Adnan Obuz has been calling the 2026 private credit turmoil a live case study since it broke in March, and the more he digs into it, the less it looks like a story about credit quality and the more it looks like a story about information architecture.
According to Adnan Obuz, the March 2026 events were a liquidity design problem, not a credit crisis
On March 6, 2026, BlackRock disclosed that its $26 billion HPS Corporate Lending Fund, known as HLEND, had received redemption requests totaling roughly $1.2 billion, about 9.3 percent of the fund’s net asset value, well above its standard 5 percent quarterly repurchase cap. BlackRock capped the payout at approximately $620 million and prorated the rest, the first time the fund had breached its own limit since launch. Shares fell about 7 percent that day, and the pressure spread: Blackstone, Apollo, and Blue Owl all faced elevated redemption requests across their own private credit vehicles in the same quarter.
None of this means the underlying loans went bad. Adnan Obuz’s read is simpler: private credit funds are built to hold illiquid assets in exchange for a yield premium, and when macro stress triggers a cluster of redemption requests at the same time, even well-underwritten portfolios hit a structural wall that has nothing to do with the quality of the loans themselves.
Where AI actually fits, according to Adnan Obuz
Private credit’s basic bargain is illiquidity for yield. That bargain holds fine in calm markets and gets tested exactly when investors get nervous, which is precisely the moment liquidity forecasting matters most. AI’s role here isn’t glamorous: it’s pattern recognition across investor behavior, redemption history, and macro indicators, built to flag rising redemption pressure before it becomes a forced-sale problem rather than after.
He has seen this pattern play out in consulting work outside private credit too: the firms that absorb shocks well usually aren’t the ones with the sharpest investment talent, they’re the ones with the cleanest data pipelines feeding decisions in real time.
Why the gap between AI’s promise and its use is so wide
McKinsey has estimated AI could add up to a trillion dollars annually to global banking, and most financial firms are nowhere close to capturing that. Adnan Obuz sees three consistent blockers in this space specifically. Data quality comes first: legacy systems across fund administration and investor records rarely integrate cleanly enough for a model to trust. Skills gaps come second, since firms need both people who understand the data and finance professionals who understand what the model’s output actually means for a portfolio. Governance comes third, and it’s the one that stalls otherwise-ready projects the longest, because explainability and regulatory expectations around AI-driven decisions in financial services are still being worked out in real time.
A pragmatic roadmap, according to Adnan Obuz
For a firm starting from close to zero, he suggests beginning narrow. Audit the data before buying any tool, since a clean pipeline matters more than a sophisticated model sitting on top of messy inputs. Pick one or two use cases with a measurable return, credit scoring and liquidity forecasting are the obvious starting points here, and aim for a working result inside 12 to 18 months rather than a multi-year platform build. Scale from what works, judged by business outcomes rather than technical elegance. Build governance into the architecture from day one rather than bolting it on once a regulator asks a hard question.
Governance can’t be an afterthought
Bodies like IOSCO and the Financial Stability Board have flagged that concentrated AI use across similar strategies can itself become a systemic risk, not just an operational one, if too many funds lean on similar models making similar calls at the same time. His own view is that audit trails and a human review layer aren’t optional extras in this space. In a market prone to cascading markdowns when sentiment turns, a model that can’t explain its own recommendation is a liability the moment something goes wrong, not a convenience.
None of this is a prediction about where private credit spreads or fund performance go from here, and it isn’t a recommendation to invest in or avoid any fund named above.
Frequently asked questions
Who is Adnan Obuz?
Adnan Obuz is a Toronto-based AI strategy consultant and capital markets analyst.
What actually happened in the 2026 private credit liquidity events?
In March 2026, BlackRock’s $26 billion HPS Corporate Lending Fund received redemption requests equal to about 9.3% of net asset value, above its 5% quarterly cap, and capped payouts at roughly $620 million. Blackstone, Apollo, and Blue Owl faced elevated redemption pressure in their own private credit funds the same quarter.
Does this mean private credit loans are performing badly?
Not necessarily. It reflected a liquidity structure mismatch rather than evidence that credit quality had deteriorated.
References
- Alternative Credit Investor, March 2026. https://alternativecreditinvestor.com/2026/03/06/blackrock-shares-slide-after-26bn-private-credit-fund-gates-withdrawals/
Adnan Obuz is a Toronto-based AI strategy consultant and capital markets analyst. Adnan Obuz publishes his notes at mrobuz.com and can be reached at businessplan@mrobuz.com.
Disclaimer: This article is for informational and educational purposes only. It does not constitute investment, financial, legal, or tax advice, or a recommendation to buy, sell, or hold any security or fund named here. Do your own research and consult a registered advisor before making investment decisions.