New modelling proposals could improve capital efficiency for senior tranches, incentivising a strategic shift toward AAA-A debt
Insurance investment across CLO debt structure could be shaken up by a series of changes to how risk-based capital requirements are calculated, if new modelling proposals are accepted by regulators.
Over the last four years, the National Association of Insurance Commissioners (NAIC) and American Academy of Actuaries have been collaborating on a project looking to assess how risk-based capital requirements are calculated for CLO debt.
If accepted, the potential changes could be in place by the end of the year, according to individuals familiar with the project, and could lead to significant changes to how the insurance sector invests in CLO bonds.
In May 2022, the structured securities group at the NAIC sent a memo to regulators highlighting risk-based capital arbitrage available in CLOs, specifically in mezzanine debt. The NAIC highlighted that if an insurer held a B-rated loan from an underlying CLO portfolio in 2022 it would receive a roughly 9% risk-based capital charge pre-tax, but if held via a CLO as a vertical slice it would be closer to 3%.
Shortly afterwards the NAIC and Academy collaborated to set up the risk-based capital Investment Risk and Evaluation (IRE) working group which was tasked with looking at the problem of how to frame risk-based capital on structured securities more broadly, starting with CLOs.
The Academy highlights, the risk-based capital IRE working group’s objective was to “define several risk buckets for CLOs according to comparable attributes and then assign a C-1 factor to each bucket”.
The group showed preliminary findings of their model in late 2025, using a sample of six BSL CLO deals, and now they have presented the results from their refined model which incorporated the entire BSL CLO deal universe.
The project involved a Monte Carlo model running 10,000 simulations around CLO securities looking at their tail risk, the worst 10th percentile of losses, and used this scenario to calculate how much money insurers should set aside to satisfy risk-based capital requirements when investing in specific CLO tranches.
The framework proposes two distinct approaches to how CLO tranche C-1 factors are modelled.
The first option would replace the current modelling framework with a simplified, ratings-based approach. Under this method, capital charges would be assigned based solely on a tranche’s credit rating, adjusted for factors such as reinvestment periods, with lower-rated bonds attracting higher charges.
The second option would retain the existing framework for senior tranches but introduce targeted adjustments for riskier parts of the capital structure. In particular, CLO tranches rated BBB- and below with a thickness of less than 4% would face higher capital charges, reflecting their reduced ability to absorb losses in stressed scenarios.
Speaking to SCI, Stephen Smith chairperson of the Academy’s C1 Subcommittee who led the risk-based capital IRE working group on this project, explains that this methodology came about after digging into the triple-B-minus tranche, commonly referred to as just triple-B, which is where a lot of insurance holdings are in the lower portion of the capital structure.
In their analysis the Academy found a common structure where the triple-B tranche is split into a senior and junior portion. The senior portion typically has a thickness of 6% with a subordination of 12%, whereas the junior portion often attaches at 10-11% and detaches at 12%, representing a much thinner portion of the overall debt stack.
“They represent the very senior-most part of what would traditionally be the double-B-minus tranche” Smith notes. “While those thin tranches were able to get a BBB- rating if they went to a ratings agency that used a first dollar of loss approach like S&P or Fitch, our modelling was not simply looking at the probability of default but how severe are the losses in the tail scenario,” he says.
“We found that those junior triple-B tranches had meaningfully more tail risk than the standard 6% thickness triple-Bs, so we believed it would be prudent to differentiate between those two different types of triple-B tranches."
Smith concludes that, “the factors we’re proposing would generally reduce the capital requirements for higher rated tranches, and starting from A- and below there would be an increase in capital charges”.
Looking at the triple-B tranche, which is where the bulk of insurance holdings of CLO debt are, the chart below shows that the additional distinction between tranche thickness significantly increases the capital charges for tranches with less cushion.
Split triple-B issuance in the CLO market has surged since 2024, according to Bank of America research, as managers used the structuring technique to take advantage of the regulatory arbitrage giving insurance investors double-B esque yields with a triple-B rating, and thus more attractive capital treatment.
BofA notes that although triple-B CLO tranches have been the widest investment grade product in the fixed income universe, implementing the modelled C-1 factors would mean CLO BBB bonds would have higher risk-based capital charge per spread compared to other assets (RMBS, IG corporates), effectively removing the regulatory arbitrage highlighted in 2022.
The research estimates that insurance accounted for roughly 20-30% of junior triple-B bonds issued in the first half of 2025, stating that it expects that if the second modelling option is adopted it will stifle insurance demand for these junior tranches, concluding that split triple-B issuance will revert to pre-2023 levels.
Insurance interest in senior tranches set to grow
Prospects for insurance demand for senior CLO debt, as Smith mentions, look set to grow under the modelled C-1 factor. Tranches with ratings between triple-A and single-A will see a reduction in risk-based capital requirements under the proposed changes, as shown in the chart below.
Pratik Gupta, head of RMBC and CLO research at BofA, tells SCI that this could mark an important shift in the CLO market, highlighting triple-A debt will be a particularly appealing proposition for insurers. He predicts the sector’s prominence in the CLO market will only grow if the Academy’s proposals come into action.
“If formalised, that’s a pretty significant change in how insurers will view CLO debt. Insurance is already a dominant buyer for mezzanine bonds but with this change AAA bonds will be even more attractive for them. With this change insurance could be a bigger force in CLOs going forward.”
The new proposals are currently exposed for comments which will be presented at the risk-based capital IRE meeting in May.
Smith adds that he expects the fate of private credit CLOs, which were not included in the group’s analysis, is still undecided, but whether or not they will be impacted by this project will ultimately become clear after the meeting in early May.
“There’s an open question at the NAIC right now as to whether the work we’ve done should also be applied to middle market CLOs,” he adds. “That is not yet resolved, and I expect it to be discussed based on comment letters that will be addressed in the 6 May meeting.”