Black Boxes All the Way Down Private credit sold itself as the smart money alternative. Better yields. Less volatility. Sophisticated underwriting. What nobody advertised: the whole system runs on faith. And faith is leaving the building. Two stories from the Wall Street Journal this month tell you where things stand. Cliffwater Corporate Lending Fund. $42 billion. Largest SEC-registered private credit interval fund in the country. Last quarter, 14% of investors wanted out. The fund bought back 7%. The rest are stuck. Why? Look inside. 3,600 holdings. Most are loans to companies you've never heard of. 71% of assets are Level 3, valued using "inputs that are both significant and unobservable." That's management's best guess dressed in accounting language. Another 28% are stakes in other private funds. For those, Cliffwater doesn't do its own valuation. It accepts the NAV reported by the other manager. Investors trust Cliffwater. Cliffwater trusts other managers. Nobody independently verifies the marks. Then there's the Ares problem. Cliffwater told investors for years that its stake in an Ares lending fund would liquidate June 30, 2025. Kept adding to the position. June 30 passed. The holding was still there. Gains kept growing. No explanation. Months later, Cliffwater called it "an error." The fund had quietly gone evergreen. That's the kind of disclosure that turns doubt into a run. Story two: Western Alliance sued Jefferies after borrower First Brands went bankrupt. Western Alliance had lent to a special-purpose vehicle Jefferies created. The SPV funded First Brands. When it collapsed, Western Alliance charged off $126 million. Now both sides are arguing over who actually owns the risk. The SPV was supposed to make the exposure clean and isolated. Instead it became the thing nobody can untangle. This is the architecture underneath private credit. Banks lend to SPVs. SPVs lend to borrowers. Regulators set lower capital requirements because SPVs supposedly isolate risk. Analysts estimate $300 billion of bank exposure built on that assumption. Zoom out and every layer has the same flaw. Fund managers mark their own illiquid assets. Banks lend against those marks. Regulators approved the structures. Nobody in the chain gets rewarded for being the first to say the numbers are wrong. Munger called it incentive-caused bias. You don't need fraud for a system like this to break. You just need enough people to stop believing the marks at the same time. That could be happening now. Bank stocks are down nearly 10% since January. Cliffwater can't keep up with redemptions. If every layer of this system depends on trusting the layer below it, what happens when the bottom layer can't explain its own numbers? https://lnkd.in/ePYeAERv