๐ง๐ต๐ฒ ๐ฏ๐ถ๐ด๐ด๐ฒ๐๐ ๐๐ ๐ฟ๐ถ๐๐ธ ๐ถ๐ป ๐๐ฒ๐ฐ๐ต ๐บ๐ฎ๐ ๐๐ถ๐ ๐ผ๐ป ๐๐ผ๐ณ๐๐๐ฎ๐ฟ๐ฒ ๐ฏ๐ฎ๐น๐ฎ๐ป๐ฐ๐ฒ ๐๐ต๐ฒ๐ฒ๐๐. A massive debt wall is forming across the software sector. Between ๐ฎ๐ฌ๐ฎ๐ฒ ๐ฎ๐ป๐ฑ ๐ฎ๐ฌ๐ฎ๐ต, ๐ป๐ฒ๐ฎ๐ฟ๐น๐ $๐ญ๐ฌ๐ฌ๐ ๐ถ๐ป ๐๐ผ๐ณ๐๐๐ฎ๐ฟ๐ฒ ๐น๐ผ๐ฎ๐ป๐ ๐บ๐ฎ๐๐๐ฟ๐ฒ. ๐ฅถ The largest concentration arrives in ๐ฎ๐ฌ๐ฎ๐ด ๐๐ถ๐๐ต ๐ฟ๐ผ๐๐ด๐ต๐น๐ $๐ฐ๐ฌ๐ ๐ฐ๐ผ๐บ๐ถ๐ป๐ด ๐ฑ๐๐ฒ. The credit quality tells the real story. A large share of this debt sits deep in ๐, ๐-, ๐ฎ๐ป๐ฑ ๐๐๐ ๐๐ฒ๐ฟ๐ฟ๐ถ๐๐ผ๐ฟ๐, ๐๐ถ๐๐ต ๐ฎ๐ฟ๐ผ๐๐ป๐ฑ ๐ต๐ฎ๐น๐ณ ๐ฟ๐ฎ๐๐ฒ๐ฑ ๐- ๐ผ๐ฟ ๐น๐ผ๐๐ฒ๐ฟ. This structure made sense during the zero-interest era. Private credit funds financed software companies because the model looked structurally safe. Recurring subscriptions created predictable cash flows, and rising valuations supported aggressive lending. AI changes that equation. AI platforms compress entire software stacks. Tasks that once required multiple SaaS products increasingly run through a single AI layer. That shift creates a structural change. The number of software tools companies need may decline. For highly leveraged vendors, this creates a refinancing challenge. Debt raised during the growth cycle eventually demands real cash flow. This is where systemic risk appears. Software represents about ๐ญ๐ฎ% ๐ผ๐ณ ๐๐ต๐ฒ ๐น๐ฒ๐๐ฒ๐ฟ๐ฎ๐ด๐ฒ๐ฑ ๐น๐ผ๐ฎ๐ป ๐บ๐ฎ๐ฟ๐ธ๐ฒ๐, making it the largest single sector exposure. If AI accelerates consolidation across SaaS, the ripple effects extend far beyond Silicon Valley. India sits at the center of that ecosystem. Indiaโs $๐ฏ๐ฌ๐ฌ๐+ ๐๐ฒ๐ฐ๐ต๐ป๐ผ๐น๐ผ๐ด๐ ๐ถ๐ป๐ฑ๐๐๐๐ฟ๐, ๐ต๐ฒ๐ฎ๐ฑ๐ถ๐ป๐ด ๐๐ผ๐๐ฎ๐ฟ๐ฑ $๐ฏ๐ฑ๐ฌ๐, operates a large share of the worldโs enterprise software infrastructure. Any restructuring or consolidation across SaaS influences implementation spending, vendor ecosystems, and enterprise technology budgets. AI will create enormous productivity gains. Before that equilibrium forms, the software industry may pass through something else first. ๐ ๐ฏ๐ฎ๐น๐ฎ๐ป๐ฐ๐ฒ ๐๐ต๐ฒ๐ฒ๐ ๐ฟ๐ฒ๐๐ฒ๐ ๐ The intersection of ๐๐ ๐ฑ๐ถ๐๐ฟ๐๐ฝ๐๐ถ๐ผ๐ป ๐ฎ๐ป๐ฑ ๐น๐ฒ๐๐ฒ๐ฟ๐ฎ๐ด๐ฒ๐ฑ ๐ฐ๐ฎ๐ฝ๐ถ๐๐ฎ๐น ๐๐๐ฟ๐๐ฐ๐๐๐ฟ๐ฒ๐ could define the next phase of the global software market. ๐ฎ๐ฌ๐ฎ๐ด ๐บ๐ฎ๐ ๐ฏ๐ฒ๐ฐ๐ผ๐บ๐ฒ ๐๐ต๐ฒ ๐๐ฒ๐ฎ๐ฟ ๐๐ต๐ผ๐๐ฒ ๐๐๐ผ ๐ณ๐ผ๐ฟ๐ฐ๐ฒ๐ ๐ฐ๐ผ๐น๐น๐ถ๐ฑ๐ฒ.
Source: Federal Reserve Board (FEDS Notes)
Research Brief
What our analysis found
The tweet's central thesis โ that a massive wall of low-rated software debt is maturing into an era of AI-driven disruption โ is well-supported by major Wall Street research. Morgan Stanley estimates software accounts for roughly 16% of the $1.5 trillion U.S. leveraged loan market (~$235 billion), making it the single largest sector exposure. Critically, about half of that debt is rated B- or lower, according to both Morgan Stanley and JPMorgan analyses. JPMorgan has specifically flagged ~$51 billion in B- or lower software loans maturing in 2028 and another ~$50 billion in 2029, meaning the tweet's "nearly $100 billion" figure may actually be conservative when accounting for all rating tiers across the 2026โ2029 window.
The AI disruption angle is where the picture gets more nuanced. A sharp early-2026 selloff in software loans โ Voya reported the software segment returned -2.97% in January 2026 alone โ shows the market is already pricing in AI displacement risk. CLO managers are actively scrubbing portfolios for vulnerable software credits, and Blue Owl raised its withdrawal cap to 17% at a tech-focused fund amid elevated redemption requests. However, Gartner still forecasts worldwide software spending to rise ~14.7% in 2026 to approximately $1.43 trillion, and enterprise SaaS portfolio data from Zylo shows app counts holding roughly flat at ~305. The refinancing risk for low-rated software borrowers is real and documented; the scale and speed of AI-driven revenue erosion across the broader SaaS sector remains an open and hotly debated question.
The India connection is directionally accurate. NASSCOM pegs India's tech industry revenues at approximately $315 billion for FY26, with significant enterprise software implementation and support operations. Any restructuring wave across SaaS vendors would ripple into India's implementation spending, vendor partnerships, and enterprise technology budgets โ though India's diversification into AI services and Global Capability Centers may partially offset that exposure.
Fact Check
Evidence from both sides
Supporting Evidence
Software is the largest single sector in the U
S. leveraged loan market. PitchBook/LCD data from July 2025 shows a 12. 3% share; Morgan Stanley's February 2026 estimate puts it higher at ~16% of the ~$1. 5 trillion market (~$235 billion in software loans). Both confirm the tweet's "~12%" claim as accurate-to-conservative.
About half of software debt is rated B- or lower
Morgan Stanley explicitly states that roughly 50% of software loans carry B- or lower ratings, corroborated by JPMorgan's granular maturity analysis. S&P Global further notes software is the largest sector in both BSL CLOs (14. 7%) and middle-market CLOs (19. 2%).
The 2028 maturity concentration is confirmed by multiple sources
LPL Research shows ~25% of software debt matures in 2028, while JPMorgan flags ~$51 billion of B- or lower software loans coming due that year and another ~$50 billion in 2029. Morgan Stanley notes ~30% of software loans mature by 2028, versus ~22% for the broader market โ a meaningfully front-loaded schedule.
The market has already begun repricing software credit risk
Voya reported software loans returned -2. 97% in January 2026 with average bids falling 364 basis points. TwentyFour Asset Management documented meaningful CLO repricing as investors reassessed software exposure and the 2028 maturity wall.
Real-world liquidity stress has emerged
Blue Owl raised its withdrawal cap to 17% at its tech-focused fund amid redemption requests reaching ~15% of NAV, and subsequently sold ~$1. 4 billion of loans across its BDCs at ~99. 7% of par to meet payouts โ a concrete sign of investor anxiety around software credit.
JPMorgan has explicitly warned about AI risk inside CLO structures
Strategy notes flagged up to ~$150 billion of loans in CLOs as potentially exposed to AI-related disruption risk, with software as the primary sector of concern.
India's tech industry scale is confirmed
NASSCOM's latest figures peg India's technology industry at ~$315 billion in FY26, heading toward $350 billion, with substantial enterprise software implementation and support operations that would be affected by SaaS vendor consolidation.
Contradicting Evidence
Global software spending is still growing robustly
Gartner forecasts worldwide software spending to rise ~14. 7% in 2026 to approximately $1. 43 trillion, suggesting broad-based demand destruction is not yet materializing and many software companies โ including those embedding AI features โ continue to grow revenue.
Enterprise app counts have not collapsed
Zylo's 2024 SaaS Management Index shows enterprise software portfolios averaging ~305 applications with essentially flat counts heading into 2026, indicating that "stack compression" from AI remains more theoretical than observed at scale so far.
Private credit managers argue losses are contained
Blue Owl's leadership publicly stated that current market loss assumptions are "disconnected from portfolio math," pointing to low loan-to-value ratios (~30%) and equity value gains since origination as cushions against credit losses.
Default rates are not projected to spike
Fitch projected U. S. leveraged loan defaults to ease slightly by end-2026 to 4. 5โ5. 0%, suggesting a challenging but manageable environment rather than a systemic collapse scenario. BlackRock data shows only ~32% of all USD loans (across sectors) are rated B- or below.
CLO exposure to software is lower than the headline index weight
TwentyFour Asset Management notes average software exposure in U. S. CLOs is ~10% โ below the 12โ16% index weight โ while Barrow Hanley confirms BSL CLOs tend to underweight software relative to the broader loan index, mitigating systemic transmission risk.
AI may be a tailwind for select software companies, not a uniform threat
Morgan Stanley specifically argued AI is a tailwind for companies like Atlassian even amid the broader sector selloff, indicating a winner-loser dynamic rather than blanket SaaS destruction.
Amend-and-extend activity has already reshaped the maturity wall
PitchBook/LCD and other sources document record amend-and-extend activity in 2024โ2025, which has pushed many near-term maturities into later years. While this concentrates risk in 2028, it also means companies have already bought additional runway โ the wall is partly a product of proactive management, not neglect.