The risk that a signed deal will nevertheless fail to reach completion has always been a focal point of public company mergers and acquisitions negotiations. This closing risk exists because the signing of a merger agreement and the completion of the planned deal do not occur simultaneously. Between the signing and closing, a multitude of factors can cause a signed deal to break: the acquirer may fail to secure the regulatory clearance or financing necessary for the purchase, a rival bidder for the seller could emerge and trump the original acquirer’s bid, market conditions may change so that one of the parties loses its appetite for the deal, and so forth.

Despite the ubiquity of closing risks in M&A practice, no prior work in either the finance or corporate law literature has provided a systemic account or analysis of these risks. In our new article, How Deals Die, we aim to draw closing risk into the open. In doing so, we show that attending to closing risk, as distinct from the deal structures and agreement terms that closing risk affects, can generate a variety of practical and doctrinal insights.

We begin by offering a comprehensive typology of outcomes for M&A transactions. The M&A scholarship, particularly empirical-financial studies, has conventionally treated deal outcomes as binary: a deal is either “completed” between the merging parties or it is “withdrawn.” But as we show, this familiar grouping is so coarse as to be misleading. “Completed” deals in fact encompass three distinct results that impact the merging parties and their shareholders in different ways: a merger could close on the originally agreed economic terms, close following an upward price revision, or close following a downward price revision. “Withdrawn” deals, too, take five distinct forms: successful third-party topping bids for the target company, target-initiated withdrawals, acquirer-initiated withdrawals, mutually agreed-to withdrawals, and regulatory blocks.

For most research designs, ignoring this granularity is a serious conceptual mistake. It would mean, for instance, treating the breakdown of Frontier’s acquisition of Spirit Airlines, which was jumped by rival JetBlue Airways’ topping bid, as no different from Elon Musk’s attempted withdrawal from his deal to buy Twitter (if he had succeeded), even though Spirit’s exit enabled its shareholders to accept a higher offer from JetBlue while Musk’s exit would have left Twitter’s shareholders with no payment at all. As we explain in the article, empirical studies that use the binary completed/withdrawn categories are vulnerable to measurement error, which in turn risks inconsistent, distorted, or even theoretically ungrounded statistical inferences.

The article’s second contribution is empirical. Using a novel, hand-collected, and rigorously verified data corpus of 5,058 definitive merger agreements involving U.S. public company targets signed between 1996 and 2020, we present a detailed picture of deal breakage in the United States, documenting how often deals break, why deals break, and how deal breakage correlates with deal structure and agreement attributes. To preview some of the findings: overall, the incidence of deal breakage (failure to consummate on the originally-announced economic terms) over the past twenty-five years follows a steady beat, peaking to around 12% during periods of market turmoil, then falling to around 9% during periods of quiescence. The stickiness of deal breakage rates over this time period is surprising because today’s merger agreements look very different—in length, complexity, and substance—from agreements from the 1990s. Despite abundant contractual innovation, the general risk of deal non-consummation remains largely unchanged.

Peeling the onion does, however, reveal evolution in the specific form of deal breakage. The frenzied dot-com bubble was the era of all-stock deals so a significant fraction of the deals that fell apart once the market tumbled occurred by mutual agreement, as sinking stock prices both lowered the value promised to targets and generated buyer’s remorse. In the years immediately preceding the 2008 crisis, private equity sponsors became major players to the M&A landscape, and their agreements contained terms that allowed them to unilaterally terminate the deal for any reason by paying a modest “reverse termination” fee. Many private equity buyers then exercised this right—despite fierce seller resistance—during the market turmoil of 2007 and 2008. Target companies quickly learned their lesson: in the post-crisis years, the structure of private equity contracts changed so that it no longer provided acquirers with a cheap escape hatch. Once again, the dynamics of deal breakage shifted. Private equity-backed buyers successfully exited or renegotiated only 1% of deals that were announced between 2008 and 2020, a sharp decline from the rate of 11.5% among pre-2008 deals.

We conclude with implications for several key doctrinal debates. For example, while empirical studies have yielded a mixed picture on the efficacy of go-shop processes to produce a higher bidder, no one has asked what happens after the new bidder appears. We find that, among the private equity deals that were announced before 2014, interlopers who emerged during the go-shop process won the bidding war in most cases, but they have lost nearly all such bidding wars since then. Recognizing this change in private equity buyers’ ability or willingness to exercise their match rights adds an important dimension to ongoing discussions about how go-shops have evolved. Our data also illuminates the extent to which institutional minority shareholders can successfully resist a controlling shareholder who pursues a freeze-out merger at an exploitative price, which has direct relevance for Delaware courts’ controversial adherence to the blueprints established by Kahn v. M & F Worldwide Corporation and In re Match Group, Inc. Derivative Litigation. Likewise, our observation that deals subject to Revlon-duties are as likely as non-Revlon deals to experience a renegotiated price bump after the definitive merger agreement is executed likely has significance for recent debates over the efficacy of pre-signing market checks and Revlon’s continued utility. Finally, we show that, contrary to some statements in the literature, merger arbitrage investors should not be expected to invariably favor deal completion—a finding that may have implications for Delaware jurisprudence that cleanses target directors of fiduciary liability if shareholders approve the transaction.