruptcy in May 2024, it owed $290 million in unpaid employee wages. Nurses weren’t being paid. Supply closets were empty. Patients were transferred to other hospitals because Steward couldn’t buy basic medical supplies.
The real estate was the play. Cerberus had sold Steward’s hospital real estate to Medical Properties Trust in a sale-leaseback, pocketing the proceeds while saddling the operating company with $6.6 billion in long-term rent obligations. When the operating company failed, Apollo Global Management, which held real estate interests, tried to retain the properties. Massachusetts responded by seizing St. Elizabeth’s Medical Center through eminent domain. A state government used emergency powers to take a hospital from a PE firm. That happened in September 2024. In America.
Discovery Behavioral Health treated patients with substance abuse disorders and mental health conditions. Webster Equity Partners and HPS Investment Partners backed it. In February 2026, two months ago, Capital One seized control after a $280 million debt default. The board was dismissed. New management was installed by the lenders. The patients in treatment didn’t get a vote. The therapists providing their care didn’t get a vote. Capital One got the keys.
Thrive Pet Healthcare was formerly Pathway Vet Alliance. TSG Consumer Partners bought it at 21 times EBITDA during COVID, when pet spending was surging and everyone believed it would never decline. It declined. Staffing costs rose. Patient volumes fell. By March 2025, Thrive completed a distressed debt exchange: extending maturities, converting cash interest to PIK, stripping the balance sheet. The company survived. The equity didn’t. In ARCC’s portfolio, Pathway Vet Alliance sits at $54.3 million fair value. Status: non-accrual.
Four major healthcare PE defaults or restructurings in three years. Over $20 billion in debt affected. Hospitals seized by state governments. Employees unpaid. Boards dismissed by lenders. Patients transferred because their hospital couldn’t buy syringes.
I want to stay with Steward for a moment because it illustrates what extraction actually looks like when it reaches its terminal phase.
In the final months before bankruptcy, Steward hospitals were so short on cash that they couldn’t afford to repair broken equipment. Ventilators went unserviced. Operating rooms were understaffed. Documented patient safety incidents included delayed emergency procedures and equipment failures attributed to understaffing and deferred maintenance. Cerberus had already extracted $1.3 billion by that point.
This is what the word “extraction” means. Not in the financial sense of dividend recapitalization. In the medical sense. The PE firm extracted the cash. The hospital extracted the remaining value from its staff and equipment until nothing was left to extract. And then it filed for bankruptcy.
Steward was not in a BDC. It was a larger, more complex structure involving hospital real estate, operating companies, and sale-leasebacks. But the dental chains, vet rollups, and behavioral health platforms in the BDC system are running the same playbook at smaller scale. Affordable Care is FSK’s Steward: a healthcare provider serving vulnerable patients, loaded with debt, paying interest with IOUs. The scale is different. The structure is the same.
The question is whether we wait for the healthcare version of Steward to happen inside a BDC, with the losses hitting retail shareholders and pension funds instead of hospital creditors, or whether the BDC structure provides enough cushion to absorb the defaults. The PIK rate says the cushion is compressing.
And 59 healthcare holdings in BDC funds are currently on PIK, the leading indicator that Fitch says drives 60% of all private credit defaults. The pipeline is loaded.
The contagion map.
I built the same cross-fund overlap analysis for healthcare that I built for the full private credit system in Pebbles II. The methodology is identical: normalize company names across all 18 BDC filings, identify companies held by multiple funds, map the contagion network.
43 healthcare companies appear in two or more BDC funds simultaneously.
GBDC and GSBD, the same fund pair that shared 175 companies in the Pebbles analysis, share 22 healthcare companies. Three of those are on PIK. The diversification illusion from Pebbles is replicated exactly in healthcare. If you own both GBDC and GSBD for “healthcare exposure diversification,” you own 22 of the same healthcare companies twice.
CGBD and ORCC share 13 healthcare companies. PetVet Care Centers is in both. National Dentex Labs is in both. Unified Women’s Healthcare, an OB/GYN rollup, is in both. When one fund marks down the OB/GYN chain, the other fund’s auditors see it on the next quarterly review.
Specialty Dental Brands, a dental rollup operating under the holding company name Vardiman Black Holdings, is on PIK in both GBDC and GSBD. A dental chain you’ve never heard of, paying interest with IOUs, in two funds simultaneously, with synchronized markdown risk.
Convey Health Solutions is on PIK in both ARCC and NMFC. Healthcare services, deferred interest, two funds exposed.
The cross-fund contagion network from Pebbles is not just replicated in healthcare. It’s concentrated. The same fund pairs (GBDC-GSBD, CGBD-ORCC) that dominated the full-system overlap dominate the healthcare overlap. The structure is fractal; zoom into any sector and the same pattern appears.
The pension connection.
In Pebbles II, I traced private credit exposure across five continents through CPP Investments’ ownership of Antares Capital. The healthcare data connects directly to that analysis.
AthenaHealth, on PIK in ARCC at $301.2 million, is a confirmed Antares deal. Global Medical Response, on PIK in ARCC at $80.3 million, is confirmed Antares-backed. These are not separate pools of risk accessed through different doors. They are the same loans, held by the same lender (Antares/CPP) and the same BDC (ARCC/Ares), exposed to the same healthcare reimbursement headwinds.
When AthenaHealth’s PIK converts to non-accrual, when the healthcare IT company that Bain and Hellman & Friedman bought for $17 billion stops even pretending to pay interest, the markdown hits ARCC shareholders and Canadian pension beneficiaries simultaneously. Through different quarterly reports. On different calendars. But the same loss.
And the relationship between Ares and Antares is deepening, not unwinding. On March 31, the Wall Street Journal reported that Antares and Ares formed a $1.7 billion continuation fund, their second in twelve months, to recapitalize a portfolio of over 300 first-lien floating-rate loans. A continuation fund is how private credit avoids marking down assets. Instead of selling at a loss, you roll the portfolio into a new vehicle with fresh capital and a longer time horizon. The financial equivalent of moving a balance from one credit card to another.
Ares is the anchor investor. Antares is the manager. CPP is Antares’s owner. The same three entities that sit atop the vertical integration chain in ARCC’s healthcare portfolio are now creating new vehicles to extend the holding period on loans that investors want to exit. The continuation fund doesn’t reduce the risk. It extends the timeline before the risk becomes visible.
The reimbursement squeeze.
Healthcare PE rollups face a problem that software rollups never had. Software companies can raise prices. Healthcare providers cannot.
Medicare reimburses hospitals at 83 cents on the dollar. That is over $100 billion in annual underpayment across the US hospital system. Physician Medicare payments have no inflation adjustment; they decline in real terms every year. The 2026 fee schedule includes a 0.5% reduction in the outpatient conversion factor. Specialists face additional cuts: audiologists -3%, speech-language pathologists -4%.
Medicaid is worse. The enhanced Federal Medical Assistance Percentage that expanded coverage during COVID expired on January 1, 2026. Medicaid pays providers 20 to 40 percent less than Medicare. For a dental rollup serving Medicaid patients, like Affordable Care in FSK’s portfolio, the reimbursement floor is dropping while the debt ceiling is rising.
The No Surprises Act, which eliminated balance billing for emergency services, cut physician-billed emergency revenue by 20.1% between 2019 and 2023. For companies like Global Medical Response, the largest ambulance company in America, $5.3 billion in debt, currently restructuring, the revenue that justified the original PE acquisition has been legislated away.
Meanwhile, healthcare labor costs increased $42.5 billion between 2021 and 2023. Contract labor surged 258%. Nurses and physicians command wages that are rising at 5.5% annually, more than double inflation. Labor represents 50 to 60 percent of provider operating costs.
Revenue is government-controlled and declining. Costs are market-driven and rising. The gap is the PIK rate. When a healthcare rollup can’t cover both debt service and rising labor costs, it defers the debt service. The PIK rate tells you how many companies have already made that choice.
59 healthcare holdings across 18 BDC funds. 10.1% of all healthcare holdings. The choice has been made. The conversions follow.
The pipeline is loaded. The conversions are coming.
What good looks like.
Not every fund is exposed. TCPC (BlackRock TCP Capital) has 10 healthcare holdings. Zero PIK. Zero non-accrual. The same “good underwriting” outlier I identified in Pebbles II. BXSL has 48 healthcare holdings with only 1 PIK. ORCC has 18 with zero PIK.
Good underwriting exists. But ARCC at 20.5% healthcare PIK, NMFC at 16.3%, and FSK at 22.2% are the rule. TCPC is the exception.
The question for investors.
When you buy ARCC shares for the 9% dividend yield, you are buying exposure to a portfolio where the same PE firm that owns the dental chain also manages the fund that holds the dental chain’s debt. The PE firm extracted $835 million in dividends from the dental chain while the dental chain’s leverage hit 7.4 times EBITDA. The PE firm earns management fees on your BDC shares regardless of whether the dental chain pays its interest or goes on PIK.
You are not the investor. You are the exit.
The PE sponsors already got paid. The $1.1 billion in dividends from Aspen Dental, the $1.3 billion Cerberus extracted from Steward, the 21x EBITDA that TSG paid for Thrive: that money left the building. What remains is the debt. And the debt is in your BDC. And your BDC is in your pension fund. And your pension fund doesn’t know it owns a dental chain that can’t pay interest.
I ran TonalityIQ on every BDC transcript I could find, 28 quarters for FSK, 28 for ARCC, 28 for TCPC. What I found changes the framing.
FSK has never had a confident earnings call. Not once. Not in 28 quarters. Not during zero rates in 2019. Not during the COVID recovery in 2021. Not in any quarter of any year. The ratio (conviction divided by constraint) has never exceeded 0.63. In most quarters, it sits below 0.25. The management team uses four to seven words of hedging for every one word of conviction, every single time they speak publicly.
TCPC, the fund with zero healthcare PIK and the best underwriting in the system, ranges from 0.45 to 1.31 over the same period. TCPC’s management speaks with two to three times more conviction than FSK’s management in every quarter. The language gap is persistent, structural, and it maps directly to portfolio quality. The fund whose CEO hedges hardest has a 24.4% PIK rate. The fund whose CEO speaks with relative confidence has a 2.3% PIK rate.
The language doesn’t predict the stress. The language IS the stress. FSK’s management has been operating a structurally fragile portfolio for seven years, and the language has reflected that fragility for seven years. The PIK rate is the symptom. The constraint score is the diagnosis.
ARCC is the more interesting case. ARCC’s ratio ranges from 0.21 to 0.92, a much wider swing than FSK. In Q2 FY2025, ARCC briefly hit 1.27, the first time above 1.0, a moment of genuine confidence. By Q4, it dropped back to 0.92. ARCC’s language moves. When it moves toward constraint, that signals something, because ARCC is not permanently stressed the way FSK is. The 16 healthcare PIK holdings and the Aspen Dental vertical integration are showing up in a fund whose management CAN be confident, and is choosing not to be.
Across the broader healthcare sector, the pattern holds. Healthcare has the highest constraint score of any sector, 8.59 out of 20. The lowest conviction-to-constraint ratio: 0.52. Management teams across the sector are hedging harder, qualifying more, and committing less than any other industry we track. This is Terminal Anxiety. The mirror image of Michael Saylor’s Terminal Confidence on quantum risk, except in healthcare, the CEOs know exactly what’s coming. They’re choosing their words very carefully. The language is deteriorating before the earnings do.
The catalyst timeline is accelerating. Since I published Pebbles, Blue Owl’s tech BDC saw 40.7% of investors ask for their money back. Apollo’s president went on television and called it “growing pains.” Blackstone injected $400 million into BCRED to prevent a run. The gating that was theoretical when I published Pebbles I is now confirmed across every major platform.
Q1 10-Q filings drop in May. That’s when the healthcare PIK holdings get their next quarterly mark. If AmeriVet or AthenaHealth or Global Medical Response converts from PIK to non-accrual, ARCC takes the markdown. And every pension fund holding ARCC takes it with them.
Jim Zelter says it’s “quite an easy conversation.” The 59 healthcare companies paying interest with IOUs suggest otherwise.
The FSK smoking gun.
I’ve been short FSK since Pebbles I. The healthcare data makes the case worse. And now the rating agencies agree.
On March 23, Moody’s downgraded FSK from Baa3 to Ba1. Junk. The largest publicly traded BDC managed by KKR lost its investment grade rating. FSK’s own investor presentation, dated March 24 (the day after the downgrade), contains the numbers that explain why.
PIK interest income: $55 million in Q4 2025 alone. That’s 16% of FSK’s total investment income paid in IOUs, not cash. NAV has eroded from $23.64 to $20.89 in twelve months, an 11.6% decline , while the fund paid $2.80/share in distributions. They’re paying dividends from a shrinking asset base. Net realized losses across FY2025: $596 million. Leverage has climbed from 1.04x to 1.22x, approaching the ceiling of their 1.00-1.25x target range.
And buried in the deck: FSK operates a $5.3 billion joint venture with the South Carolina Retirement Systems Group Trust. A state pension fund. Teachers, firefighters, public employees. Co-invested co-invested in the same portfolio that Moody’s just rated junk. Healthcare is FSK’s fourth-largest sector at 12.8% of the portfolio.
Affordable Care Inc. is a dental rollup focused on Medicaid and low-income patients. It has six tranches in FSK totaling $135.1 million. Every single one is on payment-in-kind. Not some. All six. A dental chain serving the patients with the least ability to pay private rates, funded entirely with deferred interest, held in the BDC with the highest overall PIK rate in the system.
Affordable Care cannot raise prices. Its patients are on Medicaid. Medicaid reimbursement is set by the state. The state is cutting. The debt is compounding. The interest is deferred. And the entire position sits in FSK, which already has 24.4% of its portfolio on PIK across all sectors.
This is not a company that’s going to grow its way out. This is a company that exists because a PE firm found a way to turn Medicaid reimbursement into leveraged debt issuance. The music stopped. The debt is still playing.
I’m not the only one who sees it. On April 2, FSK options volume surged to 6,070 contracts. 5,896 puts versus 174 calls, a 34-to-1 put/call ratio. Someone bought 2,910 put spreads expiring April 17 at the $15/$12.50 strikes. FSK trades at $10.16. That is not a hedge. That is a directional bet with size, placed by someone who parsed the same data I did and arrived at the same conclusion. The smart money is positioning before the Q1 filings.
The Mars contrast.
Not every healthcare rollup is leveraged to collapse. Mars Inc. bought VCA, the largest veterinary chain in North America, for $9.1 billion in 2017. Cash. No leverage. No PIK. No BDC debt. Mars makes M&Ms. They can afford to own a vet chain without loading it with 7 times EBITDA in debt.
The difference between VCA and AmeriVet is not the veterinary business. The business is the same: same patients, same procedures, same staffing challenges. The difference is the capital structure. Mars paid cash. The PE firm behind AmeriVet paid with debt. Mars absorbs rising labor costs from candy bar margins. AmeriVet absorbs them by deferring interest payments.
When you take your dog to a VCA clinic, the practice can reinvest in equipment, hire staff, maintain quality. When you take your dog to an AmeriVet clinic, a portion of the revenue services the debt that funded the acquisition, the dividend that went to the PE sponsor, and the management fee that went to the BDC manager. The quality gap isn’t theoretical. It’s structural. And it shows up in the PIK rate.
VCA: zero PIK. Zero non-accrual. Privately held by a company with $50 billion in revenue.
AmeriVet: PIK in two BDC funds. Cross-fund contagion node. Owned by a PE firm that needed other people’s money to buy it.
The healthcare rollup model doesn’t fail because healthcare is a bad business. It fails because PE turned a stable, cash-generating business into a leveraged speculation, and kept the fees while transferring the risk.
The question for ARCC shareholders.
In Pebbles II, I wrote that I was not shorting ARCC. Ares has the strongest credit selection track record among the mega-BDCs. That was true at the system level.
The healthcare data complicates it.
Bloomberg Intelligence’s senior BDC analyst wrote on March 30 that ARCC’s “credit performance remains stable, with non-accruals below historical averages.” He is right, at the portfolio level. But that is like saying the patient’s average body temperature is normal while their appendix is about to burst. The aggregate masks the sector. Nobody on the sell side is parsing the Schedule of Investments by healthcare. When you do, you find a 20.5% PIK rate, double the system-wide average, concentrated in exactly the sector where revenue is government-capped and costs are rising 5.5% annually. ARCC reports Q1 earnings on April 28. That filing will show whether the healthcare PIK pipeline has started converting.
ARCC has 78 healthcare holdings. 16 are on PIK, a 20.5% PIK rate, double the system average. 4 are in non-accrual, including Pathway Vet Alliance ($54.3 million) and Absolute Dental ($52.5 million). Ares Management simultaneously owns approximately 40% of Aspen Dental and manages ARCC. The PE firm extracted $1.1 billion in dividends from Aspen Dental since 2012 while the dental chain’s leverage hit 7.4 times EBITDA. Moody’s downgraded the outlook.
I’m not saying ARCC is badly run. I’m saying ARCC has a structural conflict. When the PE sponsor, the lender, and the fund manager are the same entity, the incentives don’t align with the BDC shareholders. The PE sponsor wants dividends. The lender wants interest payments. The fund manager wants AUM. The BDC shareholder wants all three to work. When they don’t, when the dental chain can’t pay interest after the dividend extraction, the PE sponsor has already been paid. The BDC shareholder has not.
I am short ARCC.
The pebbles landed in private credit first. The extraction is pulling them into healthcare. The same debt structures. The same cross-fund overlap. The same pension exposure. The same PIK-to-default pipeline. Different waiting room.
I’ll update this analysis after the Q1 filings.
The healthcare PIK pipeline is loaded. The conversions are coming. And ninety-three percent of the companies at the bottom of the credit stack are PE-owned.
How the pipeline converts.
I modeled three scenarios for how the healthcare PIK pipeline could play out over the next twelve to eighteen months.
Scenario one: orderly absorption. The 59 healthcare PIK holdings convert to non-accrual at the low end of the historical range, roughly 15% over 24 months. That’s 9 new healthcare defaults entering the system. Spread across 18 funds, the impact is manageable. NAV declines of 1 to 2 percent per fund from healthcare alone. ARCC absorbs its share. Nobody panics. The PE sponsors restructure quietly. This is what the fund managers expect.
Scenario two: clustered sector failure. Reimbursement cuts from Medicare and the No Surprises Act compress provider margins enough that multiple healthcare rollups hit PIK simultaneously. The dental chains, the vet rollups, and the behavioral health platforms, all facing the same labor inflation, the same rate pressure, the same PE leverage, default in the same quarter. Because the contagion map shows 43 healthcare companies in multiple funds, the markdowns are synchronized. ARCC takes its 16 PIK marks. GBDC and GSBD take their 22 shared healthcare company marks. FSK takes the Affordable Care mark. The healthcare PIK rate, currently 10.1%, jumps to 15-20%. BDC earnings guidance falls. Dividend coverage ratios tighten. BIZD drops 15-20%.
Scenario three: vertical integration crisis. The ARCC-specific risk. If Aspen Dental, the largest DSO in America, 40% owned by Ares, leveraged at 7.4x after the $835M dividend extraction, goes on PIK or requires restructuring, the loss hits Ares three times. As PE equity owner (equity marked down). As debt originator (loan marked down). As BDC manager (AUM declines, fee income declines, stock drops). The vertical integration that generated fees on the way up generates losses on the way down, three times through three doors. This scenario doesn’t require a macro event. It requires Aspen Dental’s free cash flow to fall below its debt service. At $90M annual FCF against 7.4x leverage, the margin for error is thin.
The payer paradox.
There is one more finding I haven’t shared yet. It might be the most unsettling.
I ran TonalityIQ on the health insurance companies, the payers. Cigna’s Q4 FY2025 transcript shows a conviction-to-constraint ratio of 4.46. That is Terminal Confidence. The same linguistic signature that Saylor showed on quantum risk. The same pattern that Apollo’s co-president showed before gating.
The payers are confident. The providers are stressed. This is not a coincidence. It is the mechanism.
Insurance companies are cutting reimbursement: 20.1% decline in physician-billed emergency revenue, tightening prior authorization requirements, increasing claim denials. The providers absorb the cuts because they have no negotiating leverage, especially the PE-owned ones, who are already leveraged 6 to 8 times EBITDA. The providers can’t walk away from the payer contracts because the debt service requires volume. So they accept lower rates, increase patient volume per provider, cut staff, defer maintenance, and go on PIK.
The payer’s confidence is the provider’s constraint. Cigna’s 4.46 ratio is the mirror image of healthcare’s 0.52 ratio. One side of the transaction is projecting Terminal Confidence because they’re winning the negotiation. The other side is showing Terminal Anxiety because they’re losing it.
This is the structural squeeze. Revenue compressed from above by payers and regulators. Costs inflated from below by labor markets. And in the middle, PE firms that structured the debt during a zero-rate world that no longer exists, collecting management fees on the way down.
The numbers.
582 healthcare holdings across 18 BDC funds. 294 unique companies. 59 on PIK. 8 in non-accrual. 43 in two or more funds. 22 shared between GBDC and GSBD alone. 16 healthcare PIK holdings in ARCC, the fund managed by the PE firm that owns the largest dental chain in America. 20.5% healthcare PIK rate in ARCC versus 10.1% system-wide. $1.1 billion extracted from Aspen Dental in dividends since 2012. $135.1 million in Affordable Care, a Medicaid dental chain, all six tranches on PIK in FSK. 93% of the most distressed healthcare companies are PE-owned. $20 billion in healthcare PE debt affected by defaults in the last three years. Zero healthcare PIK in TCPC.
Nobody parsed the BDC filings by sector before this. The fund managers report aggregate statistics. The healthcare analysts don’t read Schedules of Investments. The BDC analysts don’t know which holdings are dental chains versus software companies. The data was public. The cross-reference was not.
Until now.
The trades.
Disclosure: I hold short positions in FSK, OWL, APO, and ARCC through put options. I may initiate short positions in ARES, ACHC, BIZD, and related names. I am financially incentivized for these securities to decline. Positions can change at any time.
The question is not whether healthcare PE rollups are stressed. They are. The question is how to express the thesis with asymmetric risk: small cost if wrong, large payoff if right, with a catalyst you can see on the calendar.
Burry didn’t short housing stocks. He bought credit default swaps on the specific tranches that would crack first. Paulson didn’t short banks. He bought protection on the CDOs the banks held. The instrument matters more than the direction.
FSK puts remain the core position. Affordable Care (six tranches, $135.1 million, all on PIK) is the healthcare-specific reason this fund cracks before others. FSK’s overall PIK rate of 24.4% is the highest in the system. The healthcare concentration makes it worse. June and July expiry puts, purchased now, give you two months of runway before the Q1 10-Q filings arrive in May. If Affordable Care converts from PIK to non-accrual in the Q1 data, FSK’s NAV takes a $135 million hit on one name. The puts pay 3-5x. If it doesn’t convert, you lose the premium. Defined risk, defined catalyst, defined timeline.
The ARCC trade is the new addition. Not common stock puts; those fight a 9% dividend yield and a $28.7 billion NAV cushion. The precise expression is the ARCC-TCPC pair. Long TCPC, short ARCC. Same asset class. Same macro exposure. Different underwriting. TCPC has zero healthcare PIK. ARCC has 16 at a 20.5% PIK rate. If healthcare defaults accelerate, ARCC underperforms TCPC. You don’t need BDCs to crash. You just need the gap between good and bad healthcare underwriting to widen. The pair neutralizes market direction and isolates the thesis.
The reflexivity trade is ARES, the management company, not the BDC. If Aspen Dental restructures, Ares gets hit three times. As PE equity owner, the equity marks down. As debt originator, the loan marks down. As BDC manager, AUM declines and management fee income falls. ARES the stock drops. A lower stock price makes it harder to raise the next fund. Less capital means existing portfolio companies can’t refinance. More defaults. More AUM compression. More stock pressure. This is Soros’s reflexivity: the market structure amplifies the fundamental deterioration in a self-reinforcing loop. And it starts with one dental chain’s free cash flow falling below its debt service.
ACHC (Acadia Healthcare) is the publicly traded expression of behavioral health stress. PE heritage. $2.2 billion in debt. 2.8 times leverage. Discovery Behavioral Health was seized by Capital One two months ago. If the behavioral health sector follows the trajectory that dental and veterinary have already traced (PIK, restructuring, default) ACHC is where the equity market prices it.
The PE manager basket extends beyond ARES. I ran TonalityIQ on every publicly traded alternative asset manager. The results:
KKR has the lowest conviction-to-constraint ratio of any PE manager: 1.35. The most hedged language. The most defensive posture. They should be defensive. KKR backed Envision Healthcare ($9.9 billion bankruptcy), Global Medical Response ($5.3 billion restructuring), and provided a $2.3 billion unitranche to MB2 Dental. KKR is the PE firm with the most healthcare rollup debt exposure. Their language says they know it.
Blackstone has the highest absolute constraint score of the group, 5.17, alongside high conviction (10.26). Projecting confidence while hedging aggressively. They ended up owning the pieces of Envision post-bankruptcy. They injected $400 million into BCRED to prevent a run. The constraint-conviction split suggests more stress underneath the surface.
Brookfield’s TonalityIQ score, 10.86, remains the highest of any company in our 636-ticker universe. $900 billion in assets. More reassurance per word of constraint than any other public company. Terminal Confidence at institutional scale.
BIZD captures the full BDC system. One ticker, one trade. The healthcare PIK pipeline hits every fund in the index.
The catalyst timeline is mechanical, not speculative. Q1 10-Q filings drop in May. That is when the healthcare PIK holdings get their next quarterly mark. Affordable Care, AmeriVet, AthenaHealth, Global Medical Response, Modernizing Medicine, Specialty Dental Brands, all on PIK, all in multiple funds or concentrated in ARCC. If any of them converts from PIK to non-accrual in Q1, the markdown hits the fund NAV. If it hits two funds simultaneously, as the contagion map predicts, the markdown cascades through the BDC sector, through BIZD, through the pension funds holding ARCC shares.
The trades are positioned before the filings. The filings confirm or deny. That is the structure of every asymmetric trade: you pay a defined premium for the right to be proven right by data that arrives on a schedule.
I will re-parse every filing. I will rebuild the healthcare contagion map. And I will publish what I find.
The regulators are moving.
The FTC blocked Welsh Carson’s anesthesia rollup, the first time a regulator explicitly targeted the PE rollup model in healthcare. The complaint alleged that the strategy of acquiring competing anesthesia practices in the same market, consolidating them under one platform, and raising prices was anti-competitive by design. Welsh Carson settled. But the precedent was set.
FSOC’s March 25 guidance on nonbank financial company designations targets the same vertical integration structure I’ve described, where the same entity originates, manages, and values the same assets. The Ares/Aspen structure fits the template.
State attorneys general are investigating PE-owned healthcare providers. Congressional hearings have been held on Steward. The Private Equity Stakeholder Project, which tracks PE extraction from healthcare, reports that 93% of the most distressed healthcare companies are PE-owned. That number is now in the hands of policymakers.
The regulatory momentum is real. But regulation operates on a multi-year timeline. The PIK conversions operate on a quarterly timeline. The defaults will arrive before the rules change.
Let me be clear about what I’m saying and what I’m not.
These funds won’t go to zero. They don’t need to. The trade works on markdowns, not bankruptcies. A 10% NAV decline on ARCC’s $28.7 billion is $2.8 billion in value destruction. The puts are priced for a world where nothing goes wrong. Anything going wrong is enough. That’s the base case: a good asymmetric trade that pays 3-5x on healthcare markdowns over the next two to three quarters.
But there’s a scenario where this becomes something larger.
Ares extracted $1.1 billion from Aspen Dental. Ares manages ARCC, which holds 78 healthcare companies at a 20.5% PIK rate, the highest in the system. ARCC doesn’t hold Aspen directly, but the same firm that loaded Aspen with 7.4x leverage is the one deciding which healthcare loans go into the BDC that your pension fund owns. Ares sold ARCC shares to pension funds. When Aspen can’t pay, Ares keeps the dividends. The pension fund takes the loss. That is not a trade structure. That is a conflict of interest at scale. FSOC’s March 25 guidance on nonbank financial company designations targets exactly this vertical integration, where the same entity originates, manages, and values the same assets.
If the SEC or a state attorney general opens a formal investigation into the Ares/Aspen structure, and the FTC has already blocked Welsh Carson’s anesthesia rollup on the same logic, then the trade goes from a 3x put to a 10x event. Not because the fund collapses, but because the management company’s ability to raise the next fund evaporates. That is Soros reflexivity. The investigation triggers the fundraising freeze, the freeze triggers the portfolio stress, the stress validates the investigation. Self-reinforcing.
I don’t know if that happens. The base case doesn’t require it. But the option is priced as if it can’t.
$191 billion in healthcare PE deals were completed in 2025 alone, a record. 445 buyouts. 149 dental acquisitions, 95% of them add-ons by PE-owned platforms. The industry is still acquiring while the existing portfolio is defaulting. New debt is being structured while old debt is going on PIK. The machine doesn’t stop because the earlier vintages are failing. It stops when the BDC investors stop buying the debt. And the BDC investors haven’t stopped yet.
The pebbles landed in private credit. The extraction pulled them into healthcare. The pipeline runs through your dentist, your vet, your urgent care, your therapist. And the PE firms that bought them already got paid.
Apollo’s co-president called the defaults “pebbles.” Apollo’s president called the gating “growing pains” and a “skirmish on the sidelines.” Blackstone’s co-CIO said there’s a “big disconnect between the headlines and what we see in the portfolio.” Blue Owl’s investors asked for 41% of their money back and got 5%.
The language is getting smaller. The numbers are getting bigger. That’s the pattern TonalityIQ was built to detect, and it is happening in real time across the entire private credit complex.
The extraction never stops until the host can’t pay.
Eric Jackson
Founder, EventHorizonIQ | EMJ Captial | EMJX
Disclosure: I hold short positions in FSK, OWL, APO, and ARCC through put options. I may initiate short positions in ARES, ACHC, BIZD, and related names. I am financially incentivized for these securities to decline. Positions can change at any time.
Previous calls: Carvana at $15 before the 32x run, Opendoor at $0.73, IREN at $9, CIFR at $3. The Wall Street Journal, Bloomberg, and Business Insider have covered his previous calls.
See the live sensors behind this analysis:
eventhorizoniq.com/from-substack
Run TonalityIQ on any ticker, 588 companies, last 4 quarters free:
Track record: eventhorizoniq.com/scoreboard
This is not investment advice. This is structural analysis. Do your own work.