Editor’s note: This is general educational information about how the price in a tender offer is arrived at, using disclosed 2026 filings as worked examples. It is not investment advice. The SEC filings relied on are listed at the end.
A tender offer premium is usually reported as a single percentage, quoted against yesterday’s close, and treated as the deal’s headline. Inside the filings, that percentage is an output rather than an input. It is what remains after a board, a financial adviser and a bidder have argued over a per share number, and the reference price they divide it by is a choice made after the number was fixed.
The documents that show the argument are public. A third-party offer produces a Schedule 14D-9, in which the target’s board must explain what it considered and summarize its adviser’s analyses. An issuer self-tender produces a Schedule TO, and where the structure is a modified Dutch auction, the pricing decision is handed partly to the holders themselves.
The third-party case: a premium range applied to an unaffected price
The Apellis Pharmaceuticals recommendation statement filed on April 14, 2026 sets out the arithmetic plainly. Biogen’s purchaser offered $41.00 per share in cash plus one contingent value right entitling the holder to potential payments of up to an aggregate of $4.00 in cash on the achievement of specified milestones.
The board’s stated reasons record three different premium calculations on the same $41.00 upfront figure: approximately 140% over the closing price per share on March 30, 2026, the last full trading day before the merger agreement was announced; approximately 86% over the 90-day volume-weighted average price; and approximately 35% over the 52-week high. All three are true, and the gap between the first and the third is the entire question of what the shares were worth before the offer existed.
The adviser’s work shows the direction of travel. Evercore reviewed 33 transactions for control of US public biotech targets with aggregate transaction value above $1.0 billion announced since January 2024, and calculated the premium in each as the percentage by which the per share consideration exceeded the closing price one day before announcement. The distribution it found was wide: a 25th percentile of 33.7%, a median of 65.2%, a mean of 61.9% and a 75th percentile of 79.7%.
Then comes the step that turns a distribution into a number. Evercore applied a premium range of 34% to 80% to the closing price per share of $17.09 as of March 30, 2026, producing implied equity values of $23 to $31 per share. The upfront consideration of $41 sat above that range, and above the 52-week closing high of $30.05 recorded over the twelve month period ended March 30, 2026, in a stock whose low over the same window was $16.36.
That sequence is worth reading twice. The premium was not derived from the company’s cash flows. It was derived from what acquirers had recently paid over market in comparable control transactions, applied to a market price that the same document elsewhere describes as depressed relative to the stock’s own history.
One statutory rule sits behind every negotiated price. Section 14(d)(7) of the Securities Exchange Act provides that where a bidder varies the terms of a tender offer before expiration by increasing the consideration offered, it “shall pay the increased consideration to each security holder whose securities are taken up and paid for pursuant to the tender offer” whether or not those securities were taken up before the variation. A bidder that raises its price cannot confine the raise to late holdouts, which is why the opening number in a contested situation is rarely the bidder’s reservation number.
The self-tender case: the company sets a band, holders set the price
When the buyer is the issuer, there is no negotiation with a counterparty and no fairness opinion on the price. The modified Dutch auction substitutes an auction for both. The company publishes a price range and a dollar cap, holders tender at prices of their choosing inside the band, and the company pays the lowest price that lets it buy the amount sought. Everyone whose shares are accepted receives that same price, including holders who tendered lower.
The 2026 record shows how the band is anchored. Scholastic Corporation opened its offer on March 23, 2026 at not less than $36.00 nor more than $40.00 per share, against a close of $37.25 on March 20, 2026, the last full trading day before commencement, for up to $200 million of stock. Wix.com Ltd. commenced on March 5, 2026 at not less than $80.00 nor more than $92.00 per share against a March 4, 2026 close of $83.78, for up to $1,750,000,000. Expensify, Inc. commenced on May 13, 2026 at not less than $0.98 nor more than $1.20 per share for up to $25,000,000.
In each case the market price sits inside the band rather than below it. The floor is a discount to the last close and the ceiling is a premium, so the announced offer carries no single premium figure at all until it expires.
Both completed auctions cleared at their ceiling. Scholastic accepted 2,834,018 shares at $40.00, an aggregate $113,360,720 and approximately 13.7% of shares outstanding, well short of the $200 million authorized. Wix accepted 17,577,250 shares at $92.00, approximately $1.617 billion and about 29.7% of shares outstanding.
Analysis: the reference price does more work than the premium
A premium percentage is a ratio, and in a contested reading the numerator is the least interesting half of it. The Apellis filing states three premiums on one price because three defensible reference points existed: an unaffected close, a 90-day average, and a 52-week high. A reader given only the first of those learns that the offer was 140% above a single day’s print. A reader given the third learns that the offer was 35% above the best price the stock had reached in a year. Neither figure is wrong and neither is sufficient.
The adviser’s premiums paid table makes the same point structurally. A 25th percentile of 33.7% and a 75th percentile of 79.7% in a set of 33 deals spans 46 percentage points, so the interquartile range alone covers a wide band of possible per share outcomes. That is why the applied range of 34% to 80% produced implied values of $23 to $31 while the actual consideration was $41. The premiums paid work was explicitly labeled as informational and not material to the fairness opinion, and the same document also states three premium percentages in its list of reasons.
The Dutch auctions cut at the problem from the other side and expose a different limit. Because the issuer never states a premium, the only price signal is where the auction clears within the band. Both 2026 offers cleared at the maximum, which says the marginal seller would not go lower, and Scholastic’s shortfall against its own cap says how few sellers there were at that level. An auction that clears at the ceiling while leaving most of the authorized dollars unspent is a weaker demonstration of value than one that clears at the ceiling and is oversubscribed, because only the second forces the proration that shows genuine excess supply at the price.
What a careful reader would check, in either structure, is the same short list. Which date is the reference price drawn from, and was the stock already moving on rumour before it. Whether the headline consideration includes contingent components, as the Apellis offer does with its $4.00 of milestone payments, since those are not cash on the closing date. Where the price sits against the 52-week range rather than against yesterday. And in a self-tender, whether the offer was subscribed to its stated cap. Those four checks are all answerable from the filings themselves, and they change the reading of a premium far more than the premium’s own decimal place.
