What is loss of exclusivity?

Loss of exclusivity is the date a drug stops being the only version of itself for sale. After it, other companies may sell copies, and the price the original charged stops being the price the market pays.

It is almost never the patent expiry date, which is the date most people go looking for. Think of a gate with two locks. The patent office holds one key and the FDA holds the other, and the copies wait outside until both locks are open. The one that opens last sets the date, and that is the date a company puts in its annual report.

Merck writes the distinction into its own 10-K, in the flattest possible language: while a compound patent expiring generally results in loss of market exclusivity, "other patents may provide additional market exclusivity associated with certain aspects of the product that extends beyond the compound patent expiration."

The two locks have different keyholders

A patent begins when it issues and ends 20 years from the date the application was filed. Drug patents get filed years before anyone knows whether the molecule works, so a good deal of that term burns off during trials and FDA review. Congress patched this with patent term restoration, which gives some of the lost years back but caps the total so that the patent and the restored period together "do not exceed fourteen years" after approval. Patents can also be attacked. A generic company that thinks a patent is invalid can say so, and courts sometimes agree.

FDA exclusivity is a different animal. It is granted on approval, it runs from approval rather than from filing, and no one can litigate it away, because it is a bar on the agency rather than a property right. The statutory periods:

• Five years for a new chemical entity, during which no generic application may even be submitted, cut to four if the applicant is challenging a patent.
• Seven years for an orphan drug, running from the date of the approval.
• Twelve years for a biologic. A biosimilar application may not be made effective until twelve years after the reference product was first licensed, and may not even be filed for four.

The six months nobody explains

There is a fifth clock, and it is the one that surprises people, because it does not grant exclusivity at all. It moves the other clocks.

If a company runs the pediatric studies the FDA asks for, the pediatric exclusivity statute rewrites every period already running. Five years is "deemed" to be five and a half. Three years becomes three years and six months. The seven-year orphan period becomes seven years and six months. And the relevant patent date becomes "six months after the date the patent expires, including any patent extensions."

So a company can move the date its largest product faces competition by half a year, on every lock at once, by studying the drug in children. That is a real corporate decision with a number attached, and it is the sort of thing that sits in a filing without ever reaching a headline.

What a company actually tells you, and how to read it

Merck's most recent annual report names three dates in the space of a few sentences, and the way it names them differs in a way worth noticing.

Bridion loses US exclusivity in July 2026, after which Merck expects "a significant and rapid decline in U.S. sales" and says it will discontinue selling the drug in the United States by the end of the year. Januvia and Janumet lose theirs in May 2026, and Merck expects to "lose nearly all U.S. sales." Both are small molecules and both sentences are flat declaratives.

Keytruda gets different grammar. Merck says sales "will be materially negatively impacted by biosimilar competition between 2028 and 2029." A two-year window, and an adverb where the other two sentences had a quantity. Keytruda was 49% of Merck's total sales in 2025, so the vaguest sentence in the paragraph covers half the company.

The hedging is not evasion. Merck genuinely does not know, because Keytruda is a biologic, and biologics do not fall off cliffs the way pills do.

Why a pill and a biologic behave differently

A generic pill is swapped for the brand at the pharmacy counter, usually without anyone calling the doctor. A biosimilar only gets that treatment if the FDA has declared it interchangeable, which is a separate finding with its own evidence bar. The statute defines the term precisely: an interchangeable product may be substituted for the reference product "without the intervention of the health care provider who prescribed the reference product."

Without that finding, every switch needs a prescriber to decide to switch. Millions of individual decisions, made one appointment at a time, are slower than a pharmacist reaching for a different bottle.

Three years of Humira, from AbbVie's own table

Humira is the biggest test of this that exists, and AbbVie prints the answer in its revenue table. US net revenues were $12,160 million in 2023, $7,142 million in 2024, and $3,062 million in 2025.

That is a decline of about 75% and it took three years. Not a quarter, not eighteen months. The worst single year was 2025, the third one, at 57%. If your mental image of a patent cliff is a number falling off a table, the actual shape here is a staircase, and a shareholder who sold on the first step and one who sold on the third got very different prices.

Why it took that long is contested, and AbbVie discloses the argument itself. A class action filed on behalf of third-party payors alleges that AbbVie's rebating practices are impairing biosimilar competition with Humira, in violation of federal and state antitrust law. That is an allegation in a live case, not a finding, and AbbVie is contesting it. Whether slow erosion was the mechanism working as designed or something else is exactly the question the court has been asked.

The cliff is not the company

Here is the part that gets lost when loss of exclusivity is discussed as a disaster. Over the same three years that US Humira fell from $12,160 million to $3,062 million, a loss of roughly $9.1 billion, AbbVie's US sales of Skyrizi went from $6,753 million to $15,202 million and Rinvoq from $2,824 million to $5,940 million. Those two added about $11.6 billion in the United States while Humira was losing $9.1 billion.

The cliff was real. Every number in it is real. The company was bigger at the bottom of it than at the top, because it had spent the protected years building the drugs that would replace the one it was going to lose. That is what the protected years are for, and whether a company used them is a separate question from when its exclusivity ends.

What to check

• Which lock closes last. A patent expiry in a press release is not the date. The company's annual report gives the exclusivity date, and it is usually later.
• The verb. "Will lose nearly all US sales" and "will be materially negatively impacted" are management telling you how confident it is. Compare the sentences a company writes about different products in the same filing.
• Pill or biologic. A small molecule facing generics and a biologic facing biosimilars are not the same event, and only one of them resolves quickly.
• Whether an interchangeability finding exists. It decides whether substitution happens at the counter or in an appointment.
• What replaced it. The replacement products are usually already selling and already disclosed by name in the same revenue table. Their growth rate is checkable today, and the cash runway behind it is in the same filing.

This is analysis, not investment advice. Every figure above comes from the statute or from the company's own annual report, linked at the point it is used.