Q1 2026 M&A: record value, contested deal counts, and what it meant for diligence

The trackers disagree on how big Q1 2026 was, and on which way deal count moved

There is no single verifiable global total for the quarter. PitchBook, syndicated to Yahoo Finance, puts global Q1 2026 deal value at roughly US$1.6 trillion, up 50.6% year on year, across 13,877 deals, up 18.3%. Bain & Company, using Dealogic data and counting only deals above $30m, reports value up 28% and volume up 9%. Mergermarket's quarter-to-date read through 23 March 2026 had global M&A at US$1.16 trillion, up 22%. FTI Consulting reports that volume fell, down 4.4% quarter on quarter and 5.3% year on year.

The gap is methodological. PitchBook's universe includes very large private financings that league-table conventions handle differently, and Bain's floor excludes everything under $30m. Averaging them produces a number no dataset supports. We could not obtain a standalone LSEG Q1 total: LSEG data reported by Investment Executive covers only the half year, at US$2.85 trillion, up 50%, with transaction counts down 9%. A Q1 figure can be inferred arithmetically, but nobody published one, so we are not printing one. Pick one tracker and one inclusion threshold, and state which.

Value grew because a small number of very large transactions closed

Every dataset agrees on concentration. Mergermarket counted 17 megadeals of US$10bn or more worth US$413bn combined, and FTI puts strategic buyers at about 82.5% of global activity and intra-industry deals at 58.6% of transactions. On completions, WTW's Quarterly Deal Performance Monitor, using LSEG data records 12 deals above US$10bn closing against 2 in Q4 2025, and completed value of US$438bn, up 155% on Q1 2025. Pricing followed scale: PitchBook figures reported by Crowdfund Insider put the global median EV/EBITDA multiple at 10.7x, the highest since 2021, and deals above US$5bn at 13.9x.

SpaceX agreed to acquire xAI at a US$250bn valuation, announced 3 February 2026, more than 15% of global Q1 deal value on PitchBook's numbers. Paramount Skydance agreed to acquire Warner Bros. Discovery at US$81bn equity value and US$110bn enterprise value, with closing expected in Q3 2026; a US$170bn figure circulating in secondary write-ups does not match the parties' own release. McCormick's purchase of Unilever's foods business was The single link covers both figures but the Mergermarket EMEA page contains only the €37.1bn figure; the strings '44.8' and 'PitchBook' do not appear on it. The US$44.8bn figure is in the PitchBook-based Crowdfund Insider piece already cited elsewhere in this article ('the $44.8 billion sale of Unilever's food business to McCormick'). Split the link so the €37.1bn points at Mergermarket and the US$44.8bn points at the Crowdfund Insider URL., a discrepancy we could not resolve.

The regions moved in different directions, and so did acquirer returns

Mergermarket has the Americas at US$821bn, up 35.6%, on 3,782 deals, down 3.6%; EMEA at €360.2bn, up 28%, on 4,310 deals, down 16%; and APAC down on both measures at US$219bn, down 35.9%, on 2,334 deals, down 15.8%. PitchBook instead has North America at US$1,022.2bn on a deal count up 19.2%, so the direction of the Americas count is unresolved. Acquirer returns diverged as well: WTW measures acquirer share-price performance against regional indices on Q1 completions at plus 6.0 percentage points in Europe, minus 3.4 in APAC and minus 5.4 in North America.

Sector leadership depends on which dataset you open

Mergermarket attributes large-cap momentum to artificial intelligence, utilities and financial services, with the Americas led by TMT at US$301.8bn and EMEA by financial services at €82.1bn. PitchBook records IT's highest quarterly value on record at US$409.5bn. Bain attributes the most value growth to consumer products, healthcare and life sciences, and financial services. A pitch built on the AI-led quarter is defensible on one dataset and not on another.

Sponsors stayed busy without winning the top of the market

KPMG's Pulse of Private Equity puts Q1 global PE at US$436.4bn across 4,168 deals, with rolling 12-month volume falling to 19,682 deals from 21,026. Foley & Lardner, also citing PitchBook, reports 5,100 deals worth US$481.6bn and roughly US$2tn of dry powder; both attribute to the same provider, so the difference is probably data vintage.

The constraint is fundraising. KPMG has the rolling 12-month total at US$373bn across 549 funds, the lowest since Q1 2017. Sponsors were 40% of global deal count and 50% of value, but they backed only 1 of the 10 largest North American deals against 4 in Q4 2025, and the valuation gap between corporate and sponsor bids narrowed to 2.8x EBITDA from over 4x. Corporate buyers are outbidding sponsors at the top end.

The mid-market did not rebound, and financing explains most of it

Mergermarket names three reasons the recovery did not arrive: the perceived threat of AI to software valuations, elevated financing costs and geopolitical volatility, with US middle-market deal count down about 20% in January and February. The FOMC held its target range at 3-1/2 to 3-3/4 percent on an 11 to 1 vote at its March 17 to 18 meeting, and the same minutes call corporate bond credit performance solid and spreads historically narrow while noting that leveraged loan prices for software firms declined sharply, that redemption requests rose notably at several private credit funds, and that financing remained somewhat restrictive for smaller firms. PitchBook reports lenders asking for tighter covenants and larger equity contributions.

Federal merger enforcement went quiet and state challenges filled the space

Dechert's DAMITT report records no contested merger complaint from either US agency in the quarter. Review durations improved to 10.8 months on average across the four significant US investigations concluded in Q1 2026, against 12.3 months on average in 2025. Dechert's planning guidance is still to allow up to 11 months for a US agency investigation and a further 6 to 12 months if the deal is litigated.

Filing mechanics moved twice. The HSR size-of-transaction threshold rose to US$133.9m effective 17 February 2026, with fees running from US$35,000 to US$2,460,000. Wrong court. The 12 February 2026 vacatur was by the US District Court for the Eastern District of Texas (Judge Jeremy Kernodle) in the Chamber of Commerce / Longview Chamber challenge. The cited McDermott piece says the opposite of what the sentence claims: it records that on 19 March 2026 the Fifth Circuit rejected the FTC's request to pause the lower court's ruling, and that 'a federal district court in Texas had previously vacated the 2025 HSR form.' Rewrite as: 'Then a US District Court for the Eastern District of Texas ruling dated 12 February 2026 vacated the expanded 2025 HSR form under the Administrative Procedure Act; the Fifth Circuit declined to stay that ruling on 19 March 2026, reverting filers to the legacy form.' Note also that the Fifth Circuit granted a short administrative stay on 19 February 2026, so 'reverting filers to the legacy form' was not immediate., which directly reduces pre-filing document assembly. The offset is that state attorneys general became the active challengers: eight states sued to block Nexstar's US$6.2bn TEGNA acquisition despite FCC and DOJ approval, and twelve challenged DOJ's approval of HPE/Juniper. Federal clearance was no longer the end of regulatory risk.

What the quarter meant for diligence practice

The best evidence is KPMG's survey of 700 senior dealmakers across 20 jurisdictions, fielded between 19 December 2025 and 27 January 2026. AI is deployed in due diligence and valuation by 56% of respondents. The risks they rank highest are not classic financial diligence: operational disentanglement at 52%, valuation complexity at 43%, and IT and data separation at 40%. That matches the deal mix, with 71% of PE respondents open to or pursuing portfolio separations.

Carve-out diligence is a different job. The document set is incomplete by construction, because the target's records sit inside the parent's systems, and the questions that decide value concern shared contracts, shared IT and which employees transfer. Most of the effort goes into building and tracking the request list rather than reading a complete room.

Be plain about what the quarter's data does not tell you. There is no reputable published figure for commercial diligence duration in Q1 2026, and nearly every search result offering one is data-room or AI-tool marketing. SRS Acquiom's 2026 deal terms study names heightened due diligence and escrow utilisation as a focus area, but its sample is closings from 2020 to 2025 rather than this quarter. Dechert's regulatory durations are the only hard timeline evidence available.

CogniSuite is built for that workstream. Client request lists arrive as arbitrary spreadsheets and are imported by having the model describe the sheet's layout while the application reads the cells, so a request cannot be paraphrased or invented. Uploaded files are scored against open requests, duplicates are detected by embedding so an approved match inherits the earlier request's confirmed documents, and drafts written for a counterparty are grounded only in what that counterparty may read, with every citation checked against the source text before display.

The limits matter as much. It does not reconcile a figure that appears differently in two documents; the prompt asks the model to flag disagreements when both land in the same retrieval batch, and that is all. Each document is represented by a single embedding of its opening text, so retrieval points at documents, not at passages inside them. Scanned PDFs with no text layer will not enrich, since there is no OCR in the pipeline. The audit trail is readable by the advising team and not by counterparties, so it is an oversight record, not a mutual one. Controls are described on our security page.