The initial sense of doom in private markets around the so-called SaaSpocalypse is giving way to a more stoic brand of realism. For investors seeking exits from existing software investments, there remains acknowledgement that some companies will be difficult to salvage, but the tone is more betterment than bloodbath.

Asked how artificial intelligence (AI) is affecting deal outcomes in software in terms of valuation, deal structuring and stalled processes, Benedikt Joeris, a partner at Hg Capital, evoked an increasingly prevalent mood of confidence and vigilance.

“We still believe that long-term value in software comes from earnings growth, not multiples, and we continue to exit above holding values,” he said.

“But it’s also fair to say that buyers are already behaving differently. There’s more scrutiny on whether a business has a real AI position, and the ones that can show actual products and adoption are getting rewarded for it.”

Two recent exit processes for Hg help illustrate the landscape. Last year, the private equity firm sold financial software provider GTreasury to digital assets player Ripple for USD 1bn, having acquired it two years earlier for a reported USD 400m. The result was largely attributed to an intervening buildout of agentic AI capabilities.

Meanwhile, longstanding plans for an IPO of Norway-based Visma remain in limbo. Hg delisted the cloud-based enterprise software provider from the Oslo Stock Exchange in 2006 and took control in a USD 5.3bn buyout in 2017 touted as the largest ever for a European software company.

Hg began exploring a public listing for Visma as early as August 2025 and by December that year had engaged more than 10 banks to manage the process, according to Mergermarket. Last March, Bloomberg reported an IPO was unlikely within the year, citing market volatility.

Hope springs

Still, there is conviction that public market appetite for private software businesses is set to rebound. Uday Karri, a vice president focused on private capital research at MSCI, describes a gathering wave of heavyweight AI-oriented tech IPOs as a kind of cyclical market reset.

He points to evidence in the form of MSCI data on the portion of venture-backed companies that are publicly listed. Only 6.7% of the companies MSCI tracks globally were listed at the start of 2026, the lowest rate in decades. That figure could correct to the historical range around 18% in less than a year based on regulatory filings by tech companies.

“One of the biggest tailwinds for the year is the IPO market opening up. Venture is at a level of concentration that we haven’t seen since the large IPOs of Facebook, Amazon, and Yahoo. Now we’ve got companies like SpaceX, Anthropic, and OpenAI either going public or waiting in the wings,” Karri said.

“That’s contributing to concentration in venture capital, which is driving the performance. Those are mostly paper marks for now, but there is an impending IPO floodgate that promises some liquidity and distributions in the coming quarters.”

On the M&A front, exit hopes stem at least partially from the emergence of rollup strategies, whereby at-risk software companies are acquired simply to absorb their proprietary data, regulatory moats, or client networks.

Elizabeth Todd, a partner and co-lead of European private equity transactions at Ropes & Gray, is tracking this type of opportunism in the likes of Canadian legal software company Clio, which facilitated an exit for Oakley Capital last year by acquiring Spain-based vLex at a USD 1bn valuation.

“This is an exit channel to watch with some caution. Particularly for distressed SaaS companies, the valuation to be achieved is likely to be one that values client books and data rather than your actual revenue. That might not be a valuation that works for the current owners,” she said.

“There is also real integration risk in a rollup model. Acquisitions are relatively easy compared to integrating those acquisitions – that is the real challenge.”

Todd added that buyers are still willing to pay reasonable multiples for the right asset in the current market environment, albeit selectively. No one is buying software broadly on the dip. But this does not mean no one is willing to take significant risk.

Venture builders collecting at-risk software companies are the clearest case in point. They include General Catalyst, which aims to acquire software-adjacent businesses primarily for their embedded client relationships and data, and then layer AI on top. Likewise, San Francsico-based Enam Co describes its business model as buying services businesses in order to transform them with AI.

“That’s a very smart business model if you can do it. But you need expertise in both M&A and AI,” said Daisy Cai, a general partner at early to growth-stage technology investor B Capital. “That is a good exit channel. I’m already seeing these kinds of companies. They’re both financial and corporate buyers. They’re pretty active.”

Concentration effect

At first glance, software M&A appears to be holding up to AI disruption. Global deal activity amounts to around USD 623bn in 2026 to date, according to Mergermarket. That’s up 65% up versus the corresponding period in 2025 and the best half-year total since 1H21.

It has come with a significant concentration effect, however. There have been 3,345 transactions so far in 2026, an 11% decline versus the same period in the prior year.

The picture darkens further when the dataset is narrowed to sponsor exits. Transactions facilitating exits for private market funds totalled about USD 32bn globally in 1H26, down 69% year-on-year. The number of deals for the period fell 22% year-on-year to 111.

The clearest takeaway is that the market perceives safety in scale. Much of the oxygen in software M&A this year has been taken up by bumper rounds for OpenAI and Anthropic. As Bain & Company highlighted in its mid-2026 M&A outlook, this entails a “winner’s paradox,” where big bets are placed despite big uncertainty simply because it’s too risky to do nothing.

Some consensus has emerged in the recent term around the characteristics of software assets that are most likely to survive and thrive amidst the dislocation. Offerings must be mission critical. Business models must provide outcomes rather than tools. Proprietary data and control of workflows are essential. The challenge is that these traits can be difficult to identify in a rapidly evolving sector.

“You don’t know which software companies are more defensible than others,” said one global fund-of-funds investor. “For managers with positions in software, we would like them to develop an exit path. We are asking them about the exits. We haven’t seen a software company listed for a while. For trade sales, who will be the buyers?”

Secondaries buyers are increasingly watching this space, if not ready to bite. Uncertainty around AI is quickening deal flow in software in terms of potential investments but few are transacted, even as valuations decline. Marks for software buyout holdings globally were down 7.9% on average between 4Q25 and 1Q26, according to Bain & Company. This compares to a 0.3% decline for all other sectors.

“Deals could not get done off 4Q25 marks because of the substantial valuation reset that had occurred,” said Jeff Hammer, a managing director focused on secondaries in Moelis & Company’s private capital advisory group.

“Valuations moved downward in 1H26, but there is still a large bid-ask spread that needs to close before meaningful activity resumes. Structures such as earn-outs, deferrals, and risk-sharing waterfalls can help.”

The right price?

Yaron Zafir, head of secondaries at Asante, is tracking a flight to perceived quality, resulting in competitive auction processes driving up prices. The innate conservatism of secondary buyers is being tempered by the emergence of groups with specialist expertise in software.

“[They] are more comfortable with these valuations, particularly GPs that have set up secondary programs to back other GPs in continuation vehicles (CVs),” he said.

Although seen as well suited to software industry characteristics such as recurring revenue and rapid growth, CVs have proven challenging because LPs remain reluctant to transact at current valuations.

Mike Bego, managing partner at secondaries investor Kline Hill Partners, said software represented as much as 25% of his firm’s CV pipeline prior to February 2026 but is now far less prevalent.

“Buyers and sellers are far apart on price. Those few that are getting done are bridging price differences with structure – earnouts, deferred consideration,” Bego said. “We are seeing some venture-growth multi-asset deals getting done, where there’s a clear AI opportunity concentrated in a handful of core positions.”

It contributes to a sense that options in terms of exit channels and corporate lifelines are generally dwindling. Refinancing, for example, is widely framed as a shrinking possibility for many software companies due to increasing borrowing costs related to AI uncertainty.

Ropes & Gray’s Todd observed that over the next 12-18 months, discounts may be used to push deals through, but for the time being, managers don’t yet feel enough urgency to sell at lower prices.

“For weaker software assets, the more likely outcome is a clean exit at a lower valuation rather than putting them into a CV,” she said. “For the most impaired assets, restructuring – including debt for equity swaps – is increasingly the realistic outcome.”

Some owners are beginning to sell off assets just to cut their losses. Sascha Pfeiffer, Houlihan Lokey’s global head of technology, noted that investments from funds raised prior to 2020 are most likely to be put on the block simply because managers are under so much pressure to generate distributions from earlier vintages.

Local nuance

Outside of a smattering of high-profile meltdowns in the US, Bain & Company, isn’t tracking much of this kind of activity globally. Sebastien Lamy, a partner at the firm, attributes the slowdown in sale processes as much to the work required to reposition software companies for the AI era as general buyer hesitancy. “We haven’t reached that stage of picking losers yet. I’m not sure we will,” he said.

Lamy sees more risk in not addressing the sense of uncertainty among software buyers. Key pitfalls include failure to craft an AI narrative with proof points that demonstrate real progress toward transformation. Waiting for industry uncertainty to diminish before acting can be equally damaging.

“Usually by year five, six, or seven of the holding period, we’re seeing a lot of work. The risk is taking time for an asset that is in year three or year four because the market may not be that punitive and you can rely on your traditional moat. That’s dithering,” he said.

Massive geographic discrepancies in the performance of private software investments in the recent term further complicate the AI disruption narrative.

Software buyout losses averaged 7.9% during 1Q26 in the US versus 3.6% in Europe and 0.6% in Asia, according to MSCI. Ex-software investments were much closer across regions, posting an average return of 1% in the US, a 0.3% return in Europe, and a 0.3% loss in Asia.

The gap in venture returns was even more conspicuous, with software in the US returning 3% in 1Q26 versus 2.6% in Europe and 20.9% in Asia. Again, ex-software venture returns were more comparable across regions, with US, Europe, and Asia returns coming in at 4.9%, 4.4%, and 1.1%, respectively.

Anticipating criticism that the outperformance of Asia might be attributable to an incomparably small sample size, MSCI’s Karri described the number of funds and companies as healthy enough to be comfortable with the figures.

Most importantly, it speaks to the importance of companies’ entry valuations in the current environment. Private software assets in the US and Europe today were largely acquired during the high-growth, high-valuation period of 2020-2022. The AI-related markdowns from these relative highs create significant drag on overall returns.

Karri observed that the effect is muted in Asia in part because software buyout deal flow did not begin to accelerate until after 2022. Likewise in venture, Asia’s relatively higher concentration of post-pandemic vintages reflects a more prudent phase of the industry in terms of entry valuations and due diligence on business fundamentals. This appears to be smoothening software’s AI transformation.

“A lot of the software red flags in buyouts are green flags in VC just due to the nature of those companies being more AI native and thus benefiting from positive marks,” he said. “The performance of Asia VC is benefiting from that AI tailwind while also not being as hurt by the SaaSpocalypse headlines causing pain ex-Asia.”

Finding value

Bain & Company’s Lamy noted that Asia’s relative resilience to AI disruption could create contrarian arbitrage opportunities, especially given that software selloffs on public markets in the region have been comparable to those in the US and Europe. Other industry participants are circumspect on the notion of bargain hunting in Asia because irrationally spooked asset owners are venting good companies.

Paul Robine, founder of Asia-focused direct secondaries investor TR Capital, argues that high quality companies seldom trade at big discounts. The key is acquiring them as part of wider portfolios that are priced more attractively.

TR demonstrated this approach last year, picking up a position in MoEngage, a hotly pursued Indian customer engagement software platform, that was bundled with two other businesses. The deal allowed Eight Roads to partially exit three mature investments.

“Finding value is always difficult but that’s our job,” Robine added, stressing the need for forward-looking AI diligence in software. “Bargains sometimes happen but remain rare and they are not always good ones. It’s not a function of the price – it’s deeper. It’s the quality of the business and the ability to exit it within 3-5 years.”

It remains to be seen if strong investment performance in Asian software signals a reanimation the region’s most inert exit markets. In Southeast Asia, Jeremy Tan, co-founder of enterprise software-focused Tin Men Capital, believes the pressures driving corporate AI adoption and tech market consolidation will support sales to strategics.

In this view, the reasons for buying are familiar: new geography, new technology, complementary product lines. The key difference in the AI era is urgency.

“In 2026, that calculation increasingly favours buying over building internally. It’s faster to acquire an AI-native team than to build it,” Tan said.

“The buyer universe has gotten wider, and for Southeast Asian B2B software companies specifically, this is a meaningful shift. Companies in this region that have built deep vertical workflows, own real data, and have shown they can operate across borders are now on the radar of acquirers who would not have looked at Southeast Asia five years ago.”