M&A Deals Dataset 2019–2026: Global Acquisitions

Key findings

1. 123,550 M&A deals between 65,109 acquirers and 107,767 targets, January 2019 to June 2026. 84% are acquisitions of control, 6% mergers and 9% minority stakes.
2. Deal count peaked in 2025 with 20,854 deals, against 10,424 in 2019. 2026 has 10,818 in its first half.
3. 29,882 deals (24%) disclose a value, together $24.2T; the median disclosed deal is $100.0M. The largest is United Technologies / Raytheon ($120.0B, 2019).
4. 43% of deals with both countries known are cross-border. US companies are on one side of 54,058 deals, UK companies 17,114 and Indian companies 8,970; 1,725 deals bring a foreign acquirer into India.
5. The most active acquirer is Accenture with 231 deals, followed by KKR (186) and Blackstone (128).
6. 37,896 deals are reported completed and 718 terminated; the rest were last reported as announced or pending.

M&A deals by year and type

2026 covers January to June. Values are the sum of disclosed deal values.

Top 50 largest M&A deals worldwide, 2019–2026

Disclosed values only, checked against the news sentence. Values are as reported at announcement (enterprise or equity value).

Most acquisitive companies (top 50)

Deals as acquirer, after merging repeated reports of the same deal.

M&A deals by target country and year (top 30)

M&A by country worldwide (top 40)

Inbound: a foreign acquirer buys a company in the country. Outbound: a company from the country buys abroad.

Top 40 cross-border M&A corridors

M&A deals by target sector (top 30)

Top 50 largest US M&A deals, 2019–2026

Deals where the acquirer or the target is based in the United States.

Most acquisitive US companies (top 50)

US M&A by state of the target (top 30)

Location is the target company's headquarters as held in the company database.

Top 50 largest UK M&A deals, 2019–2026

Deals where the acquirer or the target is based in the United Kingdom.

Most acquisitive UK companies (top 50)

UK M&A by city of the target (top 30)

Location is the target company's headquarters as held in the company database.

Top 50 largest Indian M&A deals, 2019–2026

Deals where the acquirer or the target is based in India.

Most acquisitive Indian companies (top 50)

Indian M&A by state of the target (top 30)

Location is the target company's headquarters as held in the company database.

Column dictionary

23 columns in the sample CSV.

Method

1. Extraction. Radar's pipeline reads each business-news article, finds M&A sentences and extracts acquirer, target, value, currency, stake, status and date. Upstream, reports carrying the same event fingerprint are grouped and retrospective mentions of old deals are dropped.
2. Window and dates. The announced date is the publication date of the earliest report, read in Indian Standard Time; January 2019 to June 2026. Days carrying a bulk archive back-fill (over five times the usual volume) are dropped.
3. Parties. Descriptor prefixes are removed ("Chrysler parent Stellantis" → Stellantis). Deals are dropped when either party is a person, an undisclosed or generic party ("an unnamed buyer", "the company") or the same company on both sides. A website linked to a party is kept only when the company record matches the reported name; otherwise the name is kept and the website blanked.
4. Rumours. Reports flagged as rumoured, or worded as talks, considerations or potential bids, are dropped.
5. Deal type. Merger when reported as one; stake when the share acquired is under 50% (or a stake purchase without control wording); acquisition otherwise.
6. One row per deal. Reports naming the same target and the same (or an alias of the) acquirer within 365 days are one deal; a second pass merges target name variants (with and without a website) under one acquirer. The latest reported status wins, and completion overrides announcement.
7. Amounts. Values are converted to US$ (stored US$ value, or the native amount at 2023–2025 average rates) and kept only when they match a money figure in the sentence (0.6–1.7×) — verified — or when the sentence has no parseable figure and the value is plausible — unchecked. Values above $300B are blanked, above $25B they must be verified and reported more than once, and unchecked values above $2B are blanked. "$" in Latin-American deals is read as the local peso unless US dollars are named.

8. Countries and sectors. Countries are normalised (USA / US → United States). A company's country is the one most often reported for it across all its deals (at least two reports and a majority), else the country in the report, else the company database. Sector is the target's Fliar sector label.

9. Sample. Each acquirer key is ranked by sha1(key). Keys are taken in rank order until the file holds max(25%, 5,000) rows (capped at 100,000).

10. Profile links. Links are built from Fliar's company index and checked against it. A row whose company has no public profile has an empty link.

Limitations

• Derived from news: unreported deals are missing and coverage favours large, English-reported companies.
• ClickHouse shards were partially unavailable at build time (shard 2 down; queries skip unavailable shards), so some deals may be missing. Article links resolve for roughly half of the deals because some article-store nodes were down.
• Deal values are as reported at announcement and mix equity and enterprise values; about two-thirds of deals disclose no value.
• Merging is automatic: a few deals may remain split across name variants, and a repeat purchase from the same seller within a year is merged into one deal.
• Country, state and city come from the extraction or the company database; a multinational may appear under a national website.
• Deal types and statuses come from a classifier and some are wrong.
• 2026 covers January to June only.

Cite this dataset

@misc{fliarbi_mergers_acquisitions_2019_2026_2026,
  author    = {{Fliar BI}},
  title     = {{M&A Deals Dataset 2019–2026: Global Acquisitions}},
  year      = {2026},
  publisher = {Fliar BI},
  url       = {https://fliarbi.com/datasets/mergers-acquisitions-2019-2026/},
  note      = {Sample release. Licence CC BY-NC 4.0}
}