The steep selloff in software stocks is spreading to the debt market.
Pressured by growing worries about the disruptive potential of new AI coding tools, shares of large software companies such as Salesforce and ServiceNow have been sliding for months. But now the prices of software-company bonds and loans are also dropping.
That expanding pain is worrying many on Wall Street, because software has come to assume an outsize presence in the corporate-debt market—the result of a wave of private-equity buyouts that stretched from the late 2010s through the early 2020s. A downturn in the sector has the potential to drag down other areas of the market, cooling what has been a humming credit engine.
Software currently makes up 13% of the Morningstar LSTA U.S. Leveraged Loan Index—which tracks speculative-grade loans that are originated by banks and broadly distributed to investors—more than double the share of the next largest sector. The sector makes up an even larger percentage of private-credit loans made by asset managers directly to companies, with estimates putting the share at around 20% to a third of those loans.
Since mid-January, the prices of loans issued by the likes of Cloudera, a data analytics company, and Qlik, the maker of business intelligence software, have sunk by around 10 cents on the dollar or more, according to S&P Global Market Intelligence. Overall, the average price of software company loans in the Morningstar LSTA index has dropped to 90.51 cents on the dollar, as of Wednesday, from 94.71 cents at the end of last year, according to PitchBook LCD.
Many software-company loans will mature in just a few years, and investors and analysts say there is little indication that AI is hurting the companies right now. Still, businesses typically repay maturing loans with money raised from new loans.
That means if investors are holding a loan that matures in 2028, they have to be confident now that there will be a new lender at that time “that is confident in at least another five years,” said Michael Anderson, global head of credit strategy at Citigroup.
Right now, the threat posed by AI is vague. But investors have been alarmed by the rapid progress of Anthropic and others in developing tools that make it much easier to write software and streamline complex tasks for professionals. Those worries grew more acute this week with the release of a suite of new tools which can perform industry-specific functions such as reviewing legal contracts.
Though such advances could potentially help some software companies increase profit margins, investors still worry they could also put some out of business, as clients develop software themselves or use the plug-ins.
Andy Sieurin, a leveraged-finance analyst at Polen Capital, said debt investors are currently scouring the market for software companies that might be particularly vulnerable to AI disruption.
Those include companies that make “horizontal” software serving an array of businesses rather than a particular industry; companies that analyze publicly available data rather than their own proprietary data set; and companies that have particularly large debt loads relative to their earnings.
For loan prices to recover, investors will need to see not only good results but “almost an acceleration in beating expectations” when companies report their next quarterly earnings, Sieurin said.
The challenge, he added, is that it is almost “impossible to disprove a negative,” or convince investors that “AI isn’t going to disrupt this business.”
It took Wall Street a long time to embrace software debt in the first place.
Throughout the 1990s and early 2000s, debt investors largely avoided tech companies because they thought of them as either home runs or strikeouts. That meant they could be good for shareholders, who could reap the rewards from the successes, but bad for lenders, whose best-case scenario was regular coupon payments, and who ran the risk of significant losses in bankruptcies.
Attitudes shifted in the 2010s, with the growth of high-speed internet, cloud computing and subscription-as-a-service business models, which offered the promise of predictable revenue streams. Investors came to view the sector as so stable that they were willing to fund private-equity buyouts that loaded companies up with more debt than other types of businesses.
Even before AI worries emerged, some software companies had started to look more vulnerable, thanks in part to the 2022 jump in interest rates and rising competitive pressures.
Based on signs of stress among individual companies, KBRA DLD, which maintains a database of direct loans held by publicly traded asset managers, has forecast that the default rate for private-credit software loans will rise to 2.5% this year—above the 2% rate for all direct loans.
Some analysts say there is at least some potential for trouble in the software sector to spill into other areas of the loan market.
Citi’s Anderson noted that the software sector is even more represented in a Morningstar LSTA index of larger, more liquid loans than it is in the broader loan index, something that could spook individual investors and lead to heavier outflows from loan exchange-traded funds.
Worries about potential credit downgrades could also cool demand from managers of collateralized loan obligation funds—debt-funded vehicles that buy a significant share of speculative-grade business loans.
“There are potential transmission mechanisms” from software to other sectors, Anderson said.
Write to Sam Goldfarb at sam.goldfarb@wsj.com