The future isn’t what it used to be

It’s an amazing time to be alive. The advent of Artificial Intelligence (“AI”) has fueled a frenzy of capital spending on AI model development and infrastructure and spurred boundless speculation about how the world will dramatically change our professional and personal lives in just the next few years. Investors are busy trying to determine how staggering amounts of capital investment will translate into durable earnings power and how today’s companies will adapt to an AI revolution. Investment success will hinge on the ability to identify which businesses will experience tailwinds, which will have to navigate serious headwinds, and which will simply be displaced altogether. Time will tell, but, in some pockets of the loan market, time is shorter than some would like.

In early 2026, as the world began to better understand the power of AI, trading prices on software loans dropped precipitously and implied software spreads widened dramatically. What had long been considered one of the safest credit sectors for investment was turned on its head virtually overnight. Risk markets broadly questioned the terminal value of the sector. Software spreads now sit at 783bps relative to an index spread of 467bps and an index spread ex-Software of 419bps.

Figure 1

While private equity investors, corporate issuers and credit investors may prefer more time to understand the real impact of AI on the sector and the economy, looming loan maturities are going to force an earlier discussion. Between 2021 and 2023, of the $257bn of primary leveraged loan issuance supporting new leveraged buyouts, 24% was in the tech sector.¹ That debt is now coming due. Over $50bn of syndicated software loans mature in each of 2028 and 2029. Given it is highly unusual for debt to be allowed to come current (i.e., remain outstanding less than a year from maturity), those maturities will need to be addressed, or not, in 2H2026, 2027 and 2028. Approximately 85% of that debt is rated B- or lower at the rating agencies.

Figure 2

More alarmingly, software term loans rated B- or lower maturing in 2028-2029 currently trade at a weighted (by principal outstanding) average dollar price of ~79, with roughly half of those loans trading below 90.²

Figure 3

Software companies facing maturity pressure have several tools at their disposal. Upfront fees and spread increases to enhance returns for rolling lenders are standard, but higher cash interest burdens for highly leveraged companies with tight liquidity and an uncertain future may not be the most pragmatic solution. Additional pay-in-kind (PIK) coupons are a possible enhancement but only to the extent valued by lenders. The most powerful tool is meaningful risk reduction to rolling lenders through a repayment (at par or otherwise) from the issuance or contribution of new capital. Borrowers who can convince credit investors their business can thrive rather than be destroyed by AI adoption can pursue a high yield bond issuance or a private capital solution funding on a pari passu basis. Historically, surges in secured HY issuance have tracked with elevated loan extension activity. This cohort will also be fertile ground for opportunistic credit investors willing to invest in highly structured credit tranches on a junior basis for equity-like returns. Borrowers with loans in the 90s clearly have the easier path to either new capital solution.

Figure 4

The credit issue for software, however, is that the nature of distress here is secular, and somewhat existential, rather than cyclical. For cyclical businesses, time is typically in every investor’s interest. Existing lenders want more economics for the risk but ultimately want to allow cyclical recovery to enhance the value of their underlying collateral. New investors are attracted by the opportunity to lend new money at high interest rates with fulsome call protection to enjoy the upside of convexity once the industry recovery commences and the equity obviously wants time to keep alive the optionality of realizing a return. In the context of secular distress and high uncertainty about terminal value, time can be the enemy if the pace of financial degradation in the underlying business is high. In many instances, lenders will want more than the happy talk about a more confident future…they will likely want risk reduction through paydown, and the happy talk backed up by new equity to fund that paydown. While the market is not treating all software businesses equally, the challenge for all constituents is that the distinction between winners and losers is highly uncertain and may not become clear until well after decisions around extensions and related investments need to be made. For companies with debt trading at deep discounts, more aggressive liability management tactics and coercive structures will no doubt further complicate investment decisions for lenders. It’s about to get interesting.

^{1. JPM Credit Research.}

^{2. Morningstar, Morgan Stanley Research as of 6.21.26.}