Mega-Merger Announced: What Happens to the Buyer's Stock?

Deal headlines keep getting bigger. 2026 brought SpaceX's reported $250 billion combination with xAI, Boston Scientific's $14.5 billion agreement to buy Penumbra, and a wave of other large transactions. When a merger is announced, the first question is simple: did the market like it? Statistics offers a tidy way to ask.

The idea of an abnormal return

An event study compares a company's actual stock return around the announcement with what it would normally have earned given the market's movement. The difference is the abnormal return. Researchers average it across many deals, since one announcement tells you little. Large studies typically find big gains for target shareholders and returns near zero, sometimes slightly negative, for acquirers (Andrade, Mitchell and Stafford, 2001).

Testing twelve announcements

Suppose twelve acquirers show three-day abnormal returns, in percent, of −2.1, −1.4, −0.8, −3.0, −1.9, −0.6, −2.5, −1.2, −0.4, −2.8, −1.7 and +12.0. The one-sample t-test calculator tests whether the mean differs from zero. The mean is −0.53%, but the p-value is about 0.66, so it looks like noise. The confidence interval for a mean runs from −3.1% to +2.0%.

That conclusion is fragile, because eleven of the twelve returns are negative and one outlier does the damage. Remove the +12.0 and the mean becomes −1.67%, with p < 0.001 and an interval from −2.27% to −1.08%.

Mean, median and one big winner

The five-number summary calculator shows the median at −1.55%, far from the mean, and a maximum of 12.0 sitting alone at the top. When mean and median disagree this much, look for an outlier before reading the headline. The descriptive statistics calculator shows the same gap.

Neither answer is the whole truth. Dropping a data point because it is inconvenient is dishonest, and averaging it in without comment is misleading. Analysts report both versions, check whether the outlier was special, and consider robust methods that give extreme values less weight.

What this does and does not show

An announcement-day return reflects the market's expectations, not the eventual outcome. Investors can be wrong, and a deal's real success takes years to judge. Studies also separate deals by payment method, size and industry, because averaging very different transactions hides patterns. A broad review of the evidence is available from the US Federal Trade Commission.

Remember too that stock moves around announcements can be driven by news unrelated to the deal, so careful studies model the normal return first. Twelve deals are enough for a demonstration, not for a verdict on mergers in general.

Beyond the announcement window

Short windows capture the market's first reaction. Long-run studies track returns for months or years, but they face their own difficulty: after so long, many other events affect a company, and attributing changes to a merger becomes guesswork. Researchers also compare deals paid in cash with deals paid in shares, since payment method changes how investors read the news. Each split reduces the sample, so intervals widen, which is the price of asking finer questions.

A wide interval is not a failure of the method. It is an honest statement of how little a handful of deals can reveal.

The sensible reading of a merger headline is modest: ask how many deals the claim rests on, whether an outlier drives the average, and what the market expected before the announcement.