The BCG M&A Report Collection

BCG’s M&A research data set (the “M&A database”), which we used as the basis for our analyses, comprises approximately 1,052,000 M&A deals covering the period from January 1980 through December 2025. For our assessment of general market trends, we analyzed reported M&A transactions from 1990 through the first eight months of 2026. For our analysis of aggregate deal values and volumes, we included all announced majority transactions, including pending, partially completed, completed, unconditional, and withdrawn deals. We excluded transactions marked as self-tenders, recapitalizations, exchange offers, repurchases, acquisitions of remaining interests, minority-stake purchases, privatizations, and spinoffs, except where indicated otherwise. We apply transaction-size thresholds where indicated, based on deal value including assumed liabilities.

Our analysis draws primarily on data from LSEG, supplemented by the BCG M&A Sentiment Index and BCG M&A Explorer data as well as selected other data sources as indicated.

Short-Term Value Creation

Although analyzing different issues required us to use distinct samples, we employed the same econometric methodology to all return analyses.

To determine the “announcement return,” we derived the cumulative abnormal return (CAR), by taking the difference between the actual return on the acquirer’s stock () and the return realized in the sector index () as an approximation for expected returns, starting three days before the announcement date (–3d) and ending three days after it (3d). (See Equation 1.)


EQUATION 1


Long-Term Value Creation

For M&A deals, we tracked the stock market performance of the acquirers or the sellers over periods of different length following the acquisition announcement. We could not track the targets because, in most cases, they are delisted from the public equity markets.

First, we measured the total shareholder return (TSR) generated by the acquirer or seller over a time period of length t. (See Equation 2.)


EQUATION 2



Second, we subtracted from the TSR the return that a benchmark index delivered over the same period, in order to find the relative total shareholder return (rTSR) that the acquirer or the seller generated—in other words, the return in excess of the benchmark return. ¹ 1 (See Equation 3.)

EQUATION 3


We could not include all deals in this analysis because, for some deals, the time elapsed since the announcement was too short to allow us to calculate the returns.

1. The benchmark indexes we applied are the relevant worldwide Refinitiv (formerly Thomson Reuters) indexes.

We applied common-practice statistical significance tests to all of our quantitative results in this report. To assess whether means differed statistically from zero, we used one-sample t-tests; where appropriate, we used two-sample t-tests to determine whether the difference between means differed significantly from zero—that is, whether two groups did in fact have different means.

Generally, for longer-term analyses (such as for one- and two-year rTSR) and for the short-term analysis (that is, for CAR), we used relative measures of size impact (such as deal value compared to the enterprise value of the acquirer) as well as absolute measures of size (such as deal value) to determine whether a transaction was sufficiently material to have had an impact on overall performance.