The Cockroach Theorem
How Private Credit’s $2 Trillion Zombie Equilibrium Will Produce the Lost Decade of Alternative Assets
By Shanaka Anslem Perera The Macro Intelligence Report | February 27, 2026
I. Two Prices, One Truth
On February 18, 2026, the same portfolio of private credit loans was assigned two prices simultaneously. The first price was 99.7 cents on the dollar, paid by a consortium of institutional buyers including CalPERS, OMERS, British Columbia Investment Management, and Kuvare, the Chicago-based insurance group whose asset management operations Blue Owl Capital acquired for $750 million in 2024. The second price arrived two days later when Saba Capital Management and Cox Capital Partners filed tender offers for shares in three Blue Owl non-traded vehicles at discounts of 20 to 35 percent below stated net asset value. In the language of the bond market, the same underlying credit was simultaneously money-good and deeply distressed, depending on whether you were asking the loans or asking the investors trapped in the wrapper around them.
This is not a contradiction. It is the defining condition of the 2026 private credit market.
The response from Blue Owl was instantaneous and revealing. One day after Saba’s initial notice, the firm permanently halted quarterly share redemptions from its $1.6 billion unlisted vehicle, Blue Owl Capital Corporation II, reversing its previous plan to reopen withdrawals. It announced $2.35 per share in planned distributions, roughly 30 percent of stated net asset value, by the end of March. It shifted to a return-of-capital model that effectively placed the fund into multi-year runoff. The gates went up. The narrative locked in. The loans are fine, the managers assured everyone. It is merely the liquidity that is conditional.
But conditional liquidity is a category error in an asset class that sold itself on the promise of unconditional yield. For a decade, the wealth management industry told individual investors that private credit offered the returns of high yield without the volatility of public markets, an alchemy enabled by the simple expedient of not marking assets to market. The bargain was explicit: you surrender daily liquidity and in return you receive smoothed returns, lower reported volatility, and yields that justified the opacity. What no one explained, because no one had an incentive to explain it, was that the smoothed returns were not a feature of the underlying credit quality. They were a feature of the wrapper. And the wrapper has now been priced by an outsider for the first time in the asset class’s history.
Blue Owl’s stock fell 55 percent from its 52-week high of $22.41 to $10.08 over a period of weeks, recording eleven consecutive days of losses, the longest streak since the company’s initial public offering. Short interest reached 12.5 percent of free float. Securities class action lawsuits from Robbins Geller, Pomerantz, Rosen Law Firm, and Faruqi and Faruqi landed in the Southern District of New York. Mohamed El-Erian called it a canary in the coal mine. Jamie Dimon, nine weeks earlier, had already provided the definitive metaphor.
“When you see one cockroach,” the JPMorgan Chase chief executive told analysts on his October 2025 earnings call, “there are probably more. Everyone should be forewarned on this one.”
The consensus is waiting for the cockroaches to trigger a Lehman moment, a sudden, catastrophic liquidation that would force the $2 trillion private credit market into a fire sale of assets and a cascade of defaults. The consensus will wait forever. Private credit will not explode. It will zombify. And the mechanism through which it zombifies is already operating in plain sight, visible to anyone willing to look at three numbers that do not reconcile and ask the question that the entire industry is financially incentivized to never ask: what happens when the price of not knowing finally exceeds the yield?
Inside this analysis: the forensic anatomy of the zombie equilibrium now forming across the private credit market, the mechanisms sustaining it, the specific timeline of catalysts that will force recognition, the $300 billion bank transmission channel that connects this opaque market to the regulated financial system, the adversarial gauntlet that stress-tested every link in the causal chain, the trade structure that expresses the thesis with defined risk, and the precise conditions under which this entire framework collapses and should be abandoned. The positions are already being built.
II. The Architecture of Zombification
The amateur’s model of private credit failure is simple: rates stay high, borrowers default, lenders lose money, funds mark down, investors redeem. This model is wrong in every particular that matters. It describes the mechanism by which public credit markets reprice. It has almost nothing to do with how private credit actually fails, because private credit was engineered from inception to prevent precisely this sequence from occurring.
The architects of modern private credit understood something profound about institutional investor psychology. Most allocators do not primarily fear losses. They fear volatility. A portfolio that declines 15 percent and recovers over three years produces the same terminal wealth as one that never declined, but the allocator who reports a 15 percent quarterly drawdown faces career risk that the smooth compounder does not. Private credit was designed to exploit this asymmetry. By holding loans in structures without daily marks, by retaining the authority to value their own portfolios using internal models rather than market prices, by wrapping fundamentally illiquid assets in semi-liquid structures with quarterly redemption gates, the industry created a product that delivered the one thing institutional investors value above returns: plausible stability.
This design choice has a consequence that the industry did not anticipate. When stress arrives, the system’s response is not to reprice. It is to prevent repricing. And prevention of repricing, sustained over time, produces the precise conditions that the economists Ricardo Caballero, Takeo Hoshi, and Anil Kashyap identified in their landmark 2008 American Economic Review paper on Japan’s lost decade: zombie lending.
In Japan, banks kept insolvent borrowers alive through what Caballero and his co-authors termed “sham loan restructurings,” rolling over principal, extending maturities, and reducing interest payments to levels that allowed technically dead companies to avoid the formal recognition of death. The banks did this not because they believed the borrowers would recover but because recognizing the losses would have destroyed their own capital ratios. The zombie firms survived. The healthy firms suffered. Zombie-dominated industries experienced depressed job creation, lower productivity growth, and suppressed entry by more competitive businesses. Japan’s Nikkei index declined 80 percent over thirteen years. The government eventually deployed the equivalent of $400 billion to recapitalize the banking system.
The structural parallel to 2026 American private credit is not approximate. It is mechanically exact. The role of the impaired banks is played by private credit funds and their insurance company affiliates. The “sham restructurings” are Payment-in-Kind amendments. The institutional incentive to suppress loss recognition is the management fee, typically 1 to 1.5 percent of assets under management, which grows as PIK capitalizes unpaid interest into principal, inflating the AUM base while masking borrower distress. The absence of a termination date in evergreen and perpetual capital structures, which now comprise roughly 40 percent of the five largest managers’ combined assets under management according to the Loan Syndications and Trading Association, eliminates the forcing function that closed-end funds once imposed. In Japan, regulators were lax. In American private credit, no single regulatory body exercises comprehensive jurisdiction over the entire chain.
Here is how the zombie equilibrium operates in practice. A mid-market software company, acquired by a private equity sponsor in 2021 at 14 times adjusted EBITDA with 6 times leverage, finds that its annual recurring revenue has decelerated from 25 percent growth to 8 percent as enterprise customers consolidate software licenses in response to AI-driven automation. Cash interest coverage, once comfortably above 2.0 times, has declined to 0.9 times as SOFR remains above 3.6 percent, well above the 1.5 percent at which the loan was originated. Under normal credit market discipline, this borrower would default, restructure, or be acquired at a discount.
But the private credit lender holding this loan operates under a fundamentally different set of incentives than a bank or a public bond fund. The lender earns management fees on the fair value of its portfolio. Recognizing a loss reduces the fee base. If the lender manages a semi-liquid vehicle, a loss recognition triggers redemptions from investors who were promised stability. If the lender is an insurance company subsidiary, a loss triggers higher risk-based capital charges from state regulators. The rational response, the dominant strategy in a game-theoretic sense, is to offer the borrower a PIK amendment: convert some or all of the cash interest to payment-in-kind, which capitalizes the unpaid amount into the loan principal. The borrower avoids default. The lender avoids a mark. The loan’s fair value remains unchanged on the books. The management fee is preserved. And the actual debt burden of the borrower increases by the exact amount of the unpaid interest, compounding the problem for the next quarter, and the quarter after that, in a spiral that only terminates when someone forces price discovery.
Lincoln International, whose Private Market Index covers more than 7,000 valuations and $250 billion in indexed enterprise value, provides the most rigorous independent measurement of this dynamic. In its fourth-quarter 2025 release, Lincoln reported that 11 percent of all evaluated loans carried some form of PIK interest. Critically, 58 percent of these instances were classified as “bad PIK,” defined as situations where the borrower originally had the capacity to pay cash interest but was subsequently forced to toggle because of deteriorating liquidity. The resulting shadow default rate of 6.4 percent has nearly tripled from 2.5 percent in the fourth quarter of 2021. It is important to be precise about what this number means. It does not mean 6.4 percent of private credit borrowers have defaulted. It means 6.4 percent of borrowers have crossed the threshold at which their lenders had to choose between recognizing distress and concealing it, and in every case they chose concealment.
The average loan-to-value ratio for the bad PIK cohort expanded from a healthy 39.4 percent at origination to a deeply distressed 76.1 percent by late 2025, according to Lincoln International, indicating that the compounding debt burden is rapidly approaching the total enterprise value of the underlying businesses. Concurrently, instances of lenders executing hostile, out-of-court takeovers of struggling borrowers from private equity sponsors surged to $24.1 billion in foreclosed debt in 2025, a figure that vastly exceeds the combined totals of the previous three years. The lender-to-owner pivot is the endgame of the zombie cycle: when the accumulated PIK destroys the equity cushion entirely, the lender has no choice but to seize the business and hope to operate it better than the sponsor did. This is not restructuring. This is the shadow banking system becoming an accidental operating conglomerate.
The fee structure alignment with zombie maintenance is not an inference. It is an observable incentive. Adams Street Partners, a prominent limited partner with over $53 billion in assets under management, stated explicitly in 2025 that publicly listed managers have “little economic incentive to self-regulate, as their fees, revenues, and stock prices are correlated to growing fee-earning assets.” A manager earning 1.25 percent on a $10 billion portfolio collects $125 million annually. If PIK accruals inflate the stated portfolio value by 8 percent, the manager earns an additional $10 million in fees from capital that was never generated by actual economic activity. Multiply this dynamic across the $2 trillion deployed market and the systemic fee extraction on phantom income runs into billions annually. No manager will voluntarily mark down an asset that generates fees while it decays, any more than a Japanese banker in 1995 would voluntarily write off a loan that still paid nominal interest. The incentive architecture is identical. The outcome will rhyme.
The concealment works because of a second mechanism the Japanese analog did not have: the absence of external price discovery. In Japan, zombie firms still had publicly traded debt and equity, providing at least a noisy signal of deterioration. In American private credit, 86 percent of direct lending assets are senior first-lien loans held by the originating lender, valued by the originating lender, using models chosen by the originating lender, with inputs selected by the originating lender. The same loan in the same syndicate can be valued at 79 cents by one lender and 46 cents by another, as the cybersecurity company Magenta Buyer demonstrated. Debt tied to KBS Fashion Group was simultaneously marked at 28 cents and 100 cents across different business development companies, even within vehicles managed by the same sponsor. These are not measurement errors. They are the system functioning as designed.
Agentic artificial intelligence introduces a third amplifier that Japan’s lost decade lacked entirely. UBS estimated in February 2026 that private credit defaults could reach 15 percent in a worst-case scenario of aggressive AI disruption, driven by the fact that software companies now represent 20 to 35 percent of direct lending portfolios. Software was the favored sector because its subscription revenues appeared maximally sticky. That stickiness assumed human users paying per-seat licenses. As enterprises deploy autonomous agents that can replace ten to fifteen human workers per deployment, the per-seat revenue model faces compression that the original underwriting never contemplated. The iShares Expanded Tech-Software ETF declined 24 percent in early 2026. Salesforce’s stock dropped 28 percent. Atlassian fell 35 percent. But these are public companies with the capital and strategic capacity to adapt, transitioning to consumption-based pricing models and investing in their own AI capabilities. Salesforce reported fourth-quarter fiscal 2026 revenue of $11.2 billion, up 12 percent year over year, its fastest growth in two years. The private credit borrowers, the mid-market software firms already leveraged at 6 times or more and toggling to PIK because they cannot cover cash interest, do not have the capital to invest in transformation. They are frozen in place, too leveraged to adapt and too valuable to their lenders’ fee streams to die. This is the zombie equilibrium in its purest form.
III. The Clock That Consensus Cannot Read
The market’s consensus timeline for private credit stress is measured in credit cycles: gradual deterioration, slow marks, eventual recovery. The actual timeline is measured in governance decisions and regulatory calendars, and it is substantially shorter than consensus expects.
The first catalyst has already fired. The Saba Capital tender offers of February 17 through 20 created the first external clearing price for non-traded private credit shares. This is an event that cannot be unobserved. Every investor in every semi-liquid private credit vehicle now possesses an anchor price suggesting their holdings may be worth 20 to 35 percent less than stated NAV. The behavioral finance literature is unambiguous about anchor effects: once a number enters the information set, it distorts all subsequent judgment. Blue Owl’s gating response amplified the signal by revealing that the promise of quarterly liquidity is, in practice, the promise of quarterly liquidity only when conditions are favorable.
The second catalyst arrives on March 22, 2026, when the National Association of Insurance Commissioners convenes its Spring National Meeting in Kansas City. On the agenda is the modified exposure draft of Proposal 2025-16-L, which would replace the current uniform 6.8 percent risk-based capital charge for collateral loans with a differentiated regime. Under the February 4, 2026 revision, loans backed by joint venture or limited partnership equity interests would carry a 24 percent charge, and loans backed by residual or non-senior tranches would carry 36 percent. These figures represent a quadrupling of capital costs for the most common structures through which insurance companies provide leverage to private credit vehicles. The comment period closes on March 12. Industry participants have lobbied to defer effectiveness to 2027, but regulators appear committed to December 2026 adoption. The practical effect is to raise the cost of the capital that has been the primary funding engine for private credit’s explosive growth over the past decade.
Across the Atlantic, the European Insurance and Occupational Pensions Authority launched a public consultation on February 3, 2026, targeting supervisory expectations for private equity ownership of insurance companies, with specific focus on the shift into private credit, the use of reinsurance structures, and the complexity of affiliated transactions. The consultation closes April 30. The Bank of England launched its second System-Wide Exploratory Scenario exercise on December 4, 2025, focused explicitly on how the private markets ecosystem behaves under stress. Sixteen of the largest alternative asset managers, including Apollo, Blackstone, Ares, KKR, and Oaktree, are participating. Results are expected in early 2027. These exercises do not themselves change capital requirements. What they change is the narrative frame. The next stress event in private credit will be interpreted through the lens of systemic risk rather than idiosyncratic failure, because regulators on three continents have told the world they are looking.
The third catalyst is structural rather than calendric. Approximately $500 billion to $600 billion of leveraged loans originated during the 2021 vintage, at peak valuations and historically low rates, face refinancing between 2026 and 2028. For borrowers in the hedge finance tier of the Minsky classification, those whose cash flows comfortably cover debt service, refinancing at current SOFR levels is painful but survivable. For borrowers already in the speculative tier, those meeting interest payments but unable to repay principal, refinancing requires either a substantial equity injection from sponsors or further amendment of loan terms. For borrowers in the Ponzi finance tier, those covering neither interest nor principal and surviving entirely on PIK, the refinancing wall is not a wall at all. It is a mirror, because the act of refinancing requires an independent assessment of value that the current system is designed to avoid.
The temporal arbitrage is precise: the market prices private credit risk as a slow-moving credit cycle with gradual marks unfolding over three to five years. The governance mechanism, the tender-to-gate-to-contagion sequence, operates on a timeline of weeks to months. The regulatory mechanism operates on a timeline of quarters. The refinancing mechanism operates on a timeline of one to two years. In every case, the actual timeline is shorter than the priced timeline. The mismatch between perceived timeline and actual timeline of realizable liquidity is itself an actionable signal.
IV. The Ghost Leverage Stack
To understand who is exposed to the zombie equilibrium, it is necessary to trace the full capital stack from borrower to end investor. No single participant in this chain sees the entire chain, which is itself the most dangerous feature of the architecture.
At the borrower level, the average U.S. leveraged buyout carries debt of 4.9 times EBITDA according to Bain and Company’s 2024 Global Private Equity Report, with S&P Global calculating adjusted leverage of 7.0 times when addbacks are stripped out. Twenty percent of companies in S&P’s coverage universe operate below 1.0 times interest coverage, meaning their operating income does not cover their interest expense before any principal repayment.
At the fund level, publicly traded business development companies operate within a statutory 2:1 debt-to-equity limit, expanded from 1:1 by the Small Business Credit Availability Act of 2018. The sector average across 46 BDCs is 1.11 times, with outliers approaching the ceiling.
At the NAV facility level, a market estimated at $100 billion to $150 billion by Partners Group and the Fund Finance Association, banks provide leverage against fund portfolios at 25 to 30 percent loan-to-value ratios. The growth rate of 30 percent annually since 2019, with projections to $600 billion by 2030, indicates the pace at which this leverage layer is expanding.
At the subscription line level, a market of $800 billion to $900 billion globally according to Cadwalader and Fitch, facilities are secured by uncalled limited partner commitments rather than fund assets. Goldman Sachs has pioneered the securitization of these facilities, issuing a $475 million asset-backed deal rated triple-A by Morningstar DBRS. Seventy-five percent of private equity funds use subscription facilities.
When these layers compound, the effective systemic leverage on a dollar of mid-market EBITDA reaches 7.6 to 12 times under conservative to aggressive assumptions. But no single regulator, no single investor, and no single risk committee sees this full chain, because each layer was designed by different institutions under different regulatory frameworks for different constituencies.
The insurance-private equity nexus is where the leverage chain achieves its most concentrated and politically consequential expression. Apollo’s Athene subsidiary, with total assets of approximately $430 billion as of September 2025, represents roughly 44 percent of Apollo’s total assets under management. Athene can represent up to two-thirds of the capital alongside third-party investors in Apollo-managed credit products, according to Apollo’s Global Head of Product. In January 2026, Athene agreed to acquire the approximately $9 billion commercial real estate loan portfolio of Apollo Commercial Real Estate Finance at 99.7 percent of total commitments, a direct affiliated transaction. NAIC statutory filings show approximately 18 percent of Athene’s portfolio classified as related-party assets under SSAP No. 25, a figure that grew by $40.1 billion in a single year and represented roughly 30 percent of all new industry-wide affiliated asset growth.
The International Monetary Fund’s October 2025 Global Financial Stability Report found that private credit accounts for approximately 35 percent of investment portfolios of North American insurance companies, that PE-backed insurers hold almost twice as much in illiquid assets as their non-PE-owned counterparts, and that within one year of PE acquisition, insurers decrease corporate bonds by more than 7 percentage points while increasing private-label asset-backed securities by more than 6 points. The IMF specifically warned that specialized rating agencies rated roughly 7,000 securities in 2023 versus 2,000 in 2019, a fourfold increase, while traditional rating agencies rated approximately the same 1,000 annually. The Bank for International Settlements noted that these smaller agencies “may face commercial incentives” that compromise independence.
The banking system’s exposure to private credit operates through the fund finance market, estimated by Preqin at roughly $1 trillion. Moody’s reported in 2025 that U.S. banks hold approximately $300 billion in direct lending to private credit providers, within $1.2 trillion of total loans to non-depository financial institutions. NDFI loans now represent 10.4 percent of total bank loans, nearly triple the 3.6 percent share of a decade ago. Wells Fargo maintains nearly $60 billion in direct exposure, Bank of America over $33 billion, PNC nearly $30 billion.
When Tricolor Holdings filed for Chapter 7 liquidation in September 2025 following criminal indictments alleging systematic double-pledging of auto loan collateral and manipulation of delinquency data, the transmission to the banking system was immediate. JPMorgan Chase absorbed a $170 million charge-off. Fifth Third Bancorp disclosed an impairment approaching $200 million. Zions Bancorporation took a $50 million loss that sent its stock down 13 percent in a single session. For JPMorgan, $170 million was a rounding error. For Zions, it was a material event. The First Brands Group bankruptcy, in which a January 2026 DOJ indictment details alleged fraudulent liabilities exceeding $2.3 billion alongside total debt of over $6 billion including opaque off-balance-sheet obligations, demonstrated that the verification infrastructure of private credit cannot scale with the asset class. When a lender originates a $300 million loan backed by granular receivables across thousands of counterparties, the cost of independently verifying each receivable exceeds the economic benefit. The system necessarily relies on borrower-provided data. When that data is alleged to have been systematically compromised, as prosecutors claim in the First Brands and Tricolor cases, the collateral foundation evaporates instantaneously.
The Blue Owl founder share pledge crystallizes the reflexivity risk embedded at the top of the leverage stack. Co-CEOs Doug Ostrover and Marc Lipschultz pledged more than 106 million units, roughly 6.6 percent of outstanding shares, as collateral for personal loans. These units were valued at approximately $1.9 billion at the time of pledge. With OWL’s stock having declined 55 percent, the collateral has compressed well past typical starting loan-to-value thresholds. SEC filings warn that shares “could decline materially” if founders face forced selling. The executives used the proceeds partly to acquire a majority stake in the NHL’s Tampa Bay Lightning at a valuation of $1.8 billion. This is leverage on leverage on leverage: the founders borrowed against their stakes in a company that earns management fees on funds that are themselves leveraged and hold loans to leveraged borrowers. The arithmetic at each layer may be individually sound. The cascade, should any layer break, is not.
The precedent for what happens when semi-liquid vehicles enter sustained gating is not theoretical. Blackstone Real Estate Income Trust, BREIT, imposed redemption limits in late 2022 after investor requests exceeded its 5 percent quarterly cap. The fund required a $4.5 billion equity injection from the University of California system at guaranteed returns of 11.25 percent to stabilize outflows. Independent analysis by Chilton Capital estimated BREIT’s true NAV at $9.17 per share, some 33 percent below the stated $13.85. GAAP equity per share was just $6.16. Starwood’s SREIT has been gated for twenty-nine consecutive months, with NAV down 40 percent and fulfilling only 4 percent of redemption requests. BREIT eventually resumed full redemptions in March 2024, which the industry cites as proof that gates work. What this characterization omits is that stabilization required a multi-billion dollar bailout at above-market terms from a single counterparty with a long time horizon and no mark-to-market requirement. The question for private credit vehicles is whether such a counterparty exists at the scale required. For OBDC II, the answer was no, which is why the fund was placed into terminal runoff.
The contagion mechanism between gated vehicles is documented and immediate. BREIT’s gating in 2022 triggered a surge in redemption requests at SREIT and other non-traded REITs, as investors who observed one fund restrict access preemptively queued for exit at competing vehicles. The same dynamic is now operating in private credit. Fourth-quarter 2025 saw $2.9 billion in withdrawals from private credit BDCs with more than $1 billion in assets, a 200 percent increase from the third quarter, according to Robert A. Stanger and Company. Blue Owl’s OTIC experienced redemption requests of approximately 15 percent of net asset value, triple the 5 percent quarterly cap. The secondary market for private credit fund stakes doubled from $6 billion in 2023 to $11 billion in 2024, with Evercore projecting approximately $18 billion in 2025, but this remains a fraction of the $2 trillion deployed market.
Goldman Sachs Prime Brokerage reported that in the week through February 19, 2026, hedge fund net sales of global stocks reached negative 1.4 standard deviations from normal, the fastest selling pace since the April 2025 tariff selloff. Seven of eleven global sectors were net sold, with financials experiencing the largest disposals. This is the transmission mechanism from private credit governance stress into public equity markets: as alternative asset managers’ stocks decline on gating concerns, their positions in other asset classes are liquidated by systematic and discretionary funds that use market-wide risk-off signals as sell triggers.
The retail exposure dimension adds a final, politically volatile element. Executive Order 14330, signed by President Trump on August 7, 2025, directs the Department of Labor to facilitate access to alternative assets including private credit within 401(k) plans. The DOL submitted a proposed rule to the Office of Management and Budget on January 13, 2026. If finalized, the order would potentially channel a portion of the $12 trillion in defined-contribution retirement savings into an asset class whose valuation practices are under simultaneous scrutiny by the SEC, the DOJ, and insurance regulators in all fifty states. The timing is remarkable: retail capital would enter the market at precisely the moment when the gap between stated value and realizable value is at its measured maximum.
V. The Forensic Record
The strongest evidence for the zombie equilibrium thesis comes not from projection but from the observable bifurcation already visible in the fourth-quarter 2025 earnings of publicly traded BDCs, the one segment of private credit that must report under SEC rules.
Main Street Capital, the industry’s gold standard for discipline, reported net asset value growth to $33.29 per share, a 17 percent return on equity, and non-accruals of just 1.0 percent at fair value. It declared a regular monthly dividend of $0.26 plus a $0.30 supplemental in the first quarter of 2026. Its PIK income runs below 4 percent of total investment income. Main Street demonstrates that conservative underwriting, portfolio diversification, and operational rigor produce genuinely strong performance. This is not an indictment of private credit as an asset class. It is proof that the asset class contains a distribution, and the distribution has tails.
At the other end of that distribution, FS KKR Capital reported a fourth-quarter net realized and unrealized loss of $0.89 per share, non-accruals of 3.4 percent at fair value, net asset value erosion of 11.6 percent year over year to $20.89, and a forced dividend cut of 31 percent, from $0.70 to $0.48, announced on February 19. The company’s net debt-to-equity ratio surged to 122 percent. Full-year 2025 realized and unrealized losses totaled $2.30 per share. When a fund with nearly $15 billion in total assets announces its largest dividend cut in the company’s history while reporting losses exceeding the annual distribution, the word distress is not hyperbole. It is accounting.
Between these two poles, the speculative tier illustrates the mechanism of gradual migration. Blackstone Secured Lending reported solid fourth-quarter results with net investment income of $0.80 per share covering its $0.77 distribution, but carries PIK of approximately 8.2 percent of investment income and faces concentrated downside risk in software holdings. Ares Capital, the largest publicly traded BDC with $27 billion in portfolio assets, presents the most consequential ambiguity. ARCC reported stable core earnings of $0.50 per share in the fourth quarter and maintained its $0.48 dividend. Its PIK income runs at approximately 15.5 percent of total investment income, the highest among major BDCs. Ares management insists that this PIK is substantially structural, arising from preferred equity positions in large-cap companies averaging $480 million in EBITDA, where PIK was written into the original terms rather than representing a distressed toggle. The company carries $988 million in undistributed spillover income, providing more than two quarters of dividend coverage.
The difficulty of adjudicating the Ares case illustrates the core epistemological problem of private credit. If Ares is correct that its PIK is intentional and its spillover provides genuine resilience, then the highest-PIK entity in the public BDC universe is also one of the best-managed. If Ares is incorrect, and the distinction between structural PIK and stress-driven PIK blurs as borrower conditions deteriorate, then the industry’s largest and most respected participant is carrying the greatest latent risk. No external observer can resolve this question with public data alone, which is precisely the condition that sustains the gap between reported value and realizable value. The answer matters because Ares is the index. Where Ares goes, the market’s perception of private credit follows.
The Proskauer Default Index provides the official industry metric, reporting a private credit default rate of 2.46 percent for the fourth quarter of 2025. The Fitch Ratings broader measure, which includes distressed exchanges, PIK amendments, and other liability management exercises, pegs the rate at 5.7 percent. Lincoln International’s shadow default rate, capturing all borrowers forced to toggle to bad PIK, reaches 6.4 percent. The choice of which number to cite is itself a statement about which reality one acknowledges.
The academic foundation for the zombie mechanism is now well established across multiple peer-reviewed literatures. Caballero, Hoshi, and Kashyap documented zombie ratios exceeding 25 percent in Japan’s post-bubble economy. Acharya, Crosignani, Eisert, and Eufinger, writing in the Journal of Finance in 2024, demonstrated that cheap zombie credit in Europe produced a disinflationary effect, with the published estimate suggesting approximately 0.2 percentage points of suppressed inflation under their bank recapitalization counterfactual. In a companion NBER working paper, Acharya, Lenzu, and Wang document the congestion externality, finding that healthy firm investment and employment decline by 3 to 11 percentage points in industries dominated by zombie firms. Neither paper has been formally applied to U.S. private credit. The extrapolation is this author’s inference, and should be understood as such.
The strongest counterargument to the zombie thesis is that the top tier of the industry is performing well, and it is. The Cliffwater Direct Lending Index shows annualized credit losses of just 0.88 percent since 2010 and only one negative year in its history. Blackstone’s BCRED reports a realized loss rate below 10 basis points over two decades. The June 2025 Federal Reserve stress test explicitly concluded that private credit and hedge funds “could endure crippling losses or even fail without threatening the biggest and most complex banks.” J.P. Morgan’s own analysis notes that private credit at roughly $1.2 trillion represents “just 9 percent of all corporate borrowing,” arguing it is not large enough to threaten the broader economy. These are significant findings and should not be dismissed.
The structural differences from the 2008 crisis are genuine and material. Eighty-six percent of direct lending assets are senior first-lien, providing meaningful recovery in default scenarios, compared to the multi-tranche CDOs of the pre-crisis era. Equity contributions in U.S. leveraged buyouts have averaged 42 to 47 percent since 2015, compared to 30 to 32 percent in 2005 to 2007. Floating-rate structures eliminate the interest rate mismatch that devastated fixed-rate mortgage-backed securities. Direct lender relationships enable real-time workout intervention that was impossible with diffusely held CDOs. These differences are not rhetorical. They are structural, and they explain why the zombie thesis predicts slow-motion value destruction rather than sudden catastrophic liquidation. Anyone waiting for a Lehman moment in private credit fundamentally misunderstands the architecture.
Interest rate relief provides partial but insufficient salvation. KBRA’s rigorous analysis of 1,067 private credit borrowers found that a base rate of 3 percent would generate approximately $9 billion in annual interest savings and improve coverage ratios for roughly 70 percent of borrowers with positive EBITDA. But 30 percent of companies with near-term maturities carry leverage above 10 times or have negative EBITDA, placing them beyond the reach of monetary policy. More subtly, research from Acharya, Lenzu, and Wang demonstrates that lower rates reduce the incentive to restructure, paradoxically extending zombie survival by making forbearance cheaper. Rate cuts address cyclical stress while perpetuating structural zombification. This is not a contradiction. It is the mechanism.
What we do not yet know, and what this analysis must honestly acknowledge, is the true scale of the leverage stack at the LP level, where pension funds and insurance companies may borrow against their private credit commitments in ways that are not documented in any public source. We do not know the precise vintage decomposition of PIK at the entity level. We do not know what the Bank of England’s stress test will reveal when it reports in 2027. These gaps are not analytical failures. They are features of a system that was designed to be opaque, and they represent the carrying cost of accumulated uncertainty that compounds with each quarter of deferred recognition.
VI. The Trade
This section describes how an institutional allocator could express the zombie equilibrium thesis with defined risk. It is analytical description, not personal financial advice.
The cleanest expression is relative value that isolates the wrapper layer from the loan layer, because the thesis predicts that governance and liquidity repricing will precede fundamental credit deterioration. The structure wants long exposure to managers demonstrating genuine cash-flow generation and short exposure to managers whose reported returns depend on PIK accruals and valuation discretion.
The long side concentrates in Main Street Capital, Trinity Capital, and Gladstone Investment, BDCs in the hedge finance tier whose cash interest coverage exceeds 1.5 times, whose PIK income runs below 5 percent of total investment income, whose non-accruals sit at or below 1.0 percent, and whose dividends have been raised rather than cut. These entities demonstrate that disciplined private lending produces real returns, and they benefit from the dispersion the thesis predicts, as capital migrates from distressed managers to quality.
The short side targets FS KKR Capital, which under the Minsky framework exhibits Ponzi finance characteristics within the public BDC universe, and Blue Owl Capital at the holding-company level, where the stock decline, the founder share pledge, the OBDC II gating, and the 55 percent drawdown from the 52-week high create a reflexive vulnerability that the company’s fundamentals must overcome. The thesis does not require Blue Owl’s loans to be impaired. It requires only that the governance events produce further redemption pressure, which triggers further gates, which validates the Saba discount, which invites further activist tender offers, in a sequence that can accelerate independently of the underlying credit performance.
Hedges are essential. A generalized recession would overwhelm the relative-value structure by impairing even quality BDCs, so the short leg should be paired with long positions in liquid investment-grade credit indices that benefit from flight to quality. Gold, whose 2026 advance past $5,000 per ounce reflects the broader dynamic in which cryptographically scarce assets outperform those requiring trust in institutional reporting, provides a portfolio-level hedge against the tail scenario where zombie equilibrium transitions to systemic stress.
Sizing should treat this as event-driven convexity rather than directional beta. Initial positions at 3 to 5 percent of portfolio, scaling only on confirmation signals: a second wave of tender offers in adjacent non-traded vehicles, additional SEC-filed repurchase terminations, or evidence that asset sales can no longer clear near par at meaningful scale.
The tripwires for addition are specific. If Saba or other activists launch tenders for Blackstone’s BCRED, the contagion has crossed from the mid-tier to the top tier, and position size should increase. If NAIC adopts the modified 2025-16-L proposal at the December 2026 meeting with the 24 and 36 percent RBC charges intact, the insurance capital arbitrage closes permanently, and the long-term structural short strengthens.
The kill switch is equally specific. If Blue Owl successfully reopens redemptions in OBDC II within 2026 without meaningful NAV concessions, if the $1.4 billion asset sale is followed by additional sales at near-par prices in size, and if the Lincoln shadow default rate stabilizes or declines below 5 percent in the next two quarters, then the governance stress was genuinely idiosyncratic and the contagion thesis should be abandoned with losses taken.
VII. The Lost Decade Has Already Begun
The amateurs are waiting for a crash. The professionals should be preparing for something worse: a decade of returns that are mathematically incapable of justifying the allocation.
The zombie equilibrium is not a failure state in the engineering sense. It is a stable configuration in which every participant behaves rationally given their incentives, and the aggregate outcome is value destruction. The fund manager who toggles a struggling borrower to PIK preserves their management fee and avoids a NAV markdown that would trigger redemptions. The insurance company that holds affiliated assets at marks set by the affiliated originator avoids the capital charge increase that honest marking would require. The pension fund that reports private credit returns of 9 percent avoids the painful conversation with trustees about whether the actuarial target of 7 percent is achievable with honest marks. The regulator who defers to 2027 avoids the political crisis of forcing recognition during an election cycle. Each decision is locally rational. The system equilibrium is globally catastrophic.
The macro consequence is a slow-motion capital trap. Roughly $2 trillion in institutional capital is committed to an asset class in which a growing share of the borrowers cannot service their obligations from cash flow. That capital cannot be reallocated to productive investment because it is locked in evergreen structures with no termination date. The mid-market economy, which private credit was supposed to revitalize after banks retreated post-2008, is instead being sustained in a state of artificial survival in which zombie firms compete for labor, materials, and customers against healthy firms, depressing margins, reducing investment, and suppressing the productivity growth that is the only genuine source of long-term returns.
The IMF’s February 2026 working paper, “Banking on Nonbanks,” adds a dimension the industry has not yet processed. The authors demonstrate that when macroprudential policies tighten bank lending, banking groups offset more than half of the contraction by redirecting credit through affiliated non-bank subsidiaries. The system does not delever. It reorganizes. Capital migrates from the regulated perimeter to the unregulated perimeter, carrying the same risk with less oversight. Private credit’s explosive growth since 2015 is not merely a story of yield-seeking. It is a story of regulatory arbitrage in which the post-2008 banking reforms did not reduce systemic credit risk but displaced it to a corner of the financial system where the tools of measurement are weaker and the incentives for opacity are stronger. The cockroach theorem is, in the final analysis, a theorem about measurement. A system that cannot accurately measure its own liabilities will eventually be measured by someone else, and the measurement will arrive at the worst possible time.
The comparison to Japan is not metaphor. Japan’s banking sector entered zombie lending in the early 1990s and did not emerge until the early 2000s, after a complete restructuring of the banking system and the deployment of hundreds of billions of dollars in public capital. Europe’s non-performing loan crisis, sustained by regulatory forbearance from 2010 to 2020, produced a decade of growth that never exceeded the pre-crisis trend. In both cases, the diagnosis was the same: the system chose stability over truth, and the cost of that choice was a generation of foregone growth.
What would change this assessment? Three developments would force re-evaluation. A genuine washout, in which 25 percent or more of private credit funds experience forced liquidation with losses exceeding 30 percent, would indicate that the system has chosen crash over zombie and the short thesis should be covered. A technological rescue, in which AI-driven productivity gains expand borrower EBITDA margins by 30 percent or more within 24 months, would provide the organic cash-flow improvement that makes PIK conversion to cash interest viable for the 2021 vintage. Or a policy intervention, in which the Federal Reserve creates a facility equivalent to the Bank Term Funding Program for private credit assets, would socialize the losses and eliminate the governance stress at the cost of moral hazard that makes the next cycle worse.
In the absence of these developments, the base case, assigned a probability of 45 percent, is the zombie equilibrium: gradual recognition, permanent gates, rolling PIK conversions, declining real returns masked by nominal stability, and a lost decade for the mid-market economy that will be visible only in retrospect. The bear case, at 25 percent probability, is that the governance stress triggers a reflexive spiral as gates breed more gates, bank fund-finance covenants tighten, and the system transitions from orderly zombification to disorderly liquidation. The bull case, at 30 percent probability, is that rate cuts to SOFR of 2.5 percent by 2027 combined with a productivity surge provide enough relief to allow the worst vintages to refinance and the gates to reopen without permanent impairment.
The framework that survives regardless of which scenario materializes is the distinction between the loan price, the NAV price, and the liquidity price. For a decade, the private credit industry maintained a monopoly over all three by holding assets in structures without external marks. Saba Capital broke that monopoly in February 2026 by printing a liquidity price that differed from the NAV price by 20 to 35 percent while Blue Owl simultaneously printed a loan price of 99.7 cents. The three prices can coexist because they measure different things: the loan price measures the expected recovery on the underlying credit, the NAV price measures the manager’s assessment of portfolio value, and the liquidity price measures what an investor can actually realize if they need to exit now.
The alpha is in understanding that these three prices were never the same. The industry merely pretended they were. The cockroaches did not arrive in February 2026. They were always there. Someone simply turned on the lights.
The cockroaches do not die. They outlive the cycle. They feed on the trapped capital of the next generation. And the question the institutional investor must now answer is not whether the system crashes. It is whether the world can afford the lost decade that the zombie equilibrium has already begun to produce.
Shanaka Anslem Perera is an independent researcher and analyst delivering institutional-tier analysis across monetary theory, geopolitics, sovereign debt dynamics, and macroeconomics to central banks, sovereign wealth funds, and institutional allocators worldwide.
DISCLOSURE
This document is published for informational and educational purposes only and does not constitute investment advice, a solicitation, or a recommendation to buy, sell, or hold any security or financial instrument. The analysis reflects the author’s independent research and opinions as of the publication date and is subject to change without notice. All investments involve risk, including possible loss of principal. Past performance is not indicative of future results. The specific securities and strategies discussed are presented for analytical purposes and should not be construed as personalized recommendations. Readers should consult qualified financial, legal, and tax advisors before making investment decisions. The author may hold positions in securities discussed herein and may trade in and out of such positions without notice. Where allegations of fraud or misconduct are referenced, characterizations reflect publicly available legal filings including DOJ indictments and SEC enforcement actions and do not constitute independent determinations of guilt or liability. All persons referenced in connection with pending legal proceedings are presumed innocent unless and until proven guilty in a court of law. Sources include SEC filings, Federal Reserve publications, IMF working papers, earnings transcripts, regulatory proposals, and other publicly available materials. While every effort has been made to ensure accuracy, the author does not warrant the completeness or reliability of third-party data. This analysis is provided on an “as-is” basis. The author, publisher, and affiliated entities accept no liability for any loss or damage arising from reliance on the information presented.
© 2026 Shanaka Anslem Perera. All rights reserved.
Amazing! Great stuff always! Private Credit is Shrodingers Cat. It's in a constant state of Superposition. It is neither alive nor dead because it does not allow any of us to Observe It!
excellent