Precedent Transactions: Building the Analysis Properly

What the method is claiming

A precedent transaction analysis says that the best evidence of what someone would pay for this company is what people have recently paid for companies like it. The unit of observation is a completed deal, and the multiple is the price the acquirer paid against the target financials at the time.

Because those prices were paid to acquire control of an entire business, they include two things a share price does not: a control premium, and the buyer view of what the combination is worth. That is why the range sits above trading comparables, and it is also why the method overstates value for a company nobody is trying to buy.

Screening: the decision that sets the answer

Everything downstream is arithmetic. The analysis is made or lost when you decide which deals qualify.

• Business similarity first. Same sector is not enough. Similar end markets, similar margin structure, similar capital intensity and similar growth profile.
• Size. A $200m deal and a $5bn deal in the same industry are different markets with different buyer pools.
• Recency. Three to five years is the usual window. Older deals import a different rate environment and a different credit market, and a 2021 multiple in a 2026 set needs a footnote or an exclusion.
• Buyer type. Strategic buyers pay for synergies and sponsors pay for returns, so a set that mixes them without labelling them hides the most useful pattern in the data.
• Deal completion. Announced and terminated deals still carry information about price, but flag them. A withdrawn bid is evidence of what someone was willing to pay, not of what a business changed hands for.
• Geography and listing. Cross-border deals, take-privates and carve-outs each price differently enough to warrant separating.

Getting to the right price

The headline number in a press release is frequently not the number you want. You are building transaction enterprise value, which means starting from the offer price per share, working to equity value on a fully diluted basis, then adding net debt and other claims as at the announcement date.

Three traps recur. Contingent consideration, where part of the price depends on future performance and should be disclosed rather than silently included or excluded. Stock consideration, where the value depends on which date you price the acquirer shares. And assumed liabilities such as pensions or leases, which move enterprise value materially in industrials and retail.

• Use the target last twelve months figures as at announcement, not the latest fiscal year, and not a forecast unless the whole set is on forecasts.
• Adjust the target EBITDA for anything the market would treat as non-recurring, and apply the same policy to every deal in the set.
• Record the announcement date, the closing date, the buyer type and the consideration mix on every row. A set without those columns cannot be interrogated.

Control premia, and what they are not

The premium is usually calculated against the target undisturbed share price, most often one day, one week and one month before announcement. Undisturbed matters: if a bid leaked, the price the day before already contains the rumour, and the premium you compute against it understates what was paid.

A premium is not a valuation method on its own. Applying an average premium to a current share price produces a number, and the number assumes the current price is fair and that this buyer would pay the average. Both assumptions need saying out loud.

Where the data comes from

In a bank the set comes from a subscription database. Outside one, the primary sources are public and adequate for interview preparation and for a case study.

• Merger proxies and scheme documents, which set out the consideration, the fully diluted share count and often the adviser fairness analysis.
• Regulatory filings around the deal, including the acquirer purchase price allocation, which reveals what was actually paid and how it was allocated.
• Company announcements and investor presentations, where synergy expectations are usually stated and are the cleanest signal of why the premium was paid.

Presenting the range without hiding behind the median

The output is a range, normally quoted as the interquartile range with the median marked. Do not present the mean. One outlier deal, and there is always one, will drag it somewhere indefensible.

Then explain the dispersion. Strategic deals at the top, sponsor deals at the bottom, a distressed sale at the extreme low, one auction with three bidders at the extreme high. A precedent set that has been read rather than assembled tells a story about who pays what and why, and that story is what a client is buying.