Private equity in transition: from distribution drought to selective recovery

• Private equity is at a critical inflection point. Exit activity rebounded strongly in 2025, with global deal value reaching USD905bn. But 78% was concentrated in mega exits, leaving mid-market inventory effectively stagnant. Until liquidity broadens beyond the top tier, the normalization of distribution remains structurally incomplete.
• Three compounding pressures will define 2026. Software bifurcation is accelerating, private credit refinancing is narrowing and Limited Partner (LP) re-up decisions are arriving simultaneously. The central question is whether sponsors can convert portfolio value into realized distributions fast enough to preserve fundraising capacity, or whether synthetic liquidity becomes the default rather than the exception.
• Software is undergoing a transformation, not extinction. The winners will be defined by two criteria: proprietary data that large language models cannot replicate and mission-critical workflows with demonstrable automation value. The losers will probably be seat-based models and point solutions with replicable functionality. For 2020-2021 vintage portfolios, the composition of software exposure across these two categories will be the primary determinant of exit feasibility and valuation outcomes.
• Private credit stress is closing the refinancing channel that has underpinned distributions since 2023. Approximately 40% of private-credit borrowers now carry negative free cash flow, PIK usage is rising even as headline spreads compress and 46% of outstanding software loans mature within four years. The mechanism is direct: as lenders reassess business model durability, refinancing becomes selective, buyer leverage financing tightens, the pool of viable acquirers shrinks and exit clearing prices fall, precisely for the assets most dependent on credit availability to transact.
• The macroeconomic outlook for 2026–2027 is supportive for distributions. Easing financial conditions, resilient growth, and improving public market valuations create favorable exit conditions. However, the intersection of software transformation and private credit stress introduces material downside risk to distribution forecasts. The industry that emerges from this transition will be smaller, more specialized, more operationally focused, and more realistic about the permanence of technology-driven disruption to business models.
• Our PE distribution forecasts suggest timid positivity but a wide range of possible outcomes. Our proprietary model projects a 5pps improvement in distribution rates for 2026 under baseline assumptions, reaching 17-19% yields. However, outcomes range dramatically, from an improvement of 8pps in an upside scenario with successful software adaptation and credit resilience, to a decline of 3pps in a downside scenario with severe disruption and credit stress that extends the drought through 2028.

From IRR to DPI, following three years of elevated interest rates and persistent valuation challenges, the private equity contract has undergone significant structural changes. Currently, limited partners face multiple compounding challenges, including negative net cash flows that constrain new commitments, heightened scrutiny of performance attribution, as new benchmarking methodologies proliferate, and persistent portfolio rebalancing pressures despite improved market conditions. This urgency has manifested in concrete behavioral shifts, with limited partners reducing general partner relationships by an estimated 20-30% on average . They are concentrating capital with managers who have demonstrated consistent distribution discipline and credible monetization pathways. Re-ups are no longer automatic, and managers must actively earn their place in portfolios, where cash velocity has replaced growth narratives as the primary selection criterion. The acute denominator effect that characterized 2022-2023, when public-market declines mechanically pushed private-equity allocations above target, has largely faded with equity-market rebounds. However, portfolio constraints persist, as major US public pension plans are actively reducing private-equity exposure through allocation target cuts. Meanwhile, persistent exit backlogs and capital calls exceeding distributions effectively close many investors to all but their highest conviction relationships.

The challenges are particularly acute for sponsors overseeing portfolios from 2020-2021 vintages. In these cases, funds allocated capital at peak valuations with underwriting assumptions built around five- to six-year hold periods, implying natural exit windows concentrated in 2025-2027. However, the intersection of persistent bid-ask spreads, selective buyer demand and emerging sector-specific risks, particularly in technology and software, where business-model disruption has compressed valuations, now threatens to extend these timelines precisely when limited partner impatience is at its peak. This shift does not appear to be a mere cyclical phase that will revert automatically with lower rates. Rather, it reflects a structural recalibration of limited partner expectations, as evidenced by the shift in performance metrics. 21% of limited partners now identify distributions to paid-in capital as the most critical measure. This is a 13pps increase from just three years ago, when only 8% prioritized this metric. Meanwhile, the share ranking internal rate of return (IRR) as most critical declined from 42% to 35%. In this environment, the ability to generate distributions at scale has become critical not only for current fund performance optics but also for future fundraising capacity. The industry has entered a phase where execution quality and distribution discipline function as existential differentiators rather than marginal performance factors (Table 1).

Exit markets are reopening, but the recovery remains selective rather than broad-based. Following two subdued years, 2025 marked a genuine inflection point, with private equity exit value rebounding sharply, particularly in the US, where exits reached an estimated USD728bn, the second strongest year on record after 2021. This validates the view that exit conditions are improving and that the market is functioning again for transactions of sufficient scale and quality. However, the headline figure hides a structural reality: Liquidity remains concentrated in the upper parts of assets. Mega exits accounted for approximately 78% of the total exit value in 2025, indicating that the market is open for large, high-quality companies, while a significant portion of the broader private-equity inventory remains effectively stagnant. This dispersion directly impacts distribution forecasts. The capacity to eliminate the exit backlog is contingent not only on aggregate exit volumes, but also on the ability to identify buyers for mid-market and lower-quality assets at valuations that sponsors and LPs are willing to accept. The concentration of liquidity indicates that while the top tier of the market has reopened, full normalization, where assets across the quality spectrum can exit efficiently, remains elusive (Figure 1).

Strategic and sponsor buyers have regained confidence, supported by easing financing conditions and improved visibility on macroeconomic risks. Corporate exits were particularly strong, reaching approximately USD299bn in 2025, a record level that underscores how trade buyers have returned as credible exit partners after sitting on the sidelines in 2022-2023. Sponsor-to-sponsor activity has also been constructive, acting as a liquidity bridge for older vintages under pressure. Well-capitalized funds have been willing to pay for quality assets, driving the median sponsor-to-sponsor exit size to record levels and providing an important release valve for managers needing to demonstrate distributions. However, the most significant missing piece remains initial public offerings. While US listings showed improvements in absolute terms compared to 2023-2024, the channel remains fragile relative to historical norms. Meanwhile, IPO activity in Europe showed further deterioration. This suggests that IPO markets may resume in the future, although likely at a volume that is structurally lower than in previous cycles. The lack of a robust IPO window presents a strategic challenge for sponsors. Without the option of an IPO, the available options are limited to strategic sales and transactions among sponsors, which can increase execution risk and potentially reduce valuations for assets that previously relied on public market exits (Figure 2).

In this context, alternative liquidity tools have evolved from opportunistic to essential. These include continuation vehicles, GP-led secondaries, net asset value financing and dividend recapitalizations. These mechanisms provide optionality and help avoid forced sales into weak markets, which is constructive. In 2025, continuation-vehicle activity reached record levels as sponsors sought to extend hold periods while still providing interim liquidity to LPs. However, it's important to note that these tools are best used as complements to traditional exit channels, rather than as substitutes. The industry's exit backlog is too substantial to be cleared through secondaries and engineered liquidity solutions alone. The restoration of distributions at scale ultimately depends on traditional channels, such as strategic mergers and acquisitions and a functional IPO market. Both of these channels need to broaden beyond the largest, highest quality deals to unlock a sustainable normalization of cash returns (Figure 3 – gap between the two lines indicates the estimated distributions impact of alternative liquidity tools estimated by Pitchbook).

The software sector has been a primary component of private-equity portfolios for over a decade, accounting for approximately 25% of total deal value over the past five years. This makes the current repricing unusually important for the broader private equity ecosystem. In early February 2026, public markets saw a significant decline in software market value, with losses exceeding USD1.2trn over a period of just five trading sessions, as investors rapidly reassessed the potential impact of AI agents on major segments of the traditional SaaS model. Software indices experienced a significant decline and remain well below their recent highs, while several individual companies have encountered more substantial drawdowns (Figure 4).

The timing is particularly problematic for private equity. The peak period for software deal activity was in 2021-22, when sponsors paid high valuations based on the assumption that recurring revenues, high customer retention and predictable growth would support premium exit outcomes. A significant number of these portfolios are currently facing a dual challenge, confronted with two adverse factors. First, public market comparables have weakened considerably, which mechanically resets exit benchmarks and increases the likelihood of valuation write-downs. Second, the market is increasingly questioning the continued structural defensibility of certain software business models in an AI-driven world. This combination is widening bid-ask spreads, reducing IPO optionality and delaying monetization timelines precisely when LPs are demanding cash returns. It also raises the risk that sponsors will be forced into suboptimal exit routes, continuation vehicles, GP-led secondaries or structured solutions, rather than traditional trade sales at premium multiples (Figure 5).

The software repricing carries significant risks, particularly when it intersects with leverage and the private-debt ecosystem. The vintage problem is particularly acute as current loan books are filled with software companies acquired at peak valuations of 15-20x revenue multiples with leverage ratios of 6-7x EBITDA, underwritten on assumptions of high customer retention and stable growth trajectories. If AI-driven disruption leads to even moderate deterioration in customer retention or pricing power, the loan-to-value (LTV) ratios and leverage maintenance covenants on these deals will be breached rapidly, forcing accelerated amortization or sponsor equity injections precisely when portfolio company valuations are declining. This results in a distinct risk category compared to cyclical downturns, characterized by uncertainty in business models rather than temporary market challenges. Credit markets have a proven track record of mitigating cyclical headwinds through restructurings and payment deferrals. However, during periods when lenders question the fundamental sustainability of revenue bases and competitive positioning, credit markets often face challenges. In instances where investors are unable to assess whether a business will maintain a recognizable form for the next five years, traditional credit workout mechanisms may prove to be inadequate. This transmission channel from equity volatility to credit markets extends beyond software to any sector where business model durability is being reassessed by technological disruption, creating correlated stress across portfolio exposures that were previously considered diversified (Figure 6).

For private equity portfolios, this intersection of software repricing and credit market sensitivity creates a second-order constraint on distribution capacity. Tighter credit conditions reduce not only refinancing flexibility but also exit liquidity, because buyers require leverage financing to transact. If lenders become more cautious about business model durability, whether in software or other sectors facing structural disruption, the pool of viable buyers shrinks, competition declines, and clearing prices fall. The question is not whether credit markets can enable distribution recovery, but for which assets and under what conditions.

However, and despite all this noise, the software sector is not dying but undergoing a fundamental structural transition where value shifts from selling tools to selling outcomes, potentially expanding total addressable markets for well-positioned incumbents as enterprises increasingly allocate budgets to automation as a direct substitute for labor costs. This transformation reflects the sector's historical pattern of anticipating and adapting to technology shifts. However, it will proceed unevenly and accelerate dispersion across portfolio holdings. The winners will be companies that control mission-critical workflows, own proprietary datasets that large language models cannot replicate and successfully embed artificial intelligence deeply into their platforms to deliver genuine automation value. Conversely, companies that cannot differentiate their offerings, rely heavily on seat-based pricing models vulnerable to displacement or provide functionality that can be replicated or bypassed by AI-native solutions face material disadvantage. For private-equity sponsors, this necessitates immediate portfolio triage and active repositioning, as the window to transform business models is compressing and the cost of inaction continues to escalate with each quarter of delayed decision-making.

The software industry made an early bet on AI technology. From its earliest days, the software industry has behaved as a core innovation engine, anticipating structural technology shifts and committing capital well before full commercial visibility emerges. Its early and aggressive embrace of artificial intelligence in the initial phase of the technology’s development cycle is consistent with that legacy. As illustrated by the steady rise in capital expenditure ratios since 2010, the sector has systematically reinvested to secure future growth. This is an industry forged in volatility and adaptation: it weathered the collapse of the dot-com bubble, restructured its models and ultimately reinvented itself. The now-dominant SaaS paradigm itself emerged from the ashes of the telecom bubble, proving the sector’s ability to transform crisis into structural opportunity. Today, AI deployment, particularly in large corporate environments, poses substitution risks for certain application software segments more exposed to automation and large-language-model competition, while system-level and multi-layer platforms appear relatively more resilient. Margins in selected niches may face pressure from intensifying competition with LLM and AI agent developers. Yet, several industry leaders have already begun repositioning, as reflected in the growing contribution of AI-related revenues in 2025-2026 income statements. So far, markets have warmly welcomed the software industry’s innovation-pioneer status, and notably the stronger AI footprint on business activity that comes along with rising influence across the stock equity market, which last autumn outpaced the peak seen before the dot-com bubble crisis in the 2000s (Figure 7).

The AI ticket is expensive and comes with higher leverage. This strategic leap into AI has come at a considerable financial cost. In anticipation of automation-driven obsolescence across parts of their service stack, software companies have accelerated investment dramatically: The sector’s capex ratio has nearly doubled in less than two years, peaking at 18% in Q1 2026 (Figure 8 & 9). This unusually aggressive investment cycle has been accompanied by a rapid drawdown of cash reserves and a marked increase in leverage, pushing aggregate net debt into positive territory for the first time. While leverage metrics remain manageable, with net debt representing roughly 11% of revenue and 35% of EBIT on a trailing twelve-month basis, the upward trajectory warrants close monitoring. The situation could become more delicate should investor confidence in AI weaken and capital rotate away from technology. Indeed, since last autumn, software equities have declined by roughly -25%, offering a first warning of the market’s capacity to swiftly reprice, or even penalize, the very sector it once enthusiastically championed.

Despite the bearishness in markets these are just throwing a warning call and are reminding us that rationality does not prevail on in markets.  The recent sell-off in software equities (, roughly 30% since last autumn), appears striking in absolute terms, yet but looks far less dramatic when placed in context. The correction represents about a 60% retracement of the April-October 2025 rally and only around 20% of the powerful upswing that began after the launch of ChatGPT in late 2022. In that sense, the move resembles a normalization phase more than a structural breakdown. Elevated valuations, stalling earnings revisions, and heightened geopolitical uncertainty have provided a convenient backdrop for profit-taking, especially after several years of outsized gains. The sector’s strong prior performance and growing index weight also made it an easy source of liquidity for investors seeking to de-risk portfolios. Importantly, the correction appears more technical than fundamentally driven, though fundamentals have been invoked to rationalize it. Software sits in an uncomfortable hybrid position, caught between two opposing tail risks: AI-driven substitution pressures on parts of the application layer, and the possibility of an AI investment bubble. This dual exposure makes the sector a natural “collateral asset” in periods of doubt about the trajectory of AI development, whether optimism fades or exuberance is questioned. Still, while some concerns are legitimate, the scale of pessimism and the risk premium now embedded in valuations seem excessive relative to the industry’s innovation track record. With a price-to-earnings ratio just above 30x, well below the ~50x peaks of recent years and compressed relative to its long-term z-score (current PER at 1.4x standard deviation below its 10-year average, see Figure x), the sector has already undergone a meaningful de-rating. For an industry still delivering revenue growth above 15% and operating margins north of 30%, current levels could increasingly be viewed as an attractive re-entry point, potentially tempering the downside momentum of the sell-off.

A key development supporting our baseline forecast is the potential for improved refinancing conditions to support distribution recovery. Spreads on PE-backed direct-lending loans in the US compressed to approximately 475-500bps over SOFR by late 2025. These are the tightest levels since 2007 and well below the 650-700bps that prevailed during the stressed environment of 2023, while a 100bps reduction in the federal funds rate lowered debt-service burdens and expanded refinancing feasibility. Lower debt-service costs free up cash flow that can be distributed to sponsors and ultimately to LPs through the distribution waterfall. Equally important, reduced financing costs for buyers, both strategic acquirers and financial sponsors, support higher exit valuations by improving returns at any given purchase price.

However, the refinancing dynamic is increasingly bifurcating by sector and vintage. While aggregate spread levels reflect optimism about default rate stability, the composition of stress reveals meaningful sector-specific deterioration already underway. Approximately 40% of private credit borrowers now carry negative free cash flow, up from 25% in 2021, and software companies are among the leading sectors of payment-in-kind (PIK) loan usage, a signal of compounding risk when valuation pressure meets refinancing constraints. BDC portfolios are exhibiting a simultaneous increase in loans with PIK features even as average spreads have compressed, highlighting a growing divergence between surface level pricing signals and underlying borrower cash-flow resilience. The restructuring cycle is active but operating through other mechanisms rather than headline defaults: 64% of defaults in H1 2025 were distressed exchanges, covenant resets or maturity extensions rather than traditional missed payments or bankruptcies. This pattern underscores why default rates alone are a misleading gauge of risk in software credit.

The upcoming maturity wall is expected to increasingly transform refinancing from a broadly available liquidity channel into a selective filter on quality. Roughly 30% of outstanding software loans mature by 2028, compared to 22% for the overall leveraged loan market, with 46% of software debt coming due within the next four years versus less than 35% for the broader universe. For portfolio companies approaching these maturities, refinancing outcomes will depend not on market-wide spread levels but on lender convictions regarding specific business-model durability in an AI-disrupted landscape. This creates a decisive divergence in credit access: companies with mission-critical workflows, proprietary datasets and credible AI-integration strategies will continue to refinance smoothly, while those relying on seat-based pricing or replicable point solutions face widening spreads, reduced leverage tolerance and stricter covenant packages, or, in more challenged cases, lender-driven restructurings that dilute equity and extend holding periods.

For PE portfolios, this divergence matters because credit availability directly determines exit liquidity. Buyers require leverage financing to transact, and if lenders become more selective about software exposures, the pool of viable acquirers shrinks, competitive tension declines and clearing prices fall. This dynamic is already observable: Strategic buyers have returned for high-quality software assets (contributing to the strong 2025 corporate exit activity), but mid-market and lower-quality software deals face extended marketing periods and increased reliance on seller financing or earnouts. Concurrently, a key distribution lever weakens. Since 2023, private credit has played a pivotal role in facilitating distributions through refinancings and strategic dividend recapitalizations. If this channel is closed for software-heavy portfolios, distribution recovery becomes slower and more dependent on traditional exits. In a worst-case scenario, this creates a destabilizing feedback loop. Weaker valuations reduce equity cushions, refinancing becomes more difficult, exits are delayed and distributions remain suppressed, exactly when LP pressure is at its peak.

Systemic concerns remain contained but require monitoring. Recent developments, Department of Justice warnings on portfolio marks, SEC investigations into private credit rating agencies, over-collateralization test failures at major CLOs and the aborted Blue Owl fund merger amid NAV-price mismatches signal that enforcement authorities and market participants are increasingly scrutinizing valuation governance and liquidity mismatches in private markets. The rapid growth of private credit from approximately USD800bn in 2020 to over USD2trn today has occurred largely outside traditional stress cycles, leaving many assumptions about recovery rates, workout processes and refinancing capacity during genuine distress essentially untested by actual market conditions. For portfolios with concentrated exposure to software assets financed through covenant-lite structures, this creates potential tail risk that could manifest more rapidly than baseline assumptions reflect if conditions deteriorate. Private credit recoveries in software average roughly one-third of par, well below broadly syndicated loan benchmarks, due to intangible heavy collateral bases, meaning even senior secured positioning provides less downside protection than in asset-intensive sectors.

The baseline view remains that private credit functions as a selective enabler. Constructive for the right assets, constrained for the wrong ones. Spread compression and lower policy rates create genuine tailwinds for high-quality portfolio companies with defensible business models and well-structured capital stacks. However, the distribution recovery will proceed unevenly, with software-heavy 2020–2021 vintages facing extended timelines, increased reliance on alternative liquidity solutions and heightened sensitivity to lender risk appetite. The key implication is that distribution outcomes will be determined less by market-wide refinancing conditions and more by the quality and durability of underlying portfolio exposures.

Private equity value creation has shifted decisively toward operational improvement as elevated entry multiples and higher interest rates have rendered traditional financial engineering insufficient. Entry-buyout multiples have moderated but remained elevated at approximately 12x in 2025, while top-quartile funds now derive approximately 40% of returns from revenue growth and margin expansion compared to 60% from multiple expansion and leverage, a reversal from prior decades. Given the heightened levels of multiple ratios and the increased cost of debt, sponsors must develop more robust business models to generate acceptable returns. This shift away from relying on leverage and valuation arbitrage indicates a need for more prudent financial strategies. The operational imperative is reflected in specific portfolio management requirements that go beyond the traditional financial sponsor capabilities. Holdings now require a thorough assessment of competitive positioning and the durability of the business model, especially in sectors undergoing rapid change or margin compression. According to the data, 46% of general partners anticipate that deals executed in 2025 will outperform those from 2021 to 2024. This expectation reflects the benefits of disciplined entry pricing and operational focus. However, achieving this outperformance requires execution capabilities that some sponsors still find challenging to develop, as holding periods now exceed six years on average. Valuation discipline has evolved from being a desired practice to being an essential part of doing business. This shift is driven by the recognition that markets are now able to differentiate between companies that have defensible advantages and those that face structural headwinds. As a result, the critical question has shifted from whether multiples are high in absolute terms to whether they are justified by fundamentals capable of supporting exit valuations in selective markets.

These execution requirements directly impact fundraising outcomes, as capital consolidates among managers demonstrating operational capabilities rather than financial engineering prowess. The fundraising market experienced a significant decline in 2025, with funds closing at an average discount of 19% compared to target figures and timelines extending to 20 months. The percentage of cases requiring over two years to close increased from 9% in 2019 to 38%, reflecting a heightened level of scrutiny as limited partners evaluate current capabilities to navigate operational transformation and position assets for exits in highly selective markets. The distribution imperative creates a tension in portfolio management that differs fundamentally from prior cycles. Managers must balance maximizing long-term value against demonstrating distribution discipline in real time. They must also hold assets for incremental upside, bearing reputational cost when delayed cash returns signal execution challenges to limited partners. The ability to convert portfolio value into realized distributions has become, and will likely continue to remain, the primary evidence of manager skill from a limited partner perspective. Success in this environment requires capabilities beyond traditional private equity skill sets, including operating partners with deep sector expertise, analytical capabilities to assess competitive positioning in rapidly evolving industries and teams capable of active transformation management rather than passive monitoring. The difference between managers who effectively use this operational playbook and those who rely on legacy approaches is likely to be wider than in the past. This will directly affect current fund performance and future fundraising capacity. For this reason, this execution dimension is a primary variable in distribution forecasting. The scenarios presented below explicitly incorporate assumptions around operational transformation challenges and distribution discipline.

The collapse in private equity distributions since 2022 is the clearest manifestation of the broader "private markets liquidity crisis." However, the underlying driver has been a mix of a generalized macroeconomic slowdown, increased geopolitical and trade uncertainty and more of a specific repricing of the IT private equity (PE) universe. As policy rates increased sharply, public technology valuations decreased, initial public offering (IPO) markets effectively closed and the valuation anchor for growth assets was diluted. Many 2020-21 vintages were concentrated in software and tech-enabled business models. Consequently, the industry shifted from a traditional exit-and-recycle cycle to a regime dominated by delayed realizations, continuation vehicles, and NAV-based liquidity solutions. Therefore, the key question for 2026-2028 is not whether distributions will mechanically recover, but whether the IT/software exit engine can be restarted in a sustainable way.

Our distribution forecasting framework flows from macro conditions to exit capacity. To evaluate the normalization process, we use a macro-led distribution forecasting model that connects market conditions directly to exit feasibility. This depends on four interlinked channels: public market valuations (the pricing anchor for private assets), financing conditions (the cost and availability of leverage), exit market depth (the feasibility of IPOs and M&As) and confidence and risk appetite. We standardize these inputs using a z-score approach, which allows us to benchmark the current environment against historical cycles and quantify how much each factor contributes to the overall distribution momentum. This framework is particularly well suited to the current cycle. Unlike in traditional downturns, when cash flows remain intact, the post-2022 shock has been driven by a valuation reset in the software sector, where exit outcomes are sensitive to public comparables and sentiment. Consequently, the distribution outlook hinges on whether the exit ecosystem, rather than the underlying portfolio companies, returns to functionality.

According to our framework, tracking the reopening of the public exit window will be crucial in 2026-2028. Since IPO markets serve as a direct exit route and a marginal price-setting mechanism for PE assets, we supplement the model with a dedicated IPO indicator based on the GS IPO Barometer. The barometer is constructed as a composite of key historical issuance activity determinants, including equity market performance, market volatility, valuation levels, credit spreads, financing availability, and broader macro-policy uncertainty. Conceptually, it is a nowcast measure of whether public markets are "open" or "closed" for issuance. The recent improvement in the barometer suggests that IPO issuance is transitioning from a post-2022 freeze toward a more functional environment. Although activity remains well below the 2021 peak, the upward trend is significant. In software-driven cycles, IPO normalization typically precedes a broader recovery in sponsor exits because it restores valuation confidence and provides an external reference point for private market pricing. In other words, IPO markets do not need to return to boom conditions to matter. What matters is that they reopen enough to reestablish credible price discovery and that the software/IT hiccup does not derail the recovery.

Geopolitical dynamics and trade uncertainty also present significant constraints on global economic activity. While the IPO barometer indicates improving technical conditions for issuance, our distribution decomposition framework highlights why the normalization path remains fragile. We use standardized z-scores to analyze historical distribution cycles and identify the contributions of core macro-financial drivers, including earnings growth, CEO confidence, inflation, interest rates, high-yield spreads and trade-policy uncertainty. This decomposition provides a more nuanced narrative than a simple "rates up, distributions down" explanation, as it reveals which variables are acting as the dominant marginal constraint at different points in time. The key message from the decomposition is that the largest structural headwind today is no longer only the level of interest rates, but the persistence of geopolitical and trade-related uncertainty. This is particularly relevant for software and growth assets, where valuations are primarily driven by confidence and exit timing is contingent on stable forward assumptions. Even if financing conditions gradually improve, elevated trade policy uncertainty depresses CEO confidence and reduces the willingness of strategic buyers to transact aggressively. In practice, this implies that distribution normalization may proceed unevenly. While IPO markets may reopen mechanically, exit volumes may remain constrained if the macroeconomic environment persists in a state of uncertainty and episodic risk appetite.

In light of the current market volatility, our distribution forecasts for 2026–2028 are presented in three scenarios that are directly linked to the trajectory of the software and AI cycle. The central intuition is clear: The backlog of exits that has accumulated since 2022 can only be released if public tech valuations provide a stable anchor and if confidence returns sufficiently for buyers to clear transactions. AI is a critical factor in this equation, as it has the potential to either extend the software growth narrative and re-rate the sector, or it could prove to be a bubble that leads to another valuation shock.

Our baseline scenario: The exit reopening is selective and structurally constrained. The baseline indicates that 2026 will mark the first significant step toward normalization, with distributions improving sharply (+5pp in 2026). This year, the market will begin to stabilize, with financing conditions improving, making IPOs viable for select issuers, and sponsors restarting exits in strong software assets where growth is defensible and cash flows remain robust. The baseline foresees that this recovery is driven by pent-up exit supply, particularly from high-quality assets that have been held back purely due to market timing. However, the recovery is expected to moderate quickly in 2027 (+1pp) and 2028 (+1pp). The reason is structural: once the initial wave of "sellable" assets has passed, the industry faces the more challenging aspect of the backlog, growth software and assets acquired at high multiples, platforms with decelerating ARR and deals where the original value creation thesis placed too much reliance on multiple expansion. In this regime, exits do return, but they do so unevenly. The market is operational, but it operates under different liquidity conditions than it did before 2022. As a result, distributions are more limited than they were from 2016 to 2021.

Our upside scenario ("AI Boom"): Valuation re-rating and a robust exit cycle are both on the rise. In a favorable scenario, AI evolves into a genuine macro tailwind, rather than a transient narrative. Technological investments are surging, productivity expectations are soaring and equity markets undergo a re-rating, all of which restore confidence in long-duration growth assets. In this environment, the IPO barometer strengthens further and issuance activity broadens beyond a narrow set of mega-cap or defensive names. This provides the valuation anchor that private markets require and reactivates the full exit ecosystem, IPO exits, strategic M&A and sponsor-to-sponsor transactions. In this scenario, distributions will rise strongly to +8pp in 2026, followed by continued momentum in 2027 (+3pp) and 2028 (+2pp). It is important to note that this scenario represents more than just "better markets." This shift indicates a return to more realistic outcomes, with large platform exits becoming viable once again, valuation discrepancies being resolved and sponsors allocating funds through authentic sales rather than relying on continuation funds or NAV-based liquidity solutions. In this scenario, the software crisis is addressed as a cyclical interruption, and the exit backlog is released more expeditiously and extensively.

Our downside scenario ("tech downturn"): The AI bust and the renewed liquidity trap. In the event of unfavorable circumstances, the AI narrative may undergo a valuation adjustment. Public tech equities have experienced a significant decline, corporate investment has shown signs of moderation and confidence levels have been observed to deteriorate. This combination has historically proven to be particularly detrimental for distributions, as it tends to occur when public comps reprice lower. As a result, private valuations become more challenging to justify, potential buyers become hesitant and the IPO window tends to close again. Consequently, the distribution cycle will weaken considerably, with a reduction of -3pp in 2026, followed by a gradual stabilization in 2027 (+1pp) and 2028 (+1pp). This scenario indicates a "double-dip" outcome for PE portfolios with a significant software component. Exit markets fail to normalize, the backlog remains stuck and sponsors are forced to extend holding periods further. Liquidity continues to be generated primarily through synthetic mechanisms, NAV financing, preferred equity, continuation vehicles and selective recapitalizations,  rather than through true realizations. In this system, the software reset evolves from a valuation shock into a prolonged structural constraint on PE cash returns.

Our key takeaway for 2026 is that this year will be either a release valve or a breaking point. Across various scenarios, 2026 emerges as a pivotal year. The baseline indicates a partial reopening of exit markets, but the recovery is limited by the structural overhang of peak-multiple software vintages. On a positive note, AI plays a pivotal role in restoring price discovery and alleviating the backlog, thereby facilitating healthier distribution cycles. On the other hand, AI optimism could lead to a renewed tech downturn, which would prolong the liquidity drought and reinforce the industry's reliance on financial engineering. The key conclusion is that the distribution outlook through 2028 remains fundamentally an IT / software valuation and confidence story. The IPO conditions are improving and offer early indications of a reopening exit window. However, the decomposition highlights that geopolitical and trade uncertainty remain the primary constraints. As long as uncertainty remains elevated, exit markets may reopen but normalization will remain fragile, making the distribution path highly scenario-dependent.

All in all, private equity remains a critical asset class for long-term capital formation and return enhancement, but the industry is entering a structurally different regime. The post-2022 period has permanently raised the bar: cash distributions, not paper NAV gains, will define allocator confidence, while the market increasingly prices operational capability rather than financial engineering. The primary concern for 2026-2028 is whether the Tech/software-heavy backlog can be cleared through a sustainable reopening of IPO and M&A channels, or whether the combination of valuation uncertainty and refinancing pressure will prolong the liquidity drought.