In January 2026, private-equity-backed acquirers closed more than 25 accounting-firm deals in a single month — roughly a third of the entire 2025 deal count, in 31 days (CPA Trendlines).
The profession spends a lot of attention on AI adoption at the Big 4 and on the CPA pipeline crisis. Both matter. But the most direct force restructuring firm ownership right now is capital. This piece maps the deal data, the mechanics of the rollup, and how AI fits — as neutrally as the numbers allow.
Disclosure: I'm building Artifi, an AI tooling layer for finance teams, so I have a stake in adjacent markets. The figures below are presented as reported.
The deal data (and how to read it)
Several trackers count slightly different things, so it helps to separate them. Cornerstone's deal tracker recorded 22 PE-backed accounting transactions in 2023, 65 in 2024, and 104 in 2025, with 25+ in January 2026 alone (CPA Trendlines). A separate tally of larger platform transactions counts 53 notable deals and nearly $29B in capital since 2020 (CFO Brew). And IFAC, counting globally and including follow-on acquisitions, puts the total at 1,052 PE-touched firms since 2015 — 177 platform investments that generated 875 roll-up acquisitions, roughly five per platform (Going Concern). The three figures aren't in conflict; they count deals, notable platforms, and total firms respectively.
The headline transactions are real money. Baker Tilly and Moss Adams announced a $7B merger backed by Hellman & Friedman and Valeas in April 2025 (PR Newswire). In January 2025, Blackstone bought New Mountain's stake in Citrin Cooperman at roughly a $2B valuation — the first PE-to-PE "flip" of a US CPA firm, after revenue rose from $350M to $850M in three years (CPA Practice Advisor). EisnerAmper, the first Top-20 firm to take PE money in 2021, closed a continuation-vehicle transaction in March 2026 after 27 acquisitions and $1.2B in revenue (Inside Public Accounting). Cherry Bekaert, Aprio, PKF O'Connor Davies, and the CBIZ–Marcum merger fill in the rest of the picture.
Why now: three forces at once
Succession. The average partner age at US CPA firms is around 52.2, and in $2M–$10M firms a majority of partners are over 50, while fewer than half of firms have a written succession plan (Inside Public Accounting). Many partners assumed they'd sell to the next generation of partners; that generation is thinner than expected.
The pipeline. First-time CPA exam candidates fell from 48,004 in 2016 to 32,188 in 2021 (CFO Dive), with accounting degrees down again in 2023–24 and only early signs of recovery (Journal of Accountancy).
Capital-cost arbitrage. Independent firms historically traded around 0.8x–1.5x revenue; PE-backed acquirers now pay 8x–12x EBITDA for mid-market platforms (Auxo Capital). A partner weighing "sell to my associates at a low multiple" against "sell to a rollup at a high one with an earn-out" faces a clear gap.
The rollup playbook
The mechanics are borrowed from dental, vet, and HVAC consolidation: acquire a platform firm at 8x–10x EBITDA; bolt on smaller firms at 5x–7x; bank the spread as "multiple arbitrage" before any operational change; centralize HR, IT, payroll, billing, and procurement; standardize the tech stack; and exit in 5–7 years to a secondary buyer or via continuation vehicle (CT Acquisitions).
What's specific to accounting is a regulatory two-entity split: the licensed CPA entity (attest, audit) stays partner-owned, while a separate advisory entity holds the PE capital. Every major deal runs this structure, and Blackstone keeps its Citrin stake under 50% to manage audit-independence concerns (Accounting Today). Different sponsors run variations — Alpine's Ascend takes equal-partnership stakes; United Accountants, an ex-PwC-founded platform that raised a $2.08M seed in February 2025, targets the sub-$5M firms larger platforms ignore (PitchBook).
Where AI fits
A recurring observation from firms across the spectrum is that a 30-person independent practice struggles to adopt AI meaningfully — not because tools are unavailable (Aiwyn raised $113M from KKR and Bessemer in late 2024 and grew its firm customer base from 200 to 840 in a year, per its own figures (BusinessWire)), but because the incentives inside an independent partnership often work against it. (That dynamic is the subject of a separate piece on the innovator's dilemma.)
The rollup is the ownership structure where the AI math lines up most cleanly: cost is centralized at the holdco while benefit is distributed across portfolio firms, engineering is hired once and leveraged across many, the acquired partner takes an earn-out rather than adoption risk, and a single workflow automated at the holdco can widen across every firm — multiple arbitrage and margin expansion stacking on top of each other. IFAC's own report flagged that its count "will likely be much higher once someone perfects the burgeoning AI-forward, roll-up practice model" (Going Concern).
What this means for each group
For independent partners: the data shows pressure from several directions at once — lower-cost AI-native entrants on commodity work, acquisition offers that internal succession can't match, and larger advisory firms competing for the biggest clients. The structural responses are covered in the innovator's dilemma piece.
For the rollups: the open execution question is whether holdco-level AI deployment delivers the margin expansion the thesis assumes; the IFAC framing suggests no platform has fully demonstrated it yet.
For investors: the consolidation curve is steep and the secondary market for accounting platforms is now active (Citrin's flip being the first example). The variable to watch is operational integration, not deal supply.
Outlook
The partnership model has been the dominant ownership structure in accounting for over a century. The data points to meaningful consolidation of the $5M–$100M mid-market over the next 5–10 years, with the largest firms and the smallest lifestyle practices most likely to persist. Whether one or two AI-forward rollup platforms scale past $1B in revenue on holdco automation economics is, on current evidence, the open question — the capital is committed; the operational proof is still being written.