Software has been one of private equity’s most active and crowded sectors over the past decade, underpinned by a value creation model built on assumptions of durable revenue growth, plentiful leverage and high multiples at exit. Indeed, software transactions have accounted for approximately 40% of deal flow in recent years^[1], reflecting expectations that are now being questioned.

Key Takeaways:

• Software valuations and growth assumptions are being reset as higher rates and AI challenge long-standing private equity underwriting models.
• A significant portion of prior value creation in software was driven by multiple expansion rather than operational improvement.
• Realized outcomes, including DPI, may diverge from prior marks as exits test valuation assumptions and fund rankings.
• The next cycle may favor disciplined underwriting, cash flow generation and assets with more durable and controllable value drivers.

The current reset is not a dismissal of software’s long-term relevance, but a reassessment of business models, valuation and financing — namely the assumptions regarding growth durability, capital structures and exit multiples in a world shaped by higher rates and accelerating AI disruption.

Against a backdrop of crowding, high valuations and high leverage, we have maintained zero exposure to growth software in our last two private equity vintages. In any environment, we believe underwriting discipline — particularly around growth durability, pricing power and capital structure — is critical. Our position reflects a deliberate risk management decision grounded in downside protection, balance sheet resilience, controllable operational value creation and a strict policy of diversification.

Initially, private equity’s involvement in software was primarily through growth equity investments. This was before large-scale buyouts became common, and these were typically minority investments with little, if any, leverage. Over the past decade, the opportunity set evolved. As recurring revenue models of software companies scaled and margins expanded, software deals increasingly moved from growth equity portfolios into leveraged buyout strategies. The appeal of predictable recurring revenue, mission-critical applications, rising user growth, high customer retention and operating leverage supported higher purchase multiples, value creation opportunities and greater use of leverage.

But public software valuations have corrected sharply, with the S&P North American Technology Software Index falling as much as 35% from its September 2025 peak. In private markets, valuation adjustments tend to occur through new transactions, exits and comparable deal activity, meaning the full impact will emerge over time. That said, with mostly unlevered software companies having traded down by almost one-fifth, levered private software companies are expected to be marked down further while we await a final resolution.

The SaaSpocalypse — Or at Least a Structural Reset

The narrative that “software is eating the world” was directionally sound. But as in other cycles, a compelling trend did not mean correct price. 

A significant share of US buyout volume in recent years was concentrated in software, reflecting not only strong secular growth but also a structural shift within private equity. What was once the domain of sector specialists increasingly became a core focus for scaled generalist firms that built dedicated software capabilities. At the peak, growth software combined high purchase prices, elevated leverage and in many cases limited free cash flow. According to industry data, average purchase multiples approached 15–20x EBITDA with approximately 6x leverage and were often materially higher for premium assets.^[2]    

AI did not create the risk of structural disruption to software business models. It accelerated these dynamics by lowering barriers to entry, increasing competition and heightening uncertainty around pricing power and the durability of growth. In practical terms, this may manifest through increased pricing pressure, faster product replication and shorter innovation cycles. As AI reduces development costs and accelerates feature parity, differentiation may erode in more commoditized segments of software, compressing margins and making growth less durable than historical underwriting assumptions suggested. As a result and as we’ve noted, private equity is now confronting a reset in valuations as investors rethink the economic model and growth durability of software businesses.

Software moats are not disappearing — they are becoming more permeable and less predictable. The old assumption that seats, annual recurring revenue and multiples would expand predictably is being challenged. This is not uniform across the sector. Mission-critical systems of record, deeply embedded workflows and core infrastructure software may retain durability, while more tool-like or workflow-adjacent applications face greater risk of disruption as AI lowers barriers to entry. Meanwhile adoption is likely to be uneven and slower in enterprise environments, where reliability, integration and compliance requirements can delay disruption, even if long-term competitive dynamics are shifting — questions and uncertainty have now replaced stability and confidence, which is not conducive to strong valuations.

From Unlevered Growth to Leveraged Buyouts 

It is worth remembering that software was once largely an unlevered growth equity investment. Over time, it migrated into mainstream leveraged buyout portfolios. 

In 2021 alone, private equity deployed record capital into software — $348 billion globally — often at peak valuations.^[3] Capital structures increasingly incorporated elevated leverage and underwriting assumptions predicated on durable growth, stable cash flows and exit multiple resilience. When two of the three traditional value levers — growth and exit multiples — come under pressure, leverage magnifies the downside risk. The consequences extend beyond marks, influencing exit timing, holding periods and ultimately realized returns.

Software buyouts now represent a significant portion of industry deployment. In fact, approximately 85% of all technology deals from 2016 to 2025 in private equity have been software-related.^[4] This may underestimate the exposure, as many business services and healthcare deals, for example, are actually plays on software. Yet exits have not kept pace with deal activity, with deployment in recent years running at roughly five times exits,^[5] and recent returns were often based on marks rather than realized distributions, with initial underwriting in many cases implying return outcomes that have yet to be validated through actual exits. Ultimately, DPI reflects outcomes, not marks. As many pandemic-era investments approach the end of traditional holding periods, the question becomes: what happens when assets are sold? 

It is also fair to ask whether the quartile rankings of these largely unrealized funds will hold up, or whether they are likely to reset lower as holding periods elongate. The coming quarters will test whether current valuations have fully incorporated public market resets, particularly given that private marks are often anchored to lagging transactions, and whether realized DPI ultimately supports prior paper gains. 

DPI, Marks and Performance Rankings 

In periods of strong multiple expansion, value creation is often attributed to market sentiment. But sentiment-driven value can prove fragile and slow to convert into realized outcomes. In software specifically, a significant portion of value creation in the decade leading up to 2023 was driven by multiple expansion, much of it unrealized, rather than operational improvement.^[6] As public multiples reset, buyers and sellers have faced mismatched expectations, contributing to slower exit activity and extended hold periods. While signs of renewed M&A activity are emerging, the durability of distributions will ultimately determine whether prior paper gains translate into realized returns.

When multiple expansion reverses, marks become more difficult to defend. This raises two structural questions: 

1. How will realized outcomes compare with current carrying values in levered software portfolios? 
2. If exits occur below prior carrying values, how will that affect DPI and quartile rankings? 

Quartile rankings based on unrealized marks during a period of peak multiples may look different after realizations. In dislocated markets, realized outcomes matter more than modeled IRRs. 

Renewed Demand for HALO 

As the software cycle resets, capital may be redirected toward businesses with hard assets, lower obsolescence risk (“HALO” assets) and controllable value creation. In contrast to asset-light models facing disruption risk, HALO assets increasingly represent the physical and balance sheet foundation required to sustain the next wave of AI growth — where AI is a tool, not a disruptor. These HALO assets span our core sectors, including technology, leisure, industrials, business services, transportation and logistics, as well as media and gaming. As trillions of dollars flow into digital infrastructure, power and transmission capacity, the value of durable, capital-intensive assets with long asset lives and stable cash flow profiles is likely to rise. In that sense, HALO is not simply defensive positioning — it is participation in the infrastructure backbone of what we call the Global Industrial Renaissance.

In our view, the next cycle is likely to reward: 

• Profitable revenue growth over growth at any price 
• Resilience over narrative 
• Cash flow generation over multiple expansion 

As markets continue to digest the long-term transformational impact of AI, the advantage may accrue not to those who own commoditized code, but to those who control scaled, real-economy assets that AI can materially enhance. Increasingly, software may be the tool, not the asset. The enduring winners may not be AI businesses, but rather businesses in the traditional economy that harness AI to drive durable efficiency gains.

What About Software Buyouts Now? 

Public software valuations have corrected, while capital structures in private markets remain largely intact. The open question is not simply whether software buyouts become attractive again at lower prices, but which business models can demonstrate durable competitive advantages in a world shaped by AI and evolving cost structures. As valuations reset, the opportunity may lie in distinguishing between models that were previously overvalued and those whose resilience may now be undervalued. This dislocation may also create selective opportunities for deleveraging transactions, where high-quality businesses are saddled with overlevered capital structures, and reducing leverage frees up capital to fund growth.

A reset creates opportunity. But discipline matters. 

History shows that trends look compelling on the way up as growth creates a narrative of invincibility: railroads, telecom, dot-com, the housing market, shale energy and SaaS. The pattern is familiar: correct thesis, wrong price and ultimately poor investment outcomes — as fundamentals reassert themselves. 

We believe private equity will adapt, as it has through prior cycles — by tightening underwriting, focusing on sustainable value creation and prioritizing realized value and diversification.

But this period may be remembered as a case study in sector concentration risk, leverage layered onto growth assumptions and the limits of multiple-driven value creation, as well as a lack of risk management, experience and expertise. 

Cycles ultimately test underwriting. In the years ahead, realized outcomes — not narratives — will determine who truly created value. 

---

[1] Sources: White & Case, Deutsche Bank

[2] Sources: PitchBook LCD as of 2025 and Bain & Co. as of 2023

[3] Source: Bloomberg compiled data

[4] Source: Bain & Company Global Private Equity Report 2026

[5] Note: For software deals and exits from 2020-2025. Source: PitchBook

[6] Bain & Company Global Private Equity Report 2025