1 · What Private Credit Is, in Plain Terms

Picture a mid-sized US company - a chain of clinics, a software firm, an auto-parts plant - that wants to borrow $50 million to acquire a competitor. A big bank like JPMorgan looks at its debt-to-earnings ratio and turns it down as too risky. Where does that company go?

Over the past 15 years, the answer has increasingly been: a private credit fund. These funds raise money from large investors (pension funds, insurers) - and, more recently, from ordinary retail investors through products like BDCs (listed funds) or interval funds (sold through financial advisors). They then lend directly to companies, at rates 2-4 points above bank rates, with no Fed oversight.

From $500 billion in 2020, the market swelled to $3,000 billion by early 2025 - a sixfold increase in five years (Morgan Stanley). Morgan Stanley had previously forecast $5,000 billion by 2029.

Why did this industry grow so fast? After the 2008 crisis, new rules forced banks to hold more capital and stop lending to higher-risk borrowers. Private credit funds filled that gap. The industry marketed itself as "crisis-immune" - since, unlike banks, it has no depositors who can suddenly demand their money back.

That's the theory. Reality is turning out differently.

2 · What's Happening - From Tricolor to Dimon's "Cockroach"

In the summer of 2025, surface-level numbers still looked fine: default rates under 2%, average yields of 10%, satisfied investors. Then, in September 2025, two US companies collapsed almost simultaneously:

• Tricolor Holdings - a subprime auto lender (mostly serving the Hispanic community). It defaulted amid allegations of "pervasive fraud" within the very loan packages sold to investors.
• First Brands Group - a global auto-parts manufacturer - filed for bankruptcy on 29/9/2025 in Texas. Disclosed assets were $1-10 billion, but debts were estimated at $10-50 billion - much of it hidden through "off-balance-sheet" financing structures (Wikipedia). The two founding brothers were indicted for defrauding creditors of "billions of dollars."

The list of creditors shocked the market:

A month later, on 16/10/2025, two US regional banks - Zions Bank and Western Alliance - disclosed they had been defrauded on loans made to a group of funds. Zions fell 13% that day, Western Alliance dropped ~11%. The S&P regional bank index fell 6.3% - its worst day since April (Bloomberg).

Right around then, Jamie Dimon - CEO of JPMorgan and the most powerful figure in US banking - told analysts on a call:

The press dubbed it the "cockroach moment" - a sentiment-shifting signal. Two quarters later, more cockroaches began crawling out.

3 · Why Isn't the "Immune" Market Actually Immune?

The "private credit is crisis-proof" argument rests on a simple assumption: investors lock up their money long-term and can't withdraw it. That holds for traditional funds. But over the past 15 years, the industry has pushed out retail products - BDCs and interval funds - that allow quarterly redemptions, typically up to 5% of the fund per quarter.

The problem: mid-market corporate loans have no secondary market - if you want to sell, there's no buyer. When all investors want to withdraw at once, a fund has only three options:

1. Gate - cap redemptions at 5%, and anyone asking for more must wait until next quarter.
2. Borrow from a bank to raise cash - shifting the risk onto the bank.
3. Have management put in its own money to preserve confidence (Blackstone did this in early 2026).

In Q1 2026, redemption requests surged:

At the same time, loan quality has visibly deteriorated:

• The "reported" default rate is still <2%, but once emergency restructurings (distressed exchanges) are counted, the real figure is around 5%. Morgan Stanley warns it could reach 8% (CNBC).
• Many mid-market companies can no longer earn enough to cover interest, let alone principal.
• The downgrade-to-upgrade ratio hit 3.3 : 1 in early 2026 - the worst since 2008.

4 · AI Is Rotting Out BDCs' Biggest Client Base

This is the least-discussed but most important part. Over the past 15 years, private equity funds have loved buying up SaaS (software-as-a-subscription) companies with borrowed capital - because SaaS has wonderful characteristics: recurring revenue, sticky customers, predictable forecasts. Private credit funds happily lent against "annual recurring revenue" (ARR) - no profit required, just growing ARR.

As of 30/9/2025, software made up 29% of BDC portfolios - the single largest sector concentration, ahead of both healthcare and consumer finance (Octus). UBS estimates 25-35% of the entire private credit industry faces high AI risk - not yet priced in by the market.

What happened

On 3/2/2026, Anthropic launched Claude Cowork - a suite of AI agents capable of independently handling entire workflows: contract review, financial analysis, customer support, project management. Right after, OpenAI announced Project Operator - letting AI "operate" any piece of software the way a human would.

Within 48 hours, SaaS market capitalization lost $285 billion. By the end of March, total damage exceeded $1,000 billion (Tech Startups). The market called it the "SaaSpocalypse":

Why this is a disaster for private credit

The classic SaaS business model is selling seats: a 1,000-employee company buys 1,000 licenses at $50/person/month. When an AI agent can do the work of 5 junior analysts, that company no longer buys 1,000 seats - it buys 200, plus a few usage-based AI agent contracts. ARR growth flips from +20% to −5%.

This sets off a chain reaction:

1. A SaaS company loses ARR → breaches loan covenants (which are ARR-based).
2. The borrower asks to switch to PIK interest instead of cash, to preserve cash for operations.
3. The BDC still records the loan as "performing," but the non-accrual ratio keeps climbing.
4. By the 2026-2028 refinancing wall, new lenders refuse or demand harsh terms - and a "real" default becomes a "recorded" default.

The evidence is already in: Golub Capital (GBDC) - a BDC with 27% of its portfolio in SaaS - cut its dividend 15.4% in February 2026; analysts expect a further 10-20% cut (Seeking Alpha). The Morgan Stanley North Haven fund received redemption requests for 11%. The $33 billion Cliffwater fund has half its investors lined up to withdraw.

5 · Who's Affected - And Could It Go Global?

Retail investors - direct losses

15% of BDC fund assets now belong to ordinary retail investors - up from near zero a decade ago. When funds cut dividends, share prices fall 15-30%, or gates go up - retail investors can't get their money out. This is the biggest political flashpoint.

US banks - through a back door

Commercial banks don't directly hold private credit loans, but they lend to the funds themselves. This lending (called NDFI lending) has grown from $56 billion in 2010 to $1,320 billion by end-2025. JPMorgan alone has roughly $160 billion of exposure. If the funds fail, banks take the loss.

Insurers & pension funds - the quiet but largest channel

In Europe, insurers hold 5.8% of investment assets in private credit; occupational pension funds hold 4.4%. In the US, private-equity-owned insurers (Athene, Global Atlantic) hold roughly $1,000 billion, much of it channeled into private credit. This is the channel through which losses ultimately land on the elderly - via retirement insurance contracts (Axios).

Global spillover?

There are three transmission channels:

1. Through international banks. UBS's $500 million loss on First Brands is enough to show European banks have exposure. Deutsche Bank, BNP, and Japanese banks all carry similar loans.
2. Through pension/sovereign funds. Japan's GPIF, Canada's CPP, and Singapore's Temasek all allocate 5-10% of their alternatives portfolios to US private credit. Losses sit hidden inside NAV, not immediately visible.
3. Through market sentiment. Shares of Blackstone, Blue Owl, Apollo, KKR, Ares, and Carlyle have fallen an average 15-20% from their early-2025 peaks. A drop of 40% or more would drag down the S&P 500 and global markets.

The risk of spillover from Asia/Europe into the US is low (the scale is small). The reverse risk - the US pulling capital home to plug holes, draining USD liquidity from East Asia - is high. This is the same mechanism that played out in 1997 and 2008.

6 · Did Anyone Warn Early - And What Can They Actually Do?

Warnings have been out for two years: the IMF in 4/2024, the Fed in 5/2025, the FDIC in 2025, the BIS in 2025 - all wrote about the risk. The problem isn't a lack of warnings - it's a lack of tools.

Unlike banks, private credit funds don't fall under the Fed. The Fed can't demand stress tests or set capital requirements. The SEC can require disclosures, but can't intervene in a fund's structure. There is no lender of last resort when a fund runs into trouble - unlike banks, which have the Fed's discount window.

Steps taken in recent weeks:

• 25/3/2026: FSOC (the interagency financial stability council) held a special session on private credit.
• 30/3: Treasury met with insurance regulators.
• 10/4: The Fed sent detailed requests to major banks to report exposure to individual funds (Fortune).
• 14-15/4: The IMF published its spring 2026 Global Financial Stability Report - judging that "systemic risk is contained to date" but warning that retail exposure keeps rising (IMF GFSR).
• 24/4: The SEC, Treasury, and Fed opened a joint investigation - focused on how funds value their loans.

7 · Signals Worth Watching

If conditions shift from "normal cycle" to "systemic crisis," the following indicators will move. Here are 6 signals that are easy to track:

8 · Conclusion - Three Short Answers

Is this a crisis? As of 24/4/2026, not yet. The IMF uses the word "contained," and Dimon says it "probably does not pose systemic risk." But this is the first time this $3,000 billion market has been tested through a real credit cycle. The tricks used to mask stress (PIK, emergency restructuring, loosened lending standards) have been overused for three years - "reported" defaults and "real" defaults have diverged by 3-4 percentage points. Losses will surface; the only question is how fast.

If it worsens, how deep could it go? Base case: $90-120 billion in losses over two years, many funds cutting dividends, retail investors marked down 15-25%. Bad case: confidence in major managers evaporates, banks call in loans against NAV, alt-manager stocks collapse together, spillover into regional banks and insurers - a scope of $400-600 billion. What decides between the two scenarios is the speed of disclosure: fast, and it's containable; slow, and stress builds up before it blows.

Could it go global? Yes - through international banks, through Japanese/Canadian/Singaporean pension funds, through alt-manager stocks. But the primary transmission mechanism runs one way: the US pulling capital home to plug holes, draining USD liquidity out of Asia. Similar to 1997, similar to 2008, at a smaller scale.

And the AI shock to SaaS is a new variable with no precedent in the historical playbook. 29% of BDC portfolios sit in an industry whose very revenue model may never recover - something no prior credit cycle has ever had to face.

Financial history has one rule: every crisis was once called "this time is different." The 1980s S&L crisis "couldn't happen because of the FDIC." The 1997 Asian crisis "couldn't happen because of the Asian Miracle." 2008 "couldn't happen because of sophisticated risk models." SVB in 2023 "couldn't happen because of HTM accounting." Every time, the reason it "couldn't happen" sounded reasonable - and every time, a new mechanism found its way through anyway. Private credit may not be "the next GFC." But anyone insisting it cannot become a crisis should remember that exact sentence has preceded every major crisis of the last 40 years.