Investors do not lose money because the math is hard. They lose because the clock is wrong. In private credit, time is the hidden liability. The cash looks near at hand until it is not. Then gates close, marks slip, and the need to exit arrives at the exact moment the exit is sealed. That is not bad luck. It is the design.
A “semi-liquid” fund is not half liquid. It is liquid when the tide is calm and illiquid when the storm hits. That asymmetry, flagged in prospectuses but too often sold as convenience, is a first-mover game. Early redeemers get out near par. Late redeemers face gates and queues. In game theory terms, it is a prisoner’s dilemma disguised as patient income. The current redemption waves show the payoff matrix in action: investors rush not because they panic, but because the rules reward speed. When Jeffrey Gundlach calls the label diabolical, he is describing optionality mispriced by design. Liquidity has been offered as a feature; in reality, it is a contingent claim whose value collapses in the states of the world when investors need it most.
Quarterly marks and smooth NAVs seduce. Then come step changes. Blue Owl cut values for a $14.1 billion tech-focused business development company by about 5 percent and nearly 3 percent for a $15.3 billion BDC, even as buybacks helped cushion the optics and one fund’s dividend was reduced. Apollo’s MidCap Financial Investment Corp. posted a quarterly loss, with non-accruals climbing to about $167 million from $48.5 million a year earlier. Oaktree marked down software assets, Sixth Street Specialty Lending trimmed its dividend. None of this is catastrophic. It is the slow reveal of a convexity problem. Mark-to-model magnets suppress observed volatility until the cash flows force a re-rate. Think of a steel beam under load. Nothing happens, nothing happens, then a small crack propagates and the failure is discontinuous. Private credit’s tranquil marks are the silence before the snap.
The weak link is not random. Direct lenders chased software because revenue looked durable, contracts were sticky, and collateral was “mission critical.” The AI transition and slower growth have exposed how procyclical that thesis can be. If a sizeable share of BDC and private credit portfolios tilts toward software and services, correlation rises at the exact moment investors assumed diversification. Some funds do have better covenants than public high yield. But in recent vintages, sponsor leverage rose and terms loosened at the margin. When non-accruals move and distributions get cut, the fragility is less about any single borrower and more about shared exposure to a business model re-priced by higher rates and technological displacement. Gundlach’s line that public high yield quality is better than pre-GFC is a tell: the public market has taken its medicine in real time. Private loans have not had to, yet.
The JPMorgan-led Qualtrics financing is another stress test, not a sideshow. With a roughly $5.3 billion debt package for the Press Ganey Forsta acquisition stuck, banks are staring at more than $500 million in paper losses. The existing Qualtrics loan trades in the mid-80s, making new issuance unattractive. This is a replay of 2022’s hung pipelines around deals like Citrix and Nielsen: capacity gets tied up, fees vanish, and the cost of capital rises for everyone else while banks warehouse risk they did not intend to keep. Private credit is often marketed as outside the banking system, but the links are real: bank credit lines to BDCs and funds, warehousing, total return swaps, and sponsor bridge commitments. The Financial Stability Board has warned on these growing ties. Even if the risk is not “systemic” in the 2008 sense, it can be system-relevant in how it throttles new lending and amplifies funding frictions.
Apollo’s push to offer daily marks on private-credit funds sounds like progress. It could be, if a deep secondary market existed. Without that, daily pricing risks becoming a model refresh, not price discovery. Goodhart’s law applies: when a measure becomes a target, it ceases to be a good measure. In a stress window, more frequent model marks might actually accelerate outflows if they validate investor fears without providing an exit. The Federal Reserve has been blunt: private credit loans are illiquid because the secondary market is thin to non-existent. Frequency does not solve that. It just shortens the feedback loop between perception and constraint. If actual marking discipline improves and comparable transactions are observed at scale, daily valuations may help. Until then, they risk being a dashboard that updates faster while the vehicle still has drum brakes.
Opinions diverge. Some research points to long funding horizons, higher equity cushions, and limited bank linkages as reasons the sector will not trigger a broader crisis. That view has merit. Private credit is unlikely to be the spark for a classic run on insured deposits or a sovereign funding shock. But that is the wrong question. The right one is path dependency. As Moody’s has warned, growing interconnectedness with banks and insurers means losses can migrate. The FSB and the Fed both flag opacity and illiquidity as stress multipliers. In practice, that looks like slower deal flow, higher spreads, and selective gating that erodes trust. It looks like pensions and wealth platforms explaining why a semi-liquid sleeve is taking longer to pay, while sponsors juggle distributions and lenders debate amendments versus accelerations. Not 2008. But not a soft landing for everyone either.
In nature, redundancy is antifragile. In finance, redundancy is a drag until it is oxygen. Investors were paid an illiquidity premium and a complexity premium. Many treated those as bonuses, not as compensation for real constraints. The behavioral error is mislabeling yield as income rather than as a bundle of risks with low-frequency, high-severity tails. Quarterly reports mute discomfort. Dividends offer comfort. Then one quarter the mark drops, the dividend trims, and the queue forms. The first-mover incentive reappears, and those who believed the story the longest face the worst terms. This is how stability breeds instability, in Minsky’s old framing. The point is not to panic. It is to price time correctly.
Ignore the slogans. Track redemption queues at semi-liquid funds, non-accrual rates across BDCs, distribution cuts, and the headroom on fund leverage facilities. Watch the hung-deal pipeline at banks and the pace at which those positions clear, because that sets the tone for all new money financing. Follow sector concentration disclosures, especially software and services, where AI and slower growth are re-rating cash flows. And interrogate any “daily valuation” scheme for evidence of real secondary trades, not just model outputs. The private credit machine is not collapsing. It is repricing. Repricing is messy when the exits are narrow and the clocks run on a sponsor’s calendar, not the investor’s. Yield is not free. It is rent paid to time and opacity. The bill is coming due in increments that feel small until they are not.
