rNPV Model

Risk-adjusted net present value is the standard valuation method for clinical-stage drug assets. A naive discounted-cash-flow model assumes the product reaches market; an rNPV model does not. It treats development as a sequence of gated bets, in which each clinical phase succeeds with some probability, and weights every future cash flow by the probability that the program is still alive to receive it. Costs are spent whether or not the program ultimately succeeds, so they are weighted only by the probability of reaching the stage that incurs them; revenue arrives only if the program is approved, so it is weighted by the cumulative probability of approval.

This calculation is the spine of the platform and it is indication-agnostic: the same arithmetic values an oncology antibody, a rare-disease gene therapy, or an anti-infective small molecule. What changes between assets are the inputs (the per-stage probabilities, the revenue and cost assumptions, the exclusivity profile), not the method. This section is the assembly: how those inputs become one risk-adjusted number. The probability inputs have their own reference section (PoS Calibration); the terminal-value treatment under the Inflation Reduction Act has its own section; and the commercial COGS, SG&A, and tax priors have their own benchmark sections.

The model is deterministic and version-stamped. Every figure here is produced by model 4.1.0, which is embedded in every signed export so a reader can reproduce the exact calculation.

Each development stage carries a probability of success (PoS). The cumulative probability that the asset reaches market is the product of the per-stage probabilities: clearing Phase I tells you nothing unless you then clear Phase II, Phase III, and regulatory review.

Each stage's PoS is prefilled from the indication × modality × biomarker reference matrix and adjusted in log-odds space by any published multiplier the scenario is marked for. Every value can be changed in the asset modeler, and the model uses the per-stage values you keep. The reference figures are documented in the PoS Calibration section.

Each stage has a cost and a duration. The model discounts each stage's cost to present value at the weighted average cost of capital (WACC), timing the cash outflow at the midpoint of the stage rather than its start or end, then weights it by the cumulative probability of reaching that stage.

cumulative_PoS_at_entry is the product of the probabilities of every prior stage. The cost is incurred if and only if the program survives to enter the stage, so it is not discounted by the stage's own probability. An optional launch cost is modeled as a single outflow twelve months before approval, discounted and weighted by the probability of reaching the final stage.

On approval, the model builds an annual revenue schedule with three phases:

• Ramp. Revenue climbs from launch toward peak over a user-set number of years. The default shape is an S-curve preset interpolated onto the chosen ramp length; a linear ramp is also available. The final ramp year reaches the stated peak.
• Plateau. Revenue holds at peak for the remainder of the on-market window.
• Loss-of-exclusivity (LOE) erosion. Near the end of exclusivity, revenue erodes. Small-molecule erosion is steep (a generic cliff: revenue falls to roughly a tenth of peak and then declines further); biologic erosion is gentler (retaining about 85% per year, reflecting biosimilar dynamics).

The on-market window is exactly the asset's exclusivity period: erosion happens within that window, eroding its final years, rather than being appended as extra years. This is a deliberate modeling choice (corrected in model 2.0.0) so that turning a patent cliff on can only ever reduce value relative to holding peak, never increase it.

Gross revenue is then converted to the after-cost, after-tax figure the valuation discounts. The operating margin is additive, the investment-banking convention, so cost of goods and SG&A are each taken as a share of gross revenue, and tax is applied to the operating result.

Tax. Tax is charged on each year's operating profit from sales. Development and launch costs are not deducted from taxable income and no loss is carried forward, so the model does not credit the tax saving a profitable company would get from its research spending. The effect grows with the tax rate and the size of the development budget.

The COGS%, SG&A%, and tax-rate defaults are sourced priors keyed on modality and commercial model (COGS, SG&A, Tax); the IRA Maximum Fair Price cliff is applied to gross revenue from the Year-9 (small molecule) or Year-13 (biologic) cliff onward before costs are taken, to US revenue only since model 4.0.0. Multi-market assets are modeled as a sum of per-geography schedules with their own launch timing and revenue shares. Each geography's schedule, including its exclusivity window, starts at that geography's own launch, so a geography that launches later also loses exclusivity later.

Every cash flow is discounted to present value at the WACC with annual compounding. Revenue is discounted from the end of the development period. Peak sales arriving years after a long Phase III are worth far less than the same sales arriving sooner, which is why duration assumptions matter as much as magnitude.

Development costs are discounted at each stage's midpoint and revenue at the end of each year. The mix is conservative: taking revenue at mid-year instead would raise its present value by about 5% at a 10% WACC.

The headline value combines the probability-weighted revenue with the probability-weighted cost.

Alongside the risk-adjusted number the model reports an Unadjusted NPV: the value the asset would have if approval were certain. The gap between the Unadjusted NPV and the rNPV is the clinical-risk discount: the part of the headline value that the probability of failure removes. It is shown as a reference, not the headline, because for a pre-approval asset it overstates value by assuming success.

When a deal structure is attached (royalty, milestone, or profit-split), the model first builds the same revenue and cost series, then transforms it into the partner's economic position (royalties on gross net sales, milestones recognized at their trigger gates, profit split on net revenue), and the rNPV is the deal-adjusted value rather than the full-product value.

Three conventions apply to every deal basis. An upfront payment counts in full at the valuation date, undiscounted. Royalties, upfronts and milestones your organisation receives are gross amounts before tax, while a profit share is taken on after-tax net revenue, so the two are on different tax bases and should not be compared line for line. Who bears development cost depends on the basis: under royalty only, the partner bears it; under milestone and royalty, the licensor bears all development and launch costs; under a profit split, each side bears its cost share; under the licensee basis, your organisation bears the share it states.

Since model 3.0.0 every deal term reads one revenue series: the worldwide schedule when no geographies are set, and the sum of each geography's share with its own launch delay when they are. Royalties, royalty tiers and the profit-split pro-forma follow the same timing as the headline, and the pro-forma's yearly deal income adds up to the deal value when no launch cost is set. With a launch cost, the yearly rows leave it out and the last row is aligned to the deal value, so the rows do not add up to it.

A stage milestone pays when its stage gate is cleared, weighted by the probability of reaching that gate. A sales milestone, one with a net-sales threshold, pays in the first forecast year whose net sales reach the threshold, weighted by the probability of approval and discounted like revenue; if the forecast never reaches the threshold, it pays nothing. A sales threshold can be set through the API; the asset modeler and "Model these terms" do not send one yet, so a sales milestone entered there is timed at a stage or not modelled. Before model 3.0.0, royalties used the worldwide schedule without launch delays even when geographies were set, and sales milestones were paid at approval. A scenario keeps the model version it was computed with.

Since model 4.1.0 a deal can also hold terms your organisation pays: upfronts, milestones and one royalty, on the licensee basis or under a profit split. The licensee basis values the whole revenue passed in, with the share of development cost your organisation states, net of the terms it pays; that share must be set, because who bears development cost is never assumed. A paid upfront counts in full at the valuation date, undiscounted, with no tax saving. A paid milestone has the same timing and probability weighting as a received one, also with no tax saving. A paid milestone whose stage gate names no stage in the plan, or that has neither a gate nor a sales threshold, is refused rather than timed at the last stage, where a received milestone in that position is still timed.

A paid royalty is charged on gross net sales, tiered or at a flat rate, weighted by the probability of approval and discounted like revenue, and it is counted after tax: each year it costs the royalty minus the tax rate times the smaller of the royalty and your organisation's operating profit, which is gross sales × (1 − COGS% − SG&A%), times your share under a profit split. The result lists each paid term with its value after tax and its tax saving. From term-model mapping v5, "Model these terms" sends the terms your organisation pays, and a paid purchase price in cash as a paid upfront; the asset modeler does not offer paid terms (see Modelling deal terms).

The deterministic rNPV is a single point estimate built from point-estimate inputs. Because clinical outcomes are binary and the inputs are uncertain, the true distribution of value is wide and not symmetric: most paths end at or near zero (failure) while a minority reach the full commercial value. PhaseFolio runs a Monte Carlo simulation that re-evaluates the model many times to produce that distribution, applying the IRA cliff inside each iteration so every draw carries the correct terminal compression.

What it samples. In each iteration, each stage's PoS is drawn from a Beta distribution centred on its input, and each stage's cost varies around its input. Stage durations vary only when duration uncertainty is turned on. Each stage then passes or fails (the sequential rpNPV mode, used in the workspace and in signed dossiers): a failed path earns no revenue and bears only the costs spent before it failed. Peak sales, ramp, exclusivity, the discount rate and the margins keep the scenario's values in every iteration, so the spread reflects development risk, not sales uncertainty. PoS adjustments for biomarker quality are applied to the deterministic rNPV but not yet inside the simulation, so for a scenario that uses them the simulated mean sits away from the headline.

The expected loss figure is the average of the simulated outcomes below zero, and the value if it succeeds is the average over paths that reach the market.

The defensible central estimate of a Monte Carlo run is the expected NPV (eNPV), the mean of the simulated distribution, not the median. Because the distribution is bimodal (a large mass near zero plus a commercial-success mode), the median can sit in the empty middle and misrepresents the asset; the mean correctly reflects the probability-weighted average outcome and reconciles with the deterministic rNPV. Confidence bands (for example the 10th-to-90th-percentile range) are reported to convey the spread rather than implying false precision in the point estimate.

To show which assumptions drive the value, the model runs a one-at-a-time (tornado) sensitivity: each input is varied across a plausible range while the others are held at their base case, and the resulting swing in rNPV is ranked. This isolates the highest-leverage assumptions for a reviewer to challenge. It is deliberately one-at-a-time rather than a global sensitivity analysis, so it does not capture interaction effects between simultaneously varying inputs, a known limitation stated below.

• Point-estimate inputs. The deterministic rNPV is only as good as its inputs; it does not itself express uncertainty. The Monte Carlo distribution is the place to read uncertainty, and even that samples the assumptions the analyst provides.
• Development-input provenance. The shipped per-stage cost and duration values remain editable product defaults, not the literature estimates. The separate profiles are broad therapeutic-area point estimates with 2018-USD costs, not exact-disease, modality-, biomarker-, or asset-specific observed evidence. Broad and limited proxies stay labeled; unsupported indications remain unavailable.
• Benchmark filter boundary. Indication can select a disclosed therapeutic-area cost-and-timing profile. Modality and biomarker can change PoS only and never modify the literature cost or duration values.
• One-at-a-time sensitivity. The tornado holds all-but-one input fixed and therefore does not capture interactions between simultaneously varying assumptions.
• Annual granularity. Cash flows are modeled annually with midpoint-timed development costs and year-end revenue; intra-year timing is not modeled.
• No tax saving on development spend. Development and launch costs are not tax-deductible in the model and losses are not carried forward.
• Received terms before tax, paid royalties after tax. A royalty your organisation receives counts in full, while the same royalty paid costs less after its tax saving: when operating profit covers it, a $1 royalty paid costs $0.79 at a 21% tax rate. The licensor's and the licensee's values for one deal therefore do not add up to the full-ownership value.
• No tax saving on paid lump sums. A paid upfront or milestone is not tax-deductible in the model, which understates the payer's value.
• Royalty tax saving capped at operating profit. The saving on a paid royalty is limited to each year's operating profit and is not carried forward.
• Paid royalties run for the whole forecast. A paid royalty is charged on every forecast year of sales. A royalty term that ends earlier is not modelled, so the royalty paid is overstated.
• Full-ownership pro forma is one worldwide schedule. The yearly cash-flow table for a full-ownership scenario ignores geography launch delays and applies any IRA cap to all revenue, so with geographies it approximates, and can sit below, the Unadjusted NPV the model computes.
• Stylized LOE erosion. The small-molecule and biologic erosion curves are stylized defaults, not asset-specific generic-entry forecasts; an analyst with a specific LOE view should override them.
• Decision support, not a fairness opinion. The model is built for portfolio triage, asset stress-testing, and term-sheet work. It is not calibrated for IFRS-13 / US-GAAP fair-value reporting, Section 409A valuations, or litigation-grade damages; treat any output as a working model a qualified valuation professional must independently challenge.