The market has caught up to the trigger

Our June thesis argued that private credit and private equity are held at marks no transaction has tested, and that the forcing mechanism - the thing that turns a slow problem into a fast one - runs through the PE-owned insurers. Their capital depends on the ratings assigned to opaque assets, and those ratings are not set by the market. In the days since we published, several others have set out a version of the same concern.

Writing in the Financial Times, Gillian Tett - whose book Fool's Gold chronicled the 2008 collapse - drew the explicit parallel to pre-crisis regulatory arbitrage: bankers once sliced risky loans into products that won flattering seals of approval from rating agencies, and it worked until the rotten content was exposed. Her conclusion is that the same game is now running through private credit and insurance.

The evidence she marshals is exactly the mechanism we flagged. A new Columbia Business School study - Rating Without Market Discipline, by Xuelin Li, Sangmin Oh and Giacomo Ricciardi - quantifies it: assets carrying a private rating are about twice as likely to suffer a credit loss as debt with the same public grade. Put differently, private ratings sit two to three notches above where a public rating would put the same risk. A privately-rated BBB - the lowest rung of investment grade - behaves like a public BB+ or BB, which is below it.

"For many, many years we have taken this credit rating system as a given."

Sangmin Oh, Columbia Business School, June 2026

The most striking detail is institutional. The Bank for International Settlements has warned that private ratings appear "systematically" inflated relative to public ones. And in 2024 the NAIC - the US insurance regulator - ran its own study and found private ratings averaged 2.74 notches above public ones. After a backlash from private capital groups and the newer rating agencies, it withdrew the paper. It still circulates, and it has informed warnings from the ECB, the IMF and the Financial Stability Board. The regulator found the gap, then was pressured into un-finding it.

What a re-rating does to the capital

This is the calculation the commentary stops short of. The notch gap is not abstract - under the NAIC's risk-based capital regime it converts directly into required capital, and the conversion is steep. The current C-1 charges (2026 RBC instructions) climb sharply as a bond crosses out of investment grade:

The Columbia finding - a privately-rated BBB really belonging at BB+/BB - is precisely the move from the 1.523% row to the 3.151% row. The capital charge roughly doubles for that single correction. Where the worst of the book belongs deeper into sub-investment grade, the charge runs to three or six times higher.

Apply that to the insurers the study names - those most reliant on private ratings, expressed as a share of their rated portfolios:

Take Athene as the worked example, since it is the largest. Of roughly $330bn in assets, a privately-rated book on the order of $60bn currently sits at a blended investment-grade charge of about 1.5%, implying near $0.9bn of C-1 capital. Re-rate that book to the sub-investment-grade level the study says it truly occupies and the charge roughly doubles - an incremental call of approximately $0.9-1.0bn on Athene alone. Push the weakest tranches toward single-B and the figure is multiples of that.

The Columbia authors estimate that closing the public-private gap across the industry would raise required capital by about $4.5bn a year. That is the system-wide number. Our point is narrower and sharper: it is concentrated, not diffuse. The 24-36% reliance at the PE-owned insurers means the gap lands hardest exactly where the conflict of interest is greatest - the insurer holding assets its own sponsor originated, graded by an agency the sponsor selected. An insurer does not need to be insolvent for this to bite. It needs a re-designation, and the regulator's own retracted study supplies the basis for one.

A note on these figures. The notch-to-charge translation is illustrative, not precise: the exact capital impact depends where in each rating band an asset sits and on its duration. We present the doubling as the central case and the multiples as the tail, which brackets the Columbia trio's own $4.5bn industry estimate. The direction is not in question.

What everyone is still missing: the equity below

It seems we are not the only ones noticing this, or something close to it. But here is where the wider debate - the FT, the BIS, the ECB, the Columbia study - stops at roughly the point our thesis begins. Each is concerned with whether the debt is rated too high. Suppose it is. Debt is still the senior claim: it sits above the equity in the capital structure and is paid first. If the senior debt is over-rated by two to three notches, the private equity ranked beneath it - which absorbs the first loss - is not over-marked by two notches. It is, in many cases, gone.

And the regulatory framework everyone is scrutinising does not even look there. Under the NAIC's own rules, the rating system is purpose-built for debt; equity investments fall outside the 1-to-6 designation scale entirely. They are carried at the manager's own valuation and written down only when a loss is judged "more than temporary." So the layer attracting all the scrutiny - senior debt - is the better-governed one. The layer that is larger, junior, first-loss and most exposed to the AI disruption now repricing software has no rating to inflate, because it has no rating at all.

This is the part of our June thesis we think still has the most road ahead of it. The recent commentary is concerned with whether the debt marks are a myth. On that we agree. We also looked at what sits beneath them. That a $1.8tn credit book may be mis-rated is now being argued in several places. The $3.6tn of unsold private equity ranked junior to it has barely featured in the conversation - and where it has been forced to trade, 2021-vintage venture stakes change hands at discounts of around two-thirds to their own carrying marks.

That the ratings are inflated is increasingly argued, and we agree with it. The equity beneath those ratings is the part still largely absent from the conversation - and it is both the larger number and the more exposed one. The myth will meet the market in the debt first, because the debt is where the trigger sits. It will meet it hardest in the equity.