The Disciplines of M&A
Forty years of deal counts and deal values, what the two measures do differently, and why every "record year" has been a smaller share of a bigger economy.
Every wave history of M&A is told in adjectives — booming, frothy, frozen. The numbers tell a plainer story, and a more useful one for anyone pricing risk or timing a process. The longest continuous public series comes from the Institute for Mergers, Acquisitions and Alliances (IMAA), whose deal-count data — itself compiled from Thomson Reuters/LSEG and S&P Capital IQ — runs unbroken from 1985. Dollar values are less continuous: no single free provider has published one comparable annual value series for forty years, so the value figures below are stitched from whichever bank or data vendor published a same-year total, which means methodology shifts underneath the numbers. That caveat matters more than it sounds, and it recurs throughout this piece.
Global deal count rose from 2,676 in 1985 to 47,455 in 2007, fell hard in the 2008-09 crisis, recovered to a first post-crisis peak of roughly 48,000-50,000 deals a year in the mid-2010s, then hit its all-time high of 58,308 deals in 2021 (IMAA/Thomson Reuters, 2024) before sliding to 39,603 in 2023 and 36,067 in 2024 — the lowest count since 2005.
Forty years of global M&A, 1985–2025
Dollar value moved on a different, lumpier clock. Thomson Financial Securities Data put 1999 at $3.4tn, a 40% jump from 1998's $2.5tn — the top of the fourth wave's final, dot-com-fuelled leg. Dealogic later put 2007 at $4.6tn, the record of its era, wiped out within two years by the financial crisis (year-over-year declines on the order of 30-40% through 2008-09 were documented in press, though a clean full-year trough figure was not available for this piece). The market did not reclaim 2007's total until 2015, when Dealogic reported global M&A "surpassed $5 trillion" — a new record, by its own account 9% above 2007. That record stood until 2021, when LSEG data (as reported by Axios, 2025) put the year at $5.48tn, the current all-time high. Then came the sharpest reversal in the modern series: Bain & Company put 2023 at $3.2tn, a decade low, before a rebound to roughly $3.5tn in 2024 and then $4.8-4.9tn in 2025 (Bain, 2026) — a year LSEG's preliminary tally, cited by Axios, put closer to $4.4tn through mid-December. Both trackers agree 2025 was the second-best year on record after 2021 and the third year ever to clear $4tn, alongside 2007 and 2015.
Wave IV (mid-to-late 1980s). IMAA's count series shows deals climbing from 2,676 in 1985 to 10,814 in 1990 — a four-fold increase in five years, consistent with the leveraged-buyout boom the narrative history describes. No continuous annual dollar series survives from this period; the only aggregate found is Thomson Financial's claim that total merger value rose "100-fold" from 1980 to 1999, reaching a cumulative $15tn (Global Policy Forum, citing UNCTAD, 1999) — a cumulative figure, not a peak-year one. The wave ended with the junk-bond market's collapse and the 1990-91 recession; deal count itself barely dipped (10,814 in 1990 versus 14,722 in 1991), a reminder that count and macro turning points do not always align cleanly either.
Wave V (late 1990s). Peak value: 1999, at $3.4tn (Thomson Financial), up from $2.5tn in 1998. Deal count peaked the following year, at 39,783 in 2000 (IMAA/Thomson Reuters), ahead of 1999's 33,132 — value and count peaked in different years even within this one wave. The decline that ended it was the dot-com crash and 2001 recession; IMAA's count fell to 31,047 in 2001 and 27,201 in 2002, its lowest point since the mid-1990s.
Wave VI (2003-2007). Peak value: 2007, at $4.6tn (Dealogic). Peak count in the same run: also 2007, at 47,455 — one of the few years in this table where count and value crest together. The decline was the global financial crisis: Dealogic and Business Standard reporting document roughly 30% and 42% year-over-year drops through 2008 and into 2009, though this piece could not source a clean full-year trough total for either year.
The modern era (2010-2025). Two records inside one long cycle: 2015 at just over $5tn (Dealogic) and 2021 at $5.48tn (LSEG). Deal count peaked separately, at 58,308 in 2021 — the all-time high in the IMAA series — meaning 2021 is the rare year where both measures topped out together. The decline that followed was the sharpest and fastest in the whole series: from $5.48tn in 2021 to $3.2tn in 2023, a fall of roughly 42% in two years, driven by rate increases, collapsing tech and growth valuations (Bain, 2024, put strategic deal multiples down to 10.1x — a 15-year low — with tech multiples nearly halving from 25x to 13x), and a regulatory environment that Bain (2024) says added roughly a year to the timeline for scrutinized deals. The 2025 rebound to $4.8-4.9tn (Bain) or $4.4tn (LSEG) — take your pick of provider — restored most, but not all, of the ground lost since 2021.
This is the point the wave narrative usually skips, and it is worth stating plainly with numbers rather than asserting it as a truism. Deal count and deal value are different measurements of different things — one counts transactions, the other counts dollars, and a handful of megadeals can move the second without moving the first at all.
Count and value are different measurements and do not move together
The cleanest example in the data is 2015-into-2016. IMAA's count series shows almost no change — 48,052 deals in 2015, 49,991 in 2016, actually a 4% increase. Value, meanwhile, fell from just over $5tn to $3.69tn (Dealogic, as reported by Business Standard, 2016) — a 26% drop. More deals, a quarter less money: the difference is megadeal supply, not deal-market breadth.
The most current example is 2025 itself, and the two trackers disagree on the count's direction but agree on the underlying mechanism. Bain (2026) reports 2025 deal count up about 5% against a value increase of 36-40%; LSEG's data, per Axios, has deal count down 7% to a nine-year low against a value increase of 45%. Whichever count figure is right, both sources describe the same phenomenon: value rising several times faster than the number of transactions. Bain's own reporting explains why — megadeals of $5bn or more drove roughly 75% of 2025's strategic value growth, about 60% of them done by infrequent acquirers, and roughly 40% of them large enough to exceed half the acquirer's own market capitalization (Bain, 2025). That is a market where a small number of very large, often once-in-a-cycle transactions are doing most of the work, while the broad base of smaller dealmaking is flat or shrinking. The same pattern shows up again comparing 2023 to 2024: IMAA's count fell 9% (39,603 to 36,067) while Bain's value estimate rose from $3.2tn to roughly $3.5tn — fewer deals, more money, for a second year running.
Every "record year" headline — 2000, 2007, 2015, 2021, 2025 — is stated in nominal dollars, unadjusted for inflation and unadjusted for the fact that the pool of assets available to be acquired (the world's investable market capitalization, roughly tracking world GDP) has grown enormously over the same period. A $4.8tn year in 2025 and a $3.4tn year in 1999 are not the same-sized event merely because 4.8 is bigger than 3.4; the economy generating those deals in 2025 is roughly three-and-a-half times larger in nominal terms than the economy of the late 1990s.
Every "record year" has been a smaller share of a bigger economy
The honest comparison is M&A value as a share of the denominator it is drawn from. Using World Bank nominal world GDP figures (as compiled by StatisticsTimes, 2026) alongside the value figures above: 2007's $4.6tn against $58.6tn of world GDP is 7.9%. 2015's $5tn-plus against $75.8tn is 6.6%. 2021's $5.48tn against $98.8tn is 5.5%. 2023's $3.2tn against $107.4tn is 3.0%, and 2024's roughly $3.5tn against $111.7tn is 3.1%. Put this way, 2007 — not 2015 and not 2021 — was the largest M&A wave of the last two decades relative to the size of the world economy, even though both later years posted bigger nominal totals. Every subsequent "record" has, in fact, been a smaller share of a bigger pie. This ratio is almost never published alongside the headline dollar figure, and it changes the ranking of "biggest wave ever" entirely. (A comparable ratio for the 1999-2000 top was not computed here for lack of a matched-year GDP figure in the sources retrieved, but with world GDP at roughly $34tn in 2000 — under a third of today's level — the order of magnitude was clearly in the same range as 2007's, or higher.)
A fully sourced regional time series eluded this pass — Statista's regional-share chart, the most obvious free candidate, sits behind a paywall past its headline figure. What is directly sourced is a detailed snapshot of 2025: Bain (2025) reports the United States accounted for nearly half of global strategic deal value; Greater China was the second-largest market by value and led in deal count, with about 80% of Chinese activity domestic; Japan's deal value doubled year over year to become the third-largest market globally; and EMEA posted strong value growth even as its deal count fell 7%. That snapshot is consistent with the conventional account of the last two decades — North America as the largest single pool of value, a Europe whose deal count holds up better than its value share, and an Asia-Pacific region, China and Japan specifically, whose share of global value has become large enough to move the aggregate — but a defensible multi-decade percentage series for the three regions was not available from free sources in this pass.
Set aside sentiment and look at the mechanics. Four forces recur across every peak and trough in this table. Credit availability sets the ceiling on leveraged deal-making — the 1980s wave rode the junk-bond market, and its end coincided with that market's collapse; the 2008-09 trough and the 2022-23 trough both followed rate shocks that repriced debt-funded buyouts overnight. Equity valuations set the currency for stock deals and the multiple buyers are willing to pay in cash deals alike — Bain (2024) points to strategic multiples falling to 10.1x in 2023, the lowest in fifteen years, and tech multiples nearly halving from 25x to 13x, as the proximate cause of the 2022-23 trough, and to a "valuation gap" between what buyers would pay and sellers would accept as the reason the recovery was slower than falling rates alone would predict. CEO and board confidence is the least measurable of the four but shows up indirectly in the megadeal data: Bain's finding that roughly 60% of 2025's largest deals were done by infrequent acquirers, and that nearly 40% were large enough to exceed half the acquirer's market cap, describes boards willing to bet the company — a confidence signal, not a financing one. And regulatory posture sets the clock speed and, at the margin, the volume: Bain (2024) found at least $361bn of announced deals faced regulatory challenge over a two-year window, with contested deals taking roughly 12 months to resolve versus about three months for uncontested ones — friction that shows up as depressed count in tight-scrutiny years and, when it eases, as some of the "backlog" effect Bain has cited to explain both the 2024 build-up and the 2025 release. None of these four is sufficient on its own to explain a cycle turn; every major inflection in this table — 1990, 2001, 2008, 2022 — coincides with at least two of them moving together.