Navigating Software Exposure in Credit Markets

As Artificial Intelligence continues to disrupt traditional business models, investors have shifted their attention from equity market volatility to the software sector's footprint within credit markets. Morgan Stanley’s Vishy Tirupathor and Vishwas Parker recently examined how this exposure manifests, the unique vulnerabilities of the sector, and whether these risks pose a systemic threat to the broader financial system.

The Anatomy of Software Exposure in Credit

While software is a dominant force in equity markets, its presence in credit markets is characterized by its opacity and concentration in non-bank channels. According to Vishwas Parker, the software exposure in credit is "very different from what it is in equities." A significant portion of this market consists of private issuers, with approximately 80% of companies in the sample set being private.

This exposure is heavily represented in less liquid segments, specifically:

• BDCs (Business Development Corporations): Roughly 25% of their portfolios are software-focused.
• Private Credit CLOs: Closely follow BDCs in concentration.
• Leveraged Loan Market: Approximately 16% of the market is tied to software.

Credit Quality and Structural Vulnerabilities

The software sector’s rapid growth, particularly during the LBO (Leveraged Buyout) wave of 2020 and 2021, has left a legacy of weaker credit quality. Parker highlights that "about 50% of borrowers in the sector are rated B- or lower," placing them at the "lowest rungs of the rating spectrum."

These deals were frequently underwritten with higher leverage than the broader market. Consequently, the sector faces "more front-loaded maturities," which creates significant refinancing risks if AI-driven disruption persists and impacts the underlying companies' ability to service their debt.

The Challenge of Assessing Private Credit Risks

BDCs, which serve as the "public face of private credit," present a unique analytical challenge. Because many of their underlying holdings are private, they lack the rigorous reporting obligations of public companies. As Vishy Tirupathor notes, there are "no earnings reports, no 10-K or Qs, or broadly publicly available financials to look at."

This necessitates a "re-underwriting" of these companies to distinguish between those facing disruption from AI and those that might benefit from margin expansion. Until these outcomes are fully resolved, Tirupathor expects "credit spreads of BDCs to remain volatile," suggesting that current liability spreads, while widened, may not yet fully reflect the necessary clearing levels to account for these risks.

Is This a Systemic Threat?

Despite the clear risks, both strategists conclude that this is not a systemic issue.

Why the Risk is Contained

• Leverage Levels: Tirupathor points out that the leverage in BDCs is fairly small—roughly 2x—which is "orders of magnitude smaller" than the leverage seen in the financial system prior to the 2008 financial crisis.
• Banking System Linkage: While there is a connection to the banking system via back-leverage provided to non-bank lenders, this leverage is "substantially risk remote with very high subordination levels."
• Restrained Credit Cycle: Parker notes that historical cycles characterized by systemic risk were driven by aggressive corporate re-leveraging. In contrast, the current cycle has been "fairly restrained," with corporate debt-to-GDP ratios declining over the last five years and M&A activity remaining below trend.

Conclusion

The consensus from Morgan Stanley’s analysts is that while the software sector represents a significant risk due to its concentration, high leverage, and low credit ratings, it lacks the characteristics of a systemic risk. Investors should expect continued volatility and a necessary "valuation reset" in credit spreads as the market reconciles the impact of AI disruption, but the overall strength of corporate fundamentals suggests that contagion remains unlikely.