Leveraged Finance Asset Allocation Insights: Markets Hit an AI Inflection Point

The leveraged finance market is currently grappling with an “AI inflection point,” where the early promise of greater efficiency is giving way to a more uncertain outlook, with potential AI-related pressures on revenues and margins. This shift has introduced a sense of anxiety as investors increasingly scrutinize the actual cost of adoption on an issuer-by-issuer basis. Behind this angst is the realization that AI may cannibalize existing business models and provide increased pricing transparency in certain industries that have long benefited from opacity. This tension is manifesting in the credit markets as a “sell first” mentality, especially as fundamental cracks begin to show, with default activity rising in the technology and services sectors.

Despite these AI-related headwinds, the broader credit environment is supported by resilient macro factors and fundamental tailwinds. Revenue growth remains sustained across high yield (HY) and leveraged loan issuers, bolstered by a labor market that continues to show strength through decent job creation and manageable unemployment. The Fed’s easing stance combined with overall strong earnings from “AI ecosystem enablers” – such as tech hardware, capital goods, and utilities – have provided a necessary cushion. Furthermore, the overall high yield market continues to show resilience, with index-level spreads trading around 300 basis points (272 option-adjusted spread/298 spread-to-worst) and strong demand for new issuance. Sentiment in the loan market is not as strong given the relatively higher concentration of issuers in the software sector, but valuations are still trading well inside of longer-term averages.

Our outlook remains relatively constructive but heavily split between likely winners and losers. While the market is optimistic about hardware and power providers, fueled by datacenter buildouts, a cautious stance is warranted for software and business services companies that are most vulnerable to AI disruption. These sectors are seeing loan and bond prices come under pressure as investors reassess companies’ business models to account for potential obsolescence. Consequently, we favor higher-quality assets and moderate positioning, with a focus on navigating the volatility caused by subsector dispersion while watching for further signs of AI-led distress in lower-rated, smaller-cap issuers.

Key Data

Source: Bloomberg as of 19 February 2026. High yield represented by the Bloomberg US Corporate High Yield Index; spread is OAS and yield is yield-to-worst. Leveraged loans represented by the Morningstar LSTA Leveraged Loan Index; spread is spread-to-maturity and yield is yield-to-maturity. CLOs represented by the JPM Post-Crisis CLOIE; spread is discount margin to worst and yield is yield-to-worst. Methodologies used differ between asset classes to account for differences in characteristics, including, but not limited to, call protection, reference rate, cash flows, coupon rates, and embedded options.

High Yield Bonds

Overall, we expect positive, carry-based total returns for high yield bonds. While spreads have not widened relative to certain other credit markets (such as bank loans), we view relative valuations as reasonable given the lower concentration of technology issuers in the high yield market.

Thus far in 2026, markets have focused on the potentially negative intermediate- to long-term impact of AI technologies on a range of issuers, particularly those in the software and B2B services sectors. While we believe AI will affect many businesses, we view the breadth of related volatility as excessive. Last-12-months (LTM) par-weighted default rates rose slightly month-over-month through January to 1.97%/1.12% (with/ without distressed exchanges). Ratings activity was positive in January, with a par-weighted upgrade/ downgrade ratio of 1.83 (1.13 by issuer), versus 1.27 (1.03 by issuer) in January 2025. (Default and ratings statistics per JPMorgan as of 2 February 2026.)

From a valuation standpoint, the Bloomberg US Corporate High Yield Index spread-to-worst is at 298 (BB 189, B 331, CCC 609), 9 basis points wider month-to-date (as of 17 February), while the yield to worst (YTW) is unchanged at 6.58%. Given the focus on AI disruption, technology spreads are now the widest in the benchmark, taking the crown from communications. The high yield market has held in better than the loan market given its lower concentration of tech issuers. We maintain our 2026 spread outlook, anticipating excess returns in line with or modestly below spreads. (All statistics per Bloomberg as of 17 February 2026.)

We don’t see much of a bull case for spreads beyond where they are now, while the bear case is a material slowing of the economy. US high yield new issuance totaled $30.0 billion in January, roughly in line with average 2024 and 2025 levels. Lower-quality issuance was high, at 14.6% of the total, versus 6.2% for 2023- 2025. Fund flows were a headwind in January, at -$1.3 billion, the largest outflow since April 2025 (post- “Liberation Day”), and following strong flows of $18.9 billion for the full year. (All statistics per JPMorgan as of 2 February 2026.)

While current tight valuations can limit upside potential in fixed income in 2026, we believe high yield bonds remain an attractive asset class. All-in yields are adequately compensating investors for credit risk amid moderate default expectations. We view valuations as broadly fair in the context of our 2026 economic outlook, which would be supportive of reasonably tight valuations. With its more compelling yields, we view the belly of the high yield market – low-BB to mid-B rated bonds – as a sweet spot for investors. Overall, we expect positive, carry-based total returns for high yield bonds (but muted excess returns) and believe defaults could tick a bit higher in 2026, albeit from low current levels.

Fewer Tech Issuers in the High Yield vs. Leveraged Loan Market

The higher concentration of technology issuers in the loan market accounts for a portion of February’s underperformance of loans vis-à-vis high yield.

Source: Bloomberg as of 28 February 2026.

Leveraged Loans

Current conditions in the leveraged loan market remain conducive to disciplined credit selection, though the majority of alpha may come from avoiding credits that are heading for distressed territory rather than picking those that just appear cheap.

Loan issuers continue to show solid fundamentals, with support from a stable macro backdrop. Strong capex spending from hyperscalers remains a key economic pillar, and this investment cycle – combined with a resilient labor market, with unemployment near 4.3% and steady payroll gains – is helping offset growing consumer pressures, such as slowing wage growth and rising delinquencies. Consumer spending and overall growth expectations should hold up barring a meaningful rise in layoffs.

Despite modest, stable EBITDA growth and improving interest coverage ratios – with help from two rate cuts in late-2025 and ongoing repricing activity – market volatility has increased. Initial concerns about AI-driven disruption in software have broadened to insurance brokers, logistics firms, and parts of commercial real estate, prompting investors to reassess business models and asset valuations. As a result, the share of the Morningstar Index trading at par or above has fallen sharply, from 67% in mid-January to 30% as of 22 February (see chart). Selling pressure has been broad-based, affecting even defensive names and creating opportunities for selective credit picking.

Technicals remain supportive as net loan supply continues to lag. Primary activity remains dominated by opportunistic repricings, while broader issuance has slowed due to AI-related volatility. Repayments continue to exceed new supply, reducing outstanding loans and supporting demand for performing credits, particularly in sectors less exposed to AI disintermediation. However, conditions could turn more neutral if certain factors weigh on collateralized loan obligation (CLO) issuance – including a decline in CLO equity distributions, managers forced to realize losses in software, or rising defaults that pressure portfolio tests. Retail flows remain flat and are likely to track shifts in risk sentiment and rate cut expectations. Recent volatility has pushed loan spreads to the widest levels since just after “Liberation Day” in April 2025. Overall yields remain lower due to prior Fed cuts, but with reduced consensus on further easing, yields should stay relatively stable absent external shocks.

The Percent of Loans Trading at Par and Above Has Dropped in 2026

Source: Pitchbook/LCD as of 23 February 2026.

Software Volatility Has Driven Loan Spreads to ‘Liberation Day’ Wides

Spread to maturity

Source: Pitchbook/LCD as of 20 February 2026.

CLOs

We believe CLOs have attractive total return potential relative to other equivalently rated fixed income assets. Despite recent softness driven by the selloff of loans in the software sector, the technical picture appears supportive given strong ETF and institutional demand and high redemption and amortization volumes. We expect to see a bifurcated market favoring CLO managers who focus on credit selection to mitigate downside risks.

The recent selloff in software loans has created an environment where CLO investors are much more focused on tail risks in CLO portfolios, particularly exposure to AI risk. This has created additional bifurcation between tier 1 and tier 3 managers, along with a reluctance to invest in lower mezzanine tranches, especially for deals with lower market value overcollateralization (MVOC) cushions.

The software sector represents an average of 12.4% of US CLO collateral1 and is the highest-weighted sector of the loan market, at over 16% of the Morningstar LSTA Leveraged Loan Index. Given the high weighting of the software sector in CLOs and the broader loan market, CLOs have seen fundamentals deteriorate alongside the broad selloff. CLOs are not forced sellers during periods of increased volatility, but declining loan prices can increase haircuts on CCC and defaulted assets, leading to lower overcollateralization cushions and increasing the risk of test breaches. MVOCs for BB rated CLOs in their repayment periods have declined an average 1.2 points in 2026 to date, with individual manager results highly correlated to AI/software exposure.2

On the surface, the selloff has improved the equity arbitrage, which was under significant pressure coming into 2026. An improving equity arbitrage, in this case driven by loan spreads widening faster than CLO liability spreads, should increase a manager’s ability to issue a new CLO. Year-to-date, US CLO new-issue volume totals $24.9 billion, compared to $24.7 billion over the same period last year.3 Meanwhile, refi/reset/reissue volumes have slowed year-over-year, with year-to-date volume of $34.0 billion trailing the $57.6 billion over the same period in 2025.4 Despite the better-looking equity arbitrage and ongoing issuance, the bifurcation in the loan market and demand for performing loans relative to more stressed names still makes ramping a new deal challenging.

Despite recent softness, we believe CLO tranches offer attractive total return potential relative to other equivalently rated fixed income assets. The technical picture appears supportive: CLO ETF inflows have driven total AUM to over $44 billion,5 institutional demand is strong, and redemption and amortization volumes remain high. However, tight valuations may argue for an incrementally more defensive stance.

Overall, we view the market this year as favoring CLO managers who avoid excessive concentration and rather focus on credit selection to mitigate downside risks. As a result, we view a nimble and robust bottom-up approach to security selection as paramount in 2026 given the dispersion in the loan market, in which certain CLO portfolios holding weaker credits may eventually experience impairments to the lowest-rated debt tranches. This could result in attractive opportunities more broadly to move down the capital stack.

BB Rated CLO Spreads Versus MVOC

Source: BofA Global Research, Intex, J.P. Morgan. BofA Global Research “CLO Weekly,” as of 20 February 2026. MVOC: Market value overcollateralization.