Comprehensive Analysis of CLOs: Structure, Risks and Opportunities
26-Jan-26Comprehensive Analysis of CLOs: Structure, Risks and Opportunities
26-Jan-26March, 2026
The CLO Investor
AI, software credit and the case for European CLOs
Fair Oaks recently hosted a webinar examining two themes which have driven meaningful market volatility in recent weeks: the repricing of software-exposed credit following concerns about AI disruption, and the indirect effects of rising geopolitical tensions around Iran. The purpose of this insight is to summarise the key analytical observations from that discussion.
Fair Oaks’ current view is that the recent CLO spread widening is disproportionate to the underlying fundamental risk, particularly in European CLOs, and the market reaction has created a basis dislocation which is worth careful examination.
Software exposure: where the risk actually sits
The AI disruption narrative has prompted investors to reassess software credit across leveraged finance more broadly. However, a granular look at the data reveals that exposure levels vary materially across asset classes, and the perception of CLO risk has exceeded the reality.
US Business Development Companies (BDCs) carry the highest software and technology exposure of any sub-investment grade credit vehicle at approximately 26%, followed by US private credit CLOs at 19%. By contrast, US broadly syndicated loan (BSL) CLOs sit at 15%, and European CLOs at around 10%, in line with European loan market averages.1 The reasons for this are structural. CLO documentation typically imposes sector concentration limits of 10–12%, which have capped sector exposure during periods (such as 2000-2004) when issuance has been concentrated in certain industries. European loan markets also reflect a different borrower universe: issuers tend to be more defensive, and sponsors less focused on high-multiple software roll-ups than their US counterparts.
Furthermore, since 2023, the credit market has increasingly operated as a continuum. Stressed borrowers that the BSL market could not absorb have migrated to private credit, which has the structural flexibility to hold, restructure and extend. CLOs, constrained by OC tests, CCC buckets and reinvestment limitations, are not structurally designed to support this type of credit.
What the price data exposes
The market's response to AI-related concerns has been surgical in some respects and blunt in others. Observing the price performance of software loans, all other loans, and the BIZD ETF (a listed proxy for BDC and private credit performance) from September 2025 through end-February clearly illustrates the divergence. Software loans declined approximately 15% over the period; the BIZD ETF fell around 9%; ex-software loans were effectively flat, down just 2%.2
This is a rational outcome. The market has repriced the assets most exposed to AI displacement risk. What is less rational, and therefore of more interest from an investment standpoint, is the concurrent widening of CLO tranche spreads relative to similarly rated corporate credits.
Year-to-date EUR total return data shows that European loans have outperformed European BBB CLOs.3 This is structurally inconsistent with the relative risk profile of these instruments. BBB CLO tranches sit above substantial subordination: equity and junior tranches absorb losses before BBB noteholders are affected. They should, all else equal, exhibit greater resilience to idiosyncratic credit stress than direct loan exposure. The underperformance of CLO tranches relative to loans reflects sentiment-driven spread widening, not a deterioration in fundamental credit quality.
The fundamental health of CLO collateral
Notwithstanding the spread widening, the underlying collateral metrics for the CLO universe remain broadly sound. European loans continue to trade at higher prices than US loans on average, which is reflected in CLO weighted average portfolio levels. Most European CLOs have weighted average loan prices in the 96–98 range; the majority of US CLOs sit in the 94–96 band.4
Approximately 50% of European CLOs have BB overcollateralisation (OC) cushions (one of the key structural tests determining when cashflows begin to divert) at or above 4.5%, with the vast majority above 3%.4 Furthermore, CCC loan levels remain well below structural trigger thresholds. Around 50% of European CLOs have CCC levels below 3.75%, sitting comfortably beneath the 7.5% level4 at which OC ratios begin to be affected, and far below the mid-teens level at which cashflow diversion becomes a material risk.
The picture that emerges is one of heightened uncertainty: software prices have moved significantly, yet fundamental stress remains limited. Most CLOs have less than 5% of their loans trading below a price of 70.4 The 5–10% of portfolios trading in the 70–90 range largely reflects the software repricing, not a systemic deterioration in credit quality.
Relative value: CLO loan performance4
Sources: Bloomberg, Intex and Fair Oaks Capital data as at 02-Mar-26.
The spread basis dislocation
The widening of CLO spreads relative to equivalently rated corporate credit can be quantified. Across February, the CLO-to-corporate spread differential widened by 37 basis points at the BBB tranche level, 26 basis points at the A-rated level, and 8 basis points at the AA-rated level.5 These moves occurred on no material change in the fundamental credit quality of CLO collateral pools.
The decline in software prices has little impact on the credit quality of CLO tranches to date, particularly, as the data shows, when European CLO portfolios carry on average lower software exposure compared to private credit and US CLOs, and their structural protections remain intact. For investors able to look through short-term technical noise, the current basis offers an opportunity to access spread pickup at a given rating level without a substantial increase in fundamental credit risk.
Endnotes
- Pitchbook LCD, Morningstar, S&P Global, ICE BofA and Morgan Stanley Research as at 09-Feb-26. Software and technology exposure by asset class.
- Pitchbook LCD, Morningstar and Bloomberg as at 27-Feb-26. Weighted average bid and YTD returns; software vs broader loan market.
- JPMorgan as at 02-Mar-26. Euro (CLOIE) BBB total return; Euro loan total return. Indexed to 31-Dec-25.
- Bloomberg, Intex and Fair Oaks data as at 02-Mar-26. Universe of CLOs includes those with AAA factors > 0.85 and that are not more than 1 year post RP.
- JP Morgan as at 27-Feb-26. Euro (CLOIE) AAA, AA, A, BBB, BB primary spreads; Euro corporates asset swap spread, 3–5 year maturity



