Concerns over AI-driven disruption have intensified across risk assets in recent weeks, and the broadly syndicated loan market has not been immune. Software-sector loans — the largest segment of the leveraged loan universe — have declined 3.51% since the end of 2025, dragging overall loan returns down more than half a percent. Given the sector’s outsized representation, volatility in software credits tends to affect the broader market. How concentrated is that exposure, and where do the underlying risks reside?
Software looms large
Software is the largest sector in the broadly syndicated leveraged loan market, accounting for roughly 13% of the $1.53 trillion outstanding tracked by the Morningstar LSTA US Leveraged Loan Index. As of the end of January, speculative-grade software borrowers had $194.5 billion of leveraged loans outstanding, modestly below the $198.8 billion peak reached at year-end 2025, following a large M&A-driven repayment of WorldPay’s loan.
Outstanding loans to software issuers expanded by 70% between 2020 and 2025, driven largely by a surge in issuance in 2021 amid ultra-low interest rates and a flurry of PE activity. That year, software companies raised a record $53.5 billion in the BSL market to fund buyouts and other M&A transactions, including $47.6 billion from sponsored borrowers. The rapid build-up in tech exposure was a key contributor to the 30% expansion in the leveraged loan asset class over the five-year period.
The software sector includes some of the largest borrowers in the BSL market. Cloud Software and UKG — the two largest software constituents in the index — both rank among the top 10 borrowers overall.
Software loans lose their shine
Since the end of 2025, the weighted average bid for software names tracked by the LLI has fallen by more than four points, to 90.51 as of Feb. 4 — the lowest level in more than three years. These loans are now priced 504 bps below the LLI’s overall weighted average bid of 95.55, marking the widest discount in at least six years. For context, the gap stood at just 193 bps at the end of December. Prior to the rate-hike cycle, software names commanded a premium, with an average bid of 99.47 at year-end 2021 versus 98.64 for the broader index.
Similarly, the discounted spread-to-maturity on software loans widened sharply as prices fell, reaching 595 bps by Jan. 30 — 178 bps above the overall index. While software loans typically carry above-average nominal spreads due to their heavy concentration in the B-minus ratings bucket, the current differential is well above historical norms. By comparison, the spread premium averaged roughly 80 bps in 2025 and just 60 bps in 2024.
B-minus exposure near $100 billion
As of the end of January, half of the outstanding software loans — $97.7 billion — came from B-minus rated borrowers, with an additional 10% ($18.6 billion) in the triple-C category. Both concentrations exceed those of the broader market, where B-minus and triple-C exposures account for 23% and 6%, respectively.
Software also faces a modestly heavier near-term maturity profile: 25% of outstanding loans in the sector mature in 2027-28, compared with 21% on index loans broadly.
Until sentiment toward the sector deteriorated in recent weeks, software borrowers enjoyed broad access to refinancing in the broadly syndicated loan market. Last year, the sector raised $32.1 billion to refinance existing debt — representing 17% of all BSL refinancing activity, slightly above the sector’s overall share of the market and indicative of generally receptive conditions at the time. That window now appears to have closed: not a single new software loan has launched in February, for any purpose.
This could create opportunistic deal flow for direct lenders. Last year, some $37 billion of syndicated loans were refinanced in the private credit market, including some software companies like Mitratech (LINK) and Upland Software (LINK). Moreover, several of today’s largest software issuers in the BSL market previously relied on private credit, such as Kaseya, Qlik Technologies and Finastra. However, with private credit lenders already heavily exposed to software and tech, many might not want to increase investment in the sector.
New issue trends
With BSL lenders pulling back from software, sector diversification in the new-issue loan market is poised to depart meaningfully from historical norms. Over the past five years, software borrowers have represented roughly 20% of all sponsor-backed loan issuance and accounted for a similar share of loans financing buyouts and other M&A activity. As recently as last year, software companies made up 23% of BSL issuance tied to M&A — the second-highest level of the past decade, narrowly trailing 2022’s 24% — while the broader tech sector accounted for 26%.
Although 2026 has just started, that mix has shifted sharply so far this year: tech-related loans comprise just 8% of M&A-driven BSL volume, while healthcare has claimed the top position with 41%, mainly driven by the $7.25 billion Hologic transaction.
Overall leveraged loan issuance to software companies was $79.1 billion last year, encompassing refinancings, dividend recapitalizations, M&A, and other uses of proceeds (excluding repricings and maturity extensions). Although issuance declined 7% from 2024 levels, it still marked the third-strongest year on record.
The composition of that volume, however, shifted meaningfully from the ultra-low-rate era: just 31% of issuance came from B-minus rated borrowers, compared with roughly half of total volume during the 2019–2022 period.
As issuance shifted towards higher-rated borrowers in recent years amid the elevated cost of debt, the average leverage ratios decreased from cycle peaks. Software buyouts financed in the BSL market since the start of 2023 had an average debt/EBITDA ratio of 5.6x at closing, down almost a full turn from pre-rate-hike levels. Nonetheless, these deals maintained their half-turn difference from the overall market average, at 5.1x for all LBOs.
Software credit: divergence ahead
Where does the software sector go from here? As market volatility subsides, will the industry revert to a familiar equilibrium, or will we see a lasting divergence between winners and losers? More specifically, will credit markets begin to more clearly differentiate between software companies that investors believe can withstand disruption — and potentially unlock new growth through AI — and those that may struggle to adapt or face displacement by AI-driven alternatives?
In earnings calls, private credit providers have begun to parse the sector to focus on software companies whose products are considered “mission critical” and that produce valuable databases. Investors are more bearish on software for content creation or tool-based, single-function products.
According to PitchBook’s Rudy Torrijos and Derek Hernandez, the current pessimism regarding AI disruption largely ignores significant structural moats that define enterprise software. Their far-reaching analysis outlines risks for the credit world, and what software investors can do to mitigate those risks. A few of the many key findings in the analysis:
- Replacing a core SaaS platform effectively is open-heart surgery for an enterprise, entailing immense operational friction. It’s far easier to add an AI ‘co-pilot’ to existing operations than to fully migrate to an AI startup.
- While AI/LLMs can be powerful, they’re useless without access to customer history, inventory, or codebases that sit securely within existing SaaS platforms, giving incumbent enterprises a huge leg up.
- While AI fears might be overblown, credit investors have reasons to worry, including a fair number of pre-rate hike, 2020-2022 vintage SaaS deals completed at high valuations. Credit investors must move from passive monitoring of these deals to active operational restructuring with sponsors.
For more information and a deeper dive into the sector, see “Is AI’s threat to software overblown? PitchBook analysis” and PitchBook’s recent report Clash of the Titans: Incumbents and Challengers in the Age of Agentic AI and “Asset value drop presents buying opportunities, but who will lend to software?”
Featured image by Andriy Onufriyenko/Getty Images
Learn about our editorial standards.