Pharma case interview, valuing a drug that may never launch

Halvard Pharma, an invented drug company, makes $1.2 billion a year from one drug whose U.S. exclusivity ends in 3 years. A biotech offers to license Halvard a drug that has just finished Phase 2, for $150 million up front. Halvard would pay for Phase 3 and keep the profits. Should it sign? Halvard, its drugs and its numbers are Soreno's invention, for practice. The one outside input is FDA's rough share of Phase 3 drugs that move on, used as the range of approval odds. Money is in millions of dollars and years count from today.

1. Clarify

   What is the license meant to fix?

   Halvard's goal is to replace sales it will lose. Its best seller makes $1.2 billion a year and loses U.S. exclusivity after year 3. If it keeps 20% of its sales in the first year after generics arrive (our assumption), $960 million of yearly sales disappear by year 4. Ask what the new drug costs to finish, when it could launch and how big it could be.

2. Lay out the years

   Phase 3 costs $360 million, launch in year 5

   Phase 3 costs $120 million a year for 3 years. If it succeeds, filing and launch work cost $40 million in year 4, sales start in year 5 and reach a peak of $1 billion a year in year 8. Halvard keeps 60% of sales after cost of goods, selling and royalties, $600 million a year at peak. Exclusivity ends after year 12, and sales fall to 20% then 10% of peak as generics arrive.

3. Unrisked value

   $1.43 billion, if approval were certain

   Discounted at 10% a year, the contribution is worth $1.76 billion today, against $298.4 million of Phase 3 cost and $27.3 million of filing cost. Net, $1.43 billion. Against a $150 million fee, that looks like a bargain. It isn't, because approval is far from certain.

4. Weight by the odds

   Risk-adjusted: $134.5 million to $221 million

   FDA's clinical research page says about 25 to 30% of drugs move on from Phase 3. Phase 3 is paid either way, but filing cost and every year of sales happen only if it succeeds. So the value is 25% × $1,731.5 million − $298.4 million = $134.5 million at 25%, and $221 million at 30%.

5. Compare with the ask

   The answer flips inside FDA's range

   At 25% the drug is worth $15.5 million less than the $150 million fee. At 30% it is worth $71 million more. The fee breaks even at about a 25.9% chance of success. When the decision turns on a probability no one knows to the point, the deal terms should carry the risk, not Halvard's guess.

6. Restructure

   $75 million now, $150 million on approval

   Offer $75 million up front and $150 million more if the drug is approved at the end of year 4. Weighted and discounted, that costs $100.6 million at 25% and $105.7 million at 30%, below the drug's value at either end. The biotech gets more in total if the drug works, and Halvard pays less if it fails.

7. Recommend

   Sign on milestone terms, and keep looking

   License the drug, with most of the price tied to approval. But don't call the cliff solved: the new drug launches in year 5, after Halvard's best seller has already gone generic, and weighted by the odds its peak is worth $250 million to $300 million of yearly sales, against a $960 million hole. Halvard also needs a drug that is already approved or filed. Next step: test peak sales against a patient count and a net price that payers will accept.

Short definitions, with the source named where an agency or a company states it. Anything unlabeled is our plain-words summary.

1. Clinical phases

   Safety, then efficacy, then proof at scale

   FDA's clinical research page describes Phase 1 with 20 to 100 people, Phase 2 with up to several hundred and Phase 3 with 300 to 3,000, lasting 1 to 4 years. It puts the share that moves on at about 70% after Phase 1, 33% after Phase 2 and 25 to 30% after Phase 3. The page was last updated January 4, 2018, so treat these as rough.

2. Patent

   20 years from filing, set by statute

   FDA's patents and exclusivity FAQ says a new U.S. patent runs 20 years from the date the application was filed, and that many other factors can affect how long it lasts. The same FAQ says patents can be granted at any time during a drug's development, so the years of protected selling left at launch can be far fewer than 20.

3. Exclusivity

   A separate shield FDA grants on approval

   The same FAQ lists, among others, 5 years for a new chemical entity, 7 years for an orphan drug and 6 months of pediatric exclusivity added to existing patents and exclusivity. Patents and exclusivity may or may not run at the same time.

4. Generics

   Prices fall as makers enter

   In a December 2019 study of drugs that first faced generics from 2015 to 2017, FDA found median generic prices 39% below the brand's pre-entry price with one generic maker, 54% with two and 79% with four, using manufacturer price data. With six or more, prices were more than 95% lower.

5. Biosimilar

   The generic of a biologic, almost

   FDA describes a biosimilar as a biologic that is highly similar to an approved original biologic, the reference product, with no clinically meaningful differences, but not identical to it.

6. Risk-adjusted NPV

   Cash flows × odds × discount

   Weight each year's cash flow by the chance the drug is still alive then, then discount it. Merck's 10-K says it values acquired research projects on probability-adjusted future net cash flows that reflect each product's stage of development, the same idea in an annual report.

7. Net price

   List price minus what payers claw back

   Revenue in a case should be net of the discounts and rebates paid to insurers and government programs. A peak sales figure built on list price overstates the drug. In the U.S., pharmacy benefit managers negotiate many of those rebates for insurers and decide where a drug sits on a formulary, the list of drugs a plan covers.

Figures to test a pharma case's assumptions against. CMS's figures are for 2024, from its National Health Expenditure fact sheet, which CMS updates as each new year of data comes out, so check for a newer year before you quote them.

1. Spending

   $467.0 billion on prescription drugs

   CMS says U.S. prescription drug spending rose 7.9% to $467.0 billion in 2024, out of $5.3 trillion of total health spending. That is about 8.8% of the total, our division. CMS projects retail prescription drug spending to grow fastest of the major categories from 2025 to 2034, at 5.7% a year on average.

2. Payers

   Private insurance, Medicare, Medicaid

   Of total 2024 health spending, CMS puts private health insurance at 31%, Medicare at 21% and Medicaid at 18%. Those shares are for all health care, not just drugs, but they show why a launch case has to ask which payer covers the patients.

3. Concentration

   Merck: one drug, 49% of 2025 sales

   Merck's 10-K for 2025 says Keytruda represented 49% of its total sales. It says it expects Keytruda sales to be materially hurt by biosimilar competition between 2028 and 2029, which could begin in December 2028, when the primary compound patent expires. That is the patent cliff in one filing, as Merck described it on February 24, 2026.

4. The fall

   Nearly all U.S. sales, expected to go

   The same 10-K said Januvia and Janumet would lose U.S. market exclusivity in May 2026 and that Merck expected to lose nearly all U.S. sales of both after that, after a fall in Januvia sales in the first half of 2026 that it tied to government price setting. Use it when a case assumes a brand keeps most of its sales after generics arrive.

These first moves are ours. Each one says which number to build before any list of factors.

1. Launch price

   Value against the alternative, then the payer

   Start from what the drug does better than today's treatment, price against that treatment, then ask what insurers will cover and what rebates they will demand. Cost to make the drug barely matters.

2. License or buy a drug

   Risk-adjusted NPV against the price

   Lay out costs and sales by year, weight them by the odds of each phase and discount them. Then shape the deal so payments follow the risk, as in the Halvard case.

3. Patent cliff

   Size the hole, then the timing

   How much yearly revenue goes, how fast, and which pipeline or acquired drugs land in time to replace it. Timing is usually the harder part.

4. Launch sizing

   Patients who will actually be treated

   Prevalence, then diagnosed, treated and eligible, then peak share and net price per year. Trinity's practice case is built this way.

5. Sales force

   Prescribers, calls and reps

   Count the doctors who treat the condition, decide how often each should be visited, divide by the calls one rep can make in a year. Specialists in a few centers need far fewer reps.

6. Biosimilar or generic entry

   Price drop against share won

   Estimate how far prices fall as makers enter, what share an entrant can win and what it costs to develop and make. Being one of the first few makers is worth far more than being the sixth.

7. Pipeline choices

   Value per dollar of development

   Rank drugs by risk-adjusted value per dollar still to spend, then check strategic fit, how well each fits the company's sales force, and spread of risk.

One opening misses the risk, one builds it in. We wrote both.

1. Weak opening

   A market-entry checklist

   "I'd look at the market size, competitors, the drug's efficacy, pricing and Halvard's capabilities, then decide." Every item is fair, but it never mentions that the drug might fail Phase 3 or when its sales would arrive, so it can't say whether $150 million is a fair price.

2. Strong opening

   Odds, years, then the price

   "Two questions decide this: what the drug is worth once I weight its sales by the chance it gets through Phase 3, and whether it arrives in time to replace the best seller. I'd lay out costs and sales by year, apply the Phase 3 odds and discount it, and compare that with the $150 million fee. If the answer depends on the odds, I'd change the deal terms rather than the guess."

Practice cases set in pharma on the firm pages we read on October 31, 2026. Our healthcare consulting guide covers how these firms interview.

1. L.E.K.

   Large Pharma Co., a baldness pill

   On L.E.K.'s interview preparation page: a large drug company has a daily pill that regrows hair within three months. Estimate the market and describe how you would price it.

   L.E.K.: Interview Preparation

2. ZS

   Pricing a COPD drug

   ZS's second practice case asks how to price a drug for chronic obstructive pulmonary disease that showed a best-in-class drop in flare-ups in trials. It works through patient numbers, doses per year, two competitors and whether to double production.

   ZS: Practice case two

3. Trinity Life Sciences

   Two Phase 3 assets, which to pursue

   Trinity's practice case gives a mid-size company a choice between a hyperhidrosis drug and a wet AMD biosimilar, both in Phase 3. Its answers, which we read on October 30, 2026, put their peak revenue at about $200M and $300M.

   Trinity: Practice case study

4. ClearView

   Inhaled insulin, a $300M goal

   ClearView's first practice case asks whether a mid-sized company's inhaled insulin can reach $300M a year of U.S. peak revenue, starting from the size of the insulin market.

   ClearView: Practice case #1