Original conceptual illustration. Hypothetical calculations are labelled; book cases are retrospective accounts, and later comparisons are identified in their text.
Two different investment problems
Chapter 4 deliberately contrasts risk arbitrage with merger securities. Before completion, the investor studies whether and when promised consideration will arrive. After completion, a bond, preferred share, warrant or other instrument received as consideration may be unwanted by its new owner. The first problem is largely conditional deal exposure. The second can resemble the unwanted-security mechanism of a spin-off. Confusing them turns the chapter into a blanket endorsement of merger trading that it does not provide.
For a cash deal, the spread between target price and offer price compensates for more than waiting. Approval, financing, contractual conditions, business changes and renegotiation can affect completion. For a share deal, the consideration itself moves with the acquirer’s price. Exchange ratios, collars and election terms can make the payoff more complicated than a single announced headline value.
Break-even probability is not a forecast
In the learning lab, a fictional share costs 46, pays 50 if acquired and is worth 30 if the deal fails. Expected gross payoff at completion probability p is 50p + 30(1 − p). Subtract 46 to get expected profit. Setting the result to zero gives p = 0.80. This threshold tells you what belief the price requires under the chosen assumptions. It does not tell you the true probability, and it excludes fees, tax, financing and uncertainty about the failure value.
The failure value is especially important. It need not equal the pre-announcement price. The business, market or information set may have changed. If failure is associated with newly discovered problems, the target can be worth less than it was before the offer. A model using one fixed downside is a simplification whose limits belong beside the output, not hidden in a footnote.
Time and hedging change the result
A four-unit gain in three months is not the same use of capital as the same gain in twelve months. Annualising can help describe the difference, but it silently assumes a time interval and sometimes reinvestment opportunities that may not exist. It does not transform a conditional gain into a yield. When the expected closing date moves, update the waiting period without assuming that the economic risk has stayed constant.
In a stock-for-stock deal, a short position in the acquirer can offset some price exposure if the deal completes as specified. It brings borrowing costs, availability, margin and failure-scenario risk. If the exchange never occurs, the expected delivery of shares never arrives. A two-legged trade deserves a two-legged stress test. One cannot simply label it market-neutral and stop reading.
After the deal: read the new instrument
Imagine former shareholders receive cash plus a small bond holding. They may sell the bond because they wanted equity exposure, their mandate does not permit it, or its size is inconvenient. That creates a research candidate. The bond’s face amount is not necessarily its market value. Coupon, maturity, issuer solvency, seniority, call provisions and liquidity determine what claim the purchaser is actually receiving.
For a hypothetical bond with face value 100 and annual coupon 6, buying at 80 gives a 7.5 percent current yield. That is not the yield to maturity, because principal repayment and timing also matter. It is not a guaranteed return, because payment can fail. Similarly, a warrant with an attractive exercise price can expire worthless, and a contingent right may depend on an event that never occurs. Unwanted instruments require more precise contract reading, not less.
Greenblatt’s distinction is a lesson in choosing which uncertainty to study. A complex security after a completed transaction may have documents that allow detailed analysis. A pre-completion spread may require continuous judgement about events outside the company. Neither category is automatically appropriate for a particular learner. The useful output is an explanation of the claim and its failure conditions, not a recommendation to collect obscure securities.
Staples / Office Depot: the market-definition question
The proposed Staples acquisition of Office Depot encountered an antitrust challenge from the US Federal Trade Commission. In May 2016, a federal court granted a preliminary injunction. The companies subsequently abandoned the transaction, and the FTC dismissed its administrative case. This is a documented example of an announced acquisition that did not become the acquisition shareholders had initially been offered. It is a later comparison rather than a story from the book.
The FTC’s case focused on competition in consumable office supplies sold to large business customers under contracts. That scope is educationally important. A casual observer might look at shops, online sellers and a wide range of stationery outlets and conclude that competition was obvious. The regulator’s concern was a particular market and customer group. An investor’s intuitive description of an industry may not match the question that controls an approval process.
Our interpretation is that a deal researcher needs to translate every closing condition into an evidence problem. “There are other retailers” is an observation; “the relevant authority will find sufficient competition for the customers at issue” is a prediction requiring a different argument. The distinction can be understood without pretending to practise competition law or to assign a reliable numerical probability to a pending case. Specialist uncertainty remains uncertainty even when it has a familiar company name attached.
Consider a hypothetical target trading at 46 with a proposed cash payment of 50 and a failure value of 30. Completion gains 4; failure loses 16. An 80 percent completion probability only reaches break-even before costs because 0.8 × 4 equals 0.2 × 16. Those numbers are invented to show the payoff structure; they are not historical prices or a reconstruction of this transaction. If the failure value is worse, or the expected waiting period lengthens, the attraction changes again.
The official record establishes the legal and corporate sequence. It does not establish a particular arbitrageur’s loss, the correct probability before the ruling, or whether the spread adequately compensated a diversified specialist. Hindsight can make an outcome look predictable when reasonable disagreement existed beforehand. Good evaluation therefore asks whether the pre-event argument identified the relevant risk and had a credible downside, rather than merely whether the prediction matched the eventual result.
This case can be compared with Cypress Gardens. One illustrates a surprising business disruption; the other illustrates a formal approval obstacle. Their sources of uncertainty differ, but both show why an agreed price is conditional consideration. Neither proves that all merger arbitrage is irrational. They show why a learner should resist converting a signed agreement into the equivalent of cash already received. Ask how a research checklist would change once the relevant customer market, rather than the public-facing brand, became the unit of analysis.
Chapter connection
The Staples case shows why the relevant approval question must be specified. A broad belief that retail is competitive does not answer a narrower question about large customers under supply contracts. Compare the sourced outcome with the lab, but do not insert invented probabilities into the historical record.
Reflection
What changes if a six-month transaction takes eighteen months?
Revisit capital tied up, costs, business conditions and financing; do not only divide the same spread by a longer period.