The Liquidity Illusion: Private Debt Redemption Gates Through the Lens of Diamond-Dybvig

By The Good Guys Company

Executive Summary

In Q1 2026, over $4.6 billion in investor capital became trapped behind withdrawal limits across major private credit funds. Apollo, Ares, Blue Owl, Morgan Stanley, and Cliffwater have all capped or restructured redemptions. The U.S. private credit default rate has climbed to 5.8%, and redemptions as a share of NAV nearly tripled quarter-over-quarter in late 2025. What began as isolated concern around software loan markdowns has become the asset class's first systemic liquidity test at scale.

This paper argues that the current private credit redemption crisis is not merely a market dislocation. It is a structural re-enactment of the dynamics first formalised by Douglas Diamond and Philip Dybvig in their 1983 paper on bank runs. The parallels are precise and instructive. In both cases, an intermediary transforms illiquid long-duration assets into short-duration liquid claims, creating inherent fragility that depends entirely on confidence to remain stable.

The paper draws three sets of conclusions. First, it maps the Diamond-Dybvig framework onto the architecture of semi-liquid private credit vehicles and identifies where the analogy holds and where it breaks. Second, it assesses the impact on investors, distinguishing between those locked in current structures and those positioned to deploy capital opportunistically. Third, it argues that the crisis reveals a deeper structural truth: that the operating model of the fund manager is the primary mechanism through which liquidity risk, valuation integrity, and investor confidence are either maintained or destroyed.

1. The Diamond-Dybvig Framework

The 1983 Diamond-Dybvig model, for which Douglas Diamond received the Nobel Prize in Economics in 2022, describes the fundamental instability inherent in any institution that transforms illiquid assets into liquid liabilities. The model is elegant in its construction: an economy with three periods, a long-term productive asset, and agents with uncertain liquidity needs.

1.1 The Model's Core Mechanics

In the Diamond-Dybvig setup, investors deposit funds at t=0 into a bank. The bank invests in a long-term asset that yields a superior return at t=2 but can only be liquidated at t=1 at par or below. Some depositors will need funds at t=1 ("early types"), while others can wait until t=2 ("late types"). Individual depositors do not know in advance which type they will be.

The bank creates value by pooling deposits and offering a demand deposit contract that pays more than liquidation value to early withdrawers (r1 > 1) while still paying a return to late withdrawers (r2), financed by the surplus from long-term asset returns. Under normal conditions, only genuinely early-type depositors withdraw, the bank holds sufficient liquidity, and the arrangement is welfare-enhancing for all participants.

1.2 The Two Equilibria

The model produces two Nash equilibria. In the first, confidence holds: depositors who do not need immediate liquidity remain invested, the bank meets its obligations, and long-term assets mature at full value. In the second, confidence fails: depositors who would otherwise wait rush to withdraw because they believe others will do the same, forcing the bank to liquidate long-term assets at a discount, confirming the fears that triggered the run.

The run equilibrium does not require any deterioration in the underlying asset quality. It requires only a shift in expectations about other depositors' behaviour. The crisis is manufactured by the collective response to the fear of crisis.

1.3 Prescribed Remedies

Diamond and Dybvig identify two primary mechanisms to prevent runs. The first is suspension of convertibility: the bank halts withdrawals once a threshold is reached, removing the incentive to rush for the exit. The second is deposit insurance backed by a credible guarantor (typically the government or central bank), which eliminates the rational basis for panic by assuring depositors they will be made whole regardless of other depositors' actions.

Both solutions share a common logic: they break the strategic complementarity that makes withdrawal decisions interdependent. When each depositor's optimal action no longer depends on what others do, the run equilibrium dissolves.

2. The Private Credit Architecture

To apply the Diamond-Dybvig framework to private credit, it is necessary to understand the structural architecture of the vehicles now under stress. The private credit market has grown to approximately $2 trillion in deployed capital globally, with semi-liquid retail vehicles (non-traded BDCs, interval funds, evergreen structures) accounting for over $534 billion in AUM by year-end 2025.

2.1 The Maturity Transformation

The structural parallel to banking is direct. Private credit funds originate and hold loans to middle-market companies, typically floating-rate instruments with 3–7 year maturities, limited secondary market liquidity, and no continuous pricing mechanism. These are the illiquid long-duration assets of the Diamond-Dybvig model.

Against these assets, fund managers have issued semi-liquid liabilities: quarterly tender offers, periodic redemption windows, and NAV-based pricing that create the perception of liquidity for investors. The implicit contract is that investors can participate in private credit's yield premium without fully bearing its illiquidity cost. This is maturity transformation in its purest form.

2.2 The Irony of the Structure

Private credit originally emerged as the correction to the very asset-liability mismatch it now replicates. After 2008, as banks retreated from middle-market lending under regulatory pressure (Volcker Rule, Basel III/IV), alternative managers stepped in with a structurally sound proposition: raise committed capital from institutional investors who understood and accepted illiquidity in exchange for a yield premium.

As the asset class scaled, however, managers needed new pools of capital and found them in the retail and wealth channels. The mismatch migrated from bank balance sheets funded by deposits to fund structures supported by capital that expects periodic liquidity the underlying assets cannot reliably provide. The intermediary changed. The structural vulnerability did not.

3. The Run: Q4 2025 to Q1 2026

The sequence of events from late 2025 through Q1 2026 follows the Diamond-Dybvig dynamic with remarkable fidelity. What makes this episode instructive is that it began with a combination of fundamental credit concerns and expectation-driven contagion, making it simultaneously a solvency concern and a self-fulfilling liquidity crisis.

3.1 The Trigger Events

The initial catalyst was fundamental: the bankruptcies of First Brands and Tricolor in the U.S. auto sector in late 2025. While these failures were tied to asset-based finance rather than traditional direct lending, they forced a broader reassessment of loan quality across private credit portfolios. Simultaneously, software-as-a-service (SaaS) exposure, estimated at 20–30% of many private credit portfolios, came under pressure as agentic AI threatened the enterprise value of borrowers whose leverage had been underwritten during the low-rate era.

By early 2026, Morningstar DBRS found that credit ratings downgrades in private credit outnumbered upgrades by 3.3 times. The IMF had previously estimated that 40% of private credit borrowers carried negative free cash flow. The Morningstar LSTA US Leveraged Loan Index software segment lost 7.9% from its January peak through early March 2026.

3.2 The Contagion Sequence

The cascade from fundamental concern to systemic run proceeded through a recognisable pattern:

• Blue Owl attempted to merge its unlisted BDC (OBDC II) with its public counterpart in autumn 2025. The merger failed, exposing NAV/price mismatches and triggering redemption requests exceeding 15% of net assets from one tech-focused fund.

• In February 2026, Blue Owl halted quarterly tender offers for OBDC II entirely, replacing them with involuntary return-of-capital distributions. This converted a semi-liquid vehicle into an effective drawdown structure, unilaterally removing investor exit rights.

• In Q4 2025, redemptions as a share of beginning-of-quarter NAV in the non-listed BDC space nearly tripled to 4.71%. Among BDCs with over $1 billion in NAV, redemptions rose 217% quarter-over-quarter.

• Blackstone raised its BCRED quarterly redemption cap from 5% to 7.9% and deployed firm and employee capital to meet $1.7 billion in net withdrawals, avoiding a gate but straining its own balance sheet.

• In March 2026, Apollo gated its $25 billion ADS fund. Ares capped its $10.7 billion Strategic Income Fund at 5% after withdrawal requests hit 11.6%. Morgan Stanley and Cliffwater followed with their own caps.

• By late March 2026, Bloomberg reported over $4.6 billion in investor capital trapped behind withdrawal limits across the industry.

3.3 The Diamond-Dybvig Diagnosis

The critical observation is that much of this redemption activity was driven not by realised credit losses but by the fear of what other investors would do. Blackstone's BCRED had generated a 9.8% return since inception, and its president insisted portfolio quality remained solid. Yet the fund faced massive withdrawals because investors rationally concluded that being last in line for a limited redemption pool was the worst possible outcome.

The information asymmetry compounds the problem. Unlike publicly traded bonds, private credit loans are valued through appraisal-based, mark-to-model methodologies. Investors cannot independently verify whether NAV reflects reality. When JPMorgan marked down certain software loan portfolios and simultaneously reduced financing against them, it provided a public signal that internal valuations might be overstated. This is the Diamond-Dybvig information trigger: a credible signal that shifts expectations about asset quality, making the run rational even if the signal is only partially correct.

4. Where the Analogy Breaks

The Diamond-Dybvig framework illuminates the current crisis, but the analogy is not perfect. Three critical differences shape the dynamics in ways that make the private credit version both less systemically dangerous and more structurally intractable than a traditional bank run.

4.1 No Lender of Last Resort

Banks facing runs can access central bank liquidity through the discount window, emergency lending facilities, or deposit insurance funded by the broader banking system. Private credit funds have no equivalent backstop. The only substitute is the manager's own balance sheet (as Blackstone demonstrated) or fire sales into a thin secondary market. This absence means that the Diamond-Dybvig "good equilibrium" is harder to sustain because there is no external credible guarantor to anchor expectations.

4.2 Gates as Imperfect Suspension

Redemption gates are the private credit equivalent of suspension of convertibility. Diamond and Dybvig show that suspension can prevent runs by removing the incentive to rush for the exit. However, they also note that when the total demand for liquidity is uncertain, suspension is suboptimal. In practice, gates in private credit have the perverse effect of intensifying future redemption pressure: investors who are gated in one quarter submit larger requests the next, creating a queuing dynamic that can persist for years.

4.3 Blurred Solvency and Liquidity Lines

In the pure Diamond-Dybvig model, the run is a liquidity crisis affecting a solvent bank. In the current private credit episode, liquidity and solvency concerns are entangled. Rising defaults (5.8% by early 2026), the expansion of payment-in-kind arrangements, and growing use of "amend-and-pretend" tactics suggest genuine credit deterioration alongside the expectation-driven run. This muddies the diagnosis: investors cannot easily distinguish between a solvent fund facing a liquidity crunch and a fund whose assets are genuinely impaired.

5. Impact on Investors

5.1 Investors Locked in Current Structures

For investors trapped behind redemption gates, the consequences follow directly from the Diamond-Dybvig logic. When a fund sells assets to meet redemptions, it imposes a negative externality on remaining investors through three channels:

• Forced asset sales at discounts to NAV dilute the portfolio quality and returns for non-redeeming investors. Loans sold under pressure trade at discounts to their appraised value, crystallising losses that might never have materialised if the loans were held to maturity.

• If the manager borrows against credit lines rather than selling assets, the additional leverage increases risk for remaining investors without their consent, changing the risk profile of the fund post-hoc.

• The queuing dynamic means that even if underlying credit quality stabilises, the overhang of pending redemption requests can suppress NAV for quarters or years, as incoming cash flows are redirected to serve the queue rather than reinvested.

The distributional effect is asymmetric. Investors who redeem early (before gates are imposed) exit at full or near-full NAV. Those who remain bear the cost of supporting that liquidity. This is precisely the "first mover advantage" that creates the bank run equilibrium in Diamond-Dybvig.

5.2 Prospective and Opportunistic Investors

The crisis creates significant opportunities for investors with patient capital and strong analytical capabilities. These opportunities map onto three categories:

Secondary Market Acquisitions: Private credit secondary transactions are growing rapidly, with funds selling portfolio positions to specialised buyers at discounts to reported valuations. For buyers with the ability to underwrite individual loan quality, these discounts represent compensation for illiquidity and complexity rather than credit impairment. The secondary market for private credit is projected to grow from approximately 1% of industry AUM toward the 2–3% typical of private equity secondaries, representing a multi-billion dollar annual opportunity.

New Origination at Wider Spreads: As stressed funds reduce new lending activity to manage liquidity, the competitive intensity in private credit origination decreases. Borrowers who previously benefited from covenant-lite terms and tight spreads will face more demanding terms from the remaining lenders. Managers who maintained disciplined underwriting through the cycle are positioned to originate at historically attractive risk-adjusted returns.

Platform and Manager Selection: The crisis is creating a visible quality gap between managers. Those with robust operating models, transparent valuations, and genuine structural alignment between asset and liability duration will emerge with stronger franchises and lower cost of capital. Investors who can identify these managers early gain access to a structural premium that will widen as the weaker platforms lose their ability to raise capital.

6. The Operating Model as the Credible Backstop

Diamond and Dybvig conclude that preventing bank runs requires a credible external guarantor. In the absence of government deposit insurance or central bank facilities for private credit, the fund manager's operating model is the only available substitute. This is the paper's central strategic conclusion: operating model quality is not a back-office concern. It is the primary mechanism through which the "good equilibrium" of the Diamond-Dybvig model is sustained or lost.

6.1 Valuation Integrity as the Foundation of Confidence

In the Diamond-Dybvig framework, confidence depends on depositors' belief that the bank's assets are worth what the bank claims. In private credit, this translates directly to the valuation process. When investors cannot independently verify NAV, the manager's valuation methodology, independence, documentation, and consistency become the functional equivalent of the bank's capital adequacy.

The DOJ has publicly warned about "creative" marks and divergent valuation practices. The SEC is investigating ratings integrity in the space. Fund managers who treat valuation as a compliance exercise rather than a core operating function are structurally more vulnerable to expectation shifts. Best-practice managers maintain clearly documented methodologies, independent valuation committees, consistent treatment across portfolios, and defined escalation procedures for challenged credits.

6.2 Liquidity Management as the Suspension Mechanism

Diamond and Dybvig show that suspension of convertibility can prevent runs but only if the suspension threshold is correctly calibrated and the mechanism is credible. In private credit, this translates to the fund's liquidity management framework: the combination of cash reserves, committed credit facilities, asset sale capabilities, and redemption structures that determine whether and when gates must be imposed.

The difference between Blackstone's ability to meet all BCRED redemptions (through a combination of $8 billion in available liquidity and firm capital deployment) and Blue Owl's forced conversion of OBDC II into a drawdown vehicle is fundamentally a difference in operating model design. Blackstone built a liquidity architecture capable of absorbing stress. Blue Owl's architecture was reliant on continued subscription momentum to finance exits — a structure that fails precisely when it is needed most.

6.3 Data Infrastructure as the Early Warning System

The Diamond-Dybvig model is a game of expectations. The earlier a manager can identify emerging stress in its portfolio, the more options it has to manage the situation before expectations shift. This requires loan-level data infrastructure that goes beyond quarterly reporting: real-time covenant monitoring, automated source-system integrations, standardised data dictionaries, and reconciliation controls that support rapid credit risk escalation.

Managers who operate with manual waterfalls, spreadsheet-based fee calculations, and fragmented data systems face a structural disadvantage. They cannot move fast enough to identify deteriorating credits, adjust valuations, or communicate transparently with investors before the market does it for them. The information asymmetry that enables the bank run equilibrium is amplified by weak data infrastructure.

6.4 Investor Communication as the Confidence Anchor

The most underrated element of the operating model is investor communication. In the Diamond-Dybvig framework, the deposit insurance guarantee works by changing expectations. The guarantee does not need to be exercised; it works by existing. Similarly, a fund manager's ability to communicate proactively, transparently, and credibly during periods of stress can itself prevent the expectation shift that triggers the run.

This requires more than quarterly letters. It requires the operational capacity to produce granular, real-time portfolio analytics on demand, to explain valuation movements at the individual loan level, and to demonstrate through verifiable data that the portfolio's fundamentals support its stated NAV. Managers who can do this effectively are providing a form of private deposit insurance — using transparency and credibility as the guarantor.

7. Structural Implications and the Five Capabilities

7.1 The Structural Redesign Imperative

The current crisis will accelerate a fundamental redesign of private credit fund structures. The semi-liquid model that powered the asset class's growth from 2020–2025 has revealed its structural limitations. Going forward, the market will bifurcate between two models:

• Closed-end, committed capital structures where illiquidity is explicit, acknowledged, and priced from the outset. These structures eliminate gate risk entirely because there is no periodic redemption mechanism. Asset and liability duration are aligned by design, not simulated through structural engineering.

• Genuinely liquid structures (ETFs, listed vehicles) that invest in more liquid segments of the credit market, accept lower yields, and price continuously. These structures trade the illiquidity premium for genuine liquidity but eliminate the maturity mismatch.

The semi-liquid middle ground — which attempted to offer illiquidity premium with perceived liquidity — is the structural equivalent of a bank without deposit insurance. It works until it does not.

7.2 The Five Capabilities of an Excellent and Resilient Operating Model

For fund managers seeking to build or maintain credibility through this cycle, the operating model must deliver on five capabilities:

1. Dynamic Capital Allocation — Ability to rapidly deploy capital into high-conviction opportunities while maintaining dry powder for opportunistic buys during dislocations.

2. Advanced Risk Analytics — Sophisticated modelling of portfolio-level risks, including correlation shifts, sector concentration, and borrower-specific stress scenarios.

3. Operational Agility — Streamlined processes for deal execution, due diligence, and post-investment monitoring to capitalise on time-sensitive deals.

4. Investor Communication — Transparent, frequent, and insightful reporting that builds trust and provides clarity on portfolio performance and outlook.

5. Technology Integration — Leveraging AI and data analytics to enhance investment decision-making, operational efficiency, and risk management.

7.3 Who Must Act: The Addressee Question

The primary addressees of this analysis are not regulators. They are the individual fund managers, asset management companies, and their leadership teams. The competitive dynamics of the market will impose the discipline that regulation may or may not deliver, and they will do so faster and more decisively.

Investors who have experienced redemption gates, opaque valuations, and forced conversions of semi-liquid vehicles into drawdown structures will reallocate. They will move capital toward managers whose operating models demonstrably deliver on the five capabilities. Managers who cannot credibly deliver on these capabilities will face a structural decline in their ability to raise capital.

This is already observable. Blackstone's ability to meet all BCRED redemptions through operational preparedness and balance sheet commitment stands in direct contrast to Blue Owl's forced restructuring of OBDC II. The capital allocation consequences of this visible divergence will compound over subsequent fundraising cycles.

7.4 The Speed Imperative

If convergence toward these five capabilities is inevitable, the strategic question for individual managers is not whether to invest in their operating model, but how quickly they can do so. Speed matters for two reinforcing reasons.

First-Mover Advantage: Managers who achieve institutional-grade operating model capabilities ahead of their competitors will capture a disproportionate share of the capital reallocation that is already underway. The managers who can demonstrate, today, that their valuation processes are independently governed, their liquidity architecture is stress-tested, their data infrastructure supports real-time portfolio transparency, and their communication protocols are built for crisis conditions will be the first choice for reallocated capital. This advantage compounds: early movers attract capital, which provides operational scale, which funds further operating model investment, which attracts more capital.

The Cost of Delay: Managers who move too slowly face a compounding disadvantage that can become terminal. Each fundraising cycle in which a competitor demonstrates superior operational capabilities while the slower manager remains reliant on legacy processes, manual workflows, and reactive communication erodes the latter's competitive position. Institutional allocators increasingly treat operating model due diligence as a threshold requirement rather than a differentiator. Failing to meet that threshold does not result in a lower score; it results in exclusion from consideration.

8. Conclusions

The private credit redemption crisis of 2026 is a Diamond-Dybvig bank run re-enacted in the alternative asset management sector. The structural mechanics are identical: an intermediary performing maturity transformation against illiquid assets, creating fragility that depends on confidence to remain stable, and collapsing into a self-fulfilling run when expectations shift.

Four conclusions follow from this analysis:

1. The crisis is structural, not cyclical. The asset-liability mismatch embedded in semi-liquid private credit vehicles is a design flaw, not a market accident. It will recur in any structure that promises liquidity against inherently illiquid assets without a credible external backstop.

2. The operating model is the backstop. In the absence of deposit insurance or a lender of last resort, the fund manager's valuation integrity, liquidity architecture, data infrastructure, and investor communication capability are the only mechanisms available to sustain the "good equilibrium." Operating model quality is not operational efficiency. It is systemic stability.

3. Competitive convergence will enforce the standard. The market, not the regulator, will impose the five capabilities of an excellent and resilient operating model. Managers will either achieve this standard or lose clients to competitors who do.

4. Speed is the differentiator. The imperative is not only to build the operating model, but to build it fast. First movers will capture a disproportionate share of the capital reallocation now in progress. Those who move too slowly will not merely underperform. They will face a structural decline in their ability to compete, raise capital, and ultimately survive as independent platforms.

The liquidity illusion has been exposed. The question is no longer whether private credit funds can replicate the maturity transformation of banks. It is whether they can replicate the institutional infrastructure that makes banking stable. The answer lies in the operating model. And the time to build it is now.