Vose SoftwareModelRisk example · Pharmaceuticals and biotech
Drug Development Risk-Adjusted NPV
What is a compound in Phase II worth — and should we develop it ourselves or license it out?
From the model’s saved run
What is the compound worth?
NPV today, US$m. rNPV: the industry's stage-weighted calculation. Simulated: the mean of 10,000 development paths.
| Strategy | rNPV textbook | Mean NPV simulated | Chance of a loss | 1 in 20 worst (P5) | 1 in 20 best (P95) |
|---|---|---|---|---|---|
| Develop it alone | 76.7 | 61.5 | 82% | −143.1 | 713.2 |
| License it after Phase II | 53.5 | 45.5 | 65% | −31.8 | 379.2 |
| Chance that licensing does better | 20% |
S-curve (cumulative probability) of Develop alone, License after Phase II from the saved simulation.
The business question
A biotech company's lead compound is entering Phase II for a chronic inflammatory disease. Phase II takes about two years and US$30 million and succeeds about one time in three; Phase III takes about three years and US$150 million and succeeds 60% of the time. The regulator approves nine filings in ten.
A launched product would grow to its peak sales in about four years — between US$300 million and US$1.2 billion a year — until the patent expires at the end of 2041, when generics take most of the market.
A larger company offers to license the compound if Phase II succeeds: US$60 million on signing, US$40 million on Phase III success, US$60 million on approval and a 15% royalty, while it pays for Phase III, the filing and the launch. The board wants to know what the compound is worth, and whether to accept the offer.
- Who it is for
- Biotech and pharma business development, portfolio and finance teams; investors in drug development.
- Main outputs
- NPV of developing alone and of licensing: the textbook rNPV and the simulated mean, the chance of a loss, 1-in-20 outcomes · what each outcome is worth — failure in Phase II or III, rejection, launch · the stage-by-stage rNPV · how the value grows at each milestone.
How the model works
Three sheets after this one, read from left to right. Each one feeds the next.
Inputs sheet
For Phase II, Phase III and the regulatory review: duration and cost (PERT) and the chance of success. The peak sales (Lognormal from P10 and P90), the years to peak, the patent expiry and the sales left after it, the operating margin and the launch cost. The licensing offer. The discount rate.
Development sheet
VoseBernoulli decides whether each phase succeeds. Each phase's cost is spread over its years. The product launches when the review ends, so a late approval leaves fewer years before the patent expires. Cash flows for developing alone and for licensing after Phase II are discounted to today.
The textbook rNPV repeats the calculation on the most likely path and weights each stage by the chance of reaching it.
Results sheet
For both strategies: the rNPV, the simulated mean NPV, the chance of a loss and the 1-in-20 outcomes. The chance of each outcome and what the compound is worth in it; the rNPV stage by stage; the value once each milestone is passed.
ModelRisk techniques
Find similar models
- Product
- Type
- Application
- Uses
The results in detail
What happens, and what it is worth
The chance of each outcome, the mean NPV when it happens, and its contribution to the mean NPV (US$m).
Histogram of NPV, developed alone from the saved simulation.
| Outcome | Chance | Alone NPV if it happens | Licensed NPV if it happens | Alone share of mean | Licensed share of mean |
|---|---|---|---|---|---|
| Fails in Phase II | 65.1% | −27.9 | −27.9 | −18.2 | −18.2 |
| Fails in Phase III | 14.7% | −136.6 | 21.4 | −20.1 | 3.1 |
| Rejected by the regulator | 2.1% | −139.2 | 46.3 | −3.0 | 1.0 |
| Launched | 18.1% | 566.6 | 328.7 | 102.7 | 59.6 |
| All outcomes | 100.0% | 61.5 | 45.5 |
Share of mean: the outcome's chance × its mean NPV. The shares add up to the mean NPV.
The textbook rNPV, stage by stage
Most likely path, expected peak sales; each stage weighted by the chance of reaching it.
| Measure | Unit | Value |
|---|---|---|
| Phase II | US$m | −27.3 |
| Phase III, × the chance of starting it | US$m | −37.7 |
| Regulatory review, × the chance of filing | US$m | −0.6 |
| Market, × the chance of launching | US$m | 142.4 |
| rNPV, developed alone | US$m | 76.7 |
| NPV if every stage succeeds (unrisked) | US$m | 615.2 |
| Simulated mean NPV, developed alone | US$m | 61.5 |
| Simulated mean less rNPV | US$m | −15.2 |
How the value grows at each milestone
Expected NPV today of the futures that pass each milestone (US$m), from the outcome table.
| Measure | Unit | Value |
|---|---|---|
| Today, before Phase II | US$m | 61.5 |
| Once Phase II succeeds | US$m | 228.0 |
| Once Phase III succeeds | US$m | 492.4 |
| Once approved | US$m | 566.6 |
| Licensed, once Phase II succeeds | US$m | 182.3 |
How to read these results
The textbook rNPV — each stage's value on the most likely path, weighted by the chance of reaching it — values the compound at about US$77 million if the company develops it alone.
The simulated mean is lower, about US$61 million in this run: development often runs late, and every month late is a month less of sales before the patent expires, which a single most likely path cannot see. (About ±6 of the gap is sampling error at 10,000 samples.)
- The mean hides the shape. The company loses money in 82% of futures — about US$28 million if Phase II fails, about US$137 million if Phase III fails — and all the value comes from the 18% in which the compound launches, worth about US$570 million.
- Licensing after Phase II is worth less on average, about US$46 million, but the worst 1-in-20 outcome falls from a loss of about US$143 million to one of US$32 million: the partner carries the Phase III risk. A company that could not survive a Phase III failure should take the offer. One that could might develop alone, or negotiate a higher royalty.
- The value steps up at each milestone: from about US$61 million today to about US$228 million once Phase II succeeds and about US$492 million once Phase III succeeds. These steps are what investors pay for, and why biotech companies raise money after good data.
- A rank-correlation tornado misleads here: it ranks the Phase II cost above Phase II success, because a successful Phase II leads either to the largest losses (a failed Phase III) or to the largest gains, and their ranks cancel. Values by outcome and by milestone tell the story better.
See it move. Open the workbook, change an input on the Inputs sheet and run the simulation: every table and chart on this page updates in Excel.
Inputs and assumptions
Every assumption in the model, with its uncertainty and where it comes from. The figures are illustrative.
Development 9 inputs
Each phase's duration and cost, and its chance of success.
| Assumption | Unit | Uncertainty | Low | Most likely | High |
|---|---|---|---|---|---|
| Phase II: durationRecruitment and 24 weeks' treatment. | years | PERT | 1.50 | 2.00 | 3.00 |
| Phase II: cost | US$m | PERT | 25.0 | 30.0 | 40.0 |
| Phase II: chance of successBenchmarks for the therapeutic area: 30–40%. | probability | Fixed | 35% | ||
| Phase III: durationTwo pivotal trials. | years | PERT | 2.50 | 3.00 | 4.50 |
| Phase III: cost | US$m | PERT | 120.0 | 150.0 | 200.0 |
| Phase III: chance of success | probability | Fixed | 60% | ||
| Regulatory review: durationFiling to approval. | years | PERT | 0.80 | 1.00 | 1.50 |
| Regulatory review: cost | US$m | PERT | 4.0 | 5.0 | 7.0 |
| Regulatory review: chance of approval | probability | Fixed | 90% | ||
The market 9 inputs
| Assumption | Unit | Uncertainty | Low | Most likely | High |
|---|---|---|---|---|---|
| Peak salesThe commercial team's forecast: 10% chance below 300, 10% above 1,200. | US$m a year | Lognormal · P10, P90 | 300 | – | 1,200 |
| Years from approval to peak salesSales grow in a straight line from approval to peak. | years | Fixed | 4.0 | ||
| Patent expires at the end of | year | Fixed | 2041 | ||
| Sales after expiry: first yearGeneric entry. | of peak | Fixed | 35% | ||
| Sales after expiry: second year | of peak | Fixed | 15% | ||
| Sales after expiry: third year | of peak | Fixed | 10% | ||
| Sales after expiry: later years | of peak | Fixed | 5% | ||
| Operating marginAfter cost of goods, marketing and distribution. | of sales | Fixed | 45% | ||
| Launch costIn the year of approval. | US$m | Fixed | 80 | ||
Licensing offer 4 inputs
If Phase II succeeds, a partner takes the compound and pays for everything after Phase II.
| Assumption | Unit | Value |
|---|---|---|
| On signing, after Phase II | US$m | 60 |
| Milestone on Phase III success | US$m | 40 |
| Milestone on approval | US$m | 60 |
| Royalty on sales | of sales | 15% |
Finance 2 inputs
| Assumption | Unit | Value |
|---|---|---|
| Discount rateCost of capital for development-stage biotech. | per year | 10.0% |
| Phase II starts on 1 January | year | 2027 |
Units and conventions
- Money
- US$ million (US$m), before tax; cash flows discounted from the middle of each year.
- Time
- Year 1 is 2027; Phase II starts on 1 January 2027.
- rNPV
- Risk-adjusted NPV: each stage's cash flows on the most likely path, weighted by the chance of reaching that stage, with the expected peak sales.
- NPV, simulated
- The NPV of each simulated path; its mean is the compound's expected value.
Key assumptions and limitations
- Phases
- Back to back; a failed phase ends development, and nothing is recovered.
- Sales
- Grow in a straight line from approval to peak. After the patent expires they fall to the shares given; no tax, no terminal value after 2046.
- Licensing
- The partner pays everything from Phase III on; payments arrive in the year of the event.
- Data
- The company, the compound and all figures are invented for this example.
How to use it
- Enter your phases' durations, costs and success rates — from your plan and industry benchmarks.
- Enter the peak-sales forecast as a P10 and a P90, the uptake, the patent expiry and the margin.
- Enter the terms of the licensing offer, then run a simulation (10,000 samples with a fixed seed are set).
- Compare the strategies on value and on risk: the company that cannot survive a US$140 million loss may prefer the lower value with the smaller downside.
ModelRisk functions used
- VoseBernoulli
- Whether each phase succeeds.
- VosePERT
- Durations and costs of the phases.
- VoseLognormalAlt
- Peak sales from the forecast's P10 and P90.
- VoseInput, VoseOutput
- Mark the uncertain inputs and the outputs.
- VoseSimMean, VoseSimPercentile, VoseSimProbability
- Statistics of the simulation on the Results sheet.
Further reading
- Monte Carlo simulation explained (Vose guide)
- Stewart, J. J., Allison, P. N. & Johnson, R. S. (2001). Putting a price on biotechnology. Nature Biotechnology 19, 813–817.
- BIO, Informa & QLS (2021). Clinical Development Success Rates 2011–2020.
- Vose, D. (2008). Risk Analysis: A Quantitative Guide, 3rd edition. Wiley.
Download the workbook and run it yourself. It opens in Excel with ModelRisk. The free trial runs every example model.
Related example models
- ModelChoice★★★★★Clinical Trial Go/No-Go DecisionAfter Phase I: run Phase II, partner the compound, or shelve it?Four decisions and eight chance events. Best first decision: Proceed to Phase II, expected value $135.1m.
- ModelChoice★★★★★Pharmaceutical Pipeline DecisionDevelop drug A, drug B or both, out-license both, or keep the money?Three decisions and ten chance events, driven by linked inputs. Best first decision: Develop Both (Parallel), expected value $776.7m.
- ModelRiskOptimizer★★★★★R&D Portfolio Chosen by the OptimizerWhich projects to fund for the most value, with limits on budget risk and on the chance of no launch?Eight projects with phase-by-phase attrition and correlated Phase II results; the portfolio chosen by the Optimizer from 256 combinations.
An illustrative example: the figures are invented for teaching and do not describe a real company, project or asset. Last updated 9 October 2026.