Pipeline SOTP and Peak Sales Estimation

Introduction

Building a complete biotech valuation requires aggregating the rNPV of each pipeline asset into a sum-of-the-parts framework, then adding net cash and subtracting corporate costs. This pipeline SOTP is the standard valuation approach in healthcare banking, equity research, and pharma business development. Unlike pharma SOTP (which values existing products using DCF and pipeline assets using rNPV), biotech pipeline SOTP is almost entirely rNPV-driven because most biotech companies have little or no product revenue.

The challenge is that the key input, peak sales estimation, has a documented average forecast error of approximately 71%, making the methodology only as good as the assumptions that feed it. This article covers the pipeline SOTP framework, the patient funnel methodology for peak sales estimation, how to build the revenue curve from peak sales to year-by-year projections, the role of multi-indication expansion, and the cross-checking process that validates the bottom-up analysis.

Each component serves a specific purpose: the sum of individual asset rNPVs captures the probability-weighted value of the pipeline, net cash provides the balance sheet floor, and the corporate cost NPV reflects the ongoing overhead required to operate the company during the development period. This formula applies to both standalone valuation (what is the biotech worth as an independent company?) and M&A valuation (what should an acquirer pay?), though the inputs differ because an acquirer eliminates corporate costs and may project higher peak sales due to commercial synergies.

Why Net Cash Matters So Much

For clinical-stage biotechs with no product revenue, net cash (cash plus short-term investments minus debt) is a floor value for the equity. If all pipeline assets fail, the company is worth approximately its net cash (minus wind-down costs, which typically consume 6-12 months of operating expenses for employee severance, lease terminations, and clinical trial wind-down). This creates a dynamic where net cash represents an unusually large percentage of market capitalization:

The declining percentage as companies advance reflects the increasing value the market assigns to the pipeline as clinical risk is progressively resolved through positive data. A Phase III company with a validated asset has substantial pipeline value layered on top of its cash position, so cash becomes a smaller fraction of total market cap.

Cash Runway

The number of months a biotech can continue operating at its current cash burn rate before running out of money. Calculated as net cash divided by quarterly cash burn (typically operating expenses plus capital expenditures minus any revenue). Cash runway is one of the most important metrics in biotech investing because a company that runs out of cash before its next value-inflecting catalyst is forced into dilutive financing, a fire-sale partnership, or liquidation. Healthcare bankers monitor cash runway closely: companies with less than 12 months of runway face existential financing pressure, while companies with 24+ months of runway have strategic optionality to wait for better market conditions or negotiate from a position of strength.

Peak sales is the single most impactful assumption in an rNPV model. A doubling of peak sales approximately doubles the rNPV (all else equal). The standard approach is the patient funnel, a top-down-to-bottom-up framework that starts with disease prevalence and applies successive filters to arrive at the treatable patient population and expected revenue per patient.

Scenario Analysis for Peak Sales

The weighted average peak sales flows into the rNPV calculation, producing a more robust valuation than any single scenario alone. In practice, healthcare bankers present all three scenarios and the weighted average, allowing the reader (whether a PE investment committee, a pharma BD team, or a biotech board) to form their own view on which scenario is most likely.

Peak sales is a single number representing one year. The rNPV model requires a year-by-year revenue projection covering the entire commercial life of the drug (typically 12-15 years from launch to loss of exclusivity and generic/biosimilar erosion). Building this revenue curve from the peak sales estimate requires modeling three phases:

Launch ramp (years 1-4). Revenue grows as physicians adopt the drug, payers add it to formularies, and the company's sales force builds awareness. The ramp shape depends on the disease area and competitive context. Oncology drugs in areas of high unmet need (no effective alternatives) ramp fastest, sometimes reaching 70-80% of peak sales by year 2. Drugs in crowded markets with established competitors ramp more slowly, taking 4-5 years to reach peak. The ramp assumptions significantly affect the NPV because earlier revenue is worth more in present value terms.

Peak plateau (years 4-8). Revenue stabilizes near the peak as the drug reaches maximum penetration in its approved indications. The plateau duration depends on patent protection remaining, the timing of competitive entry, and whether the company secures additional indication approvals (which can extend or elevate the peak). A drug with 10 years of remaining patent life at launch has a longer plateau than one with 7 years.

Decline phase (years 8-15+). Revenue declines as patent protection expires and generics or biosimilars enter the market. The decline shape depends critically on whether the drug is a small molecule (steep decline, 80-90% erosion in 18 months) or a biologic (gradual decline, 30-50% erosion over 3-5 years). This distinction can represent billions of dollars in NPV difference.

Revenue Curve Shape

The year-by-year revenue trajectory from launch through loss of exclusivity. The standard revenue curve resembles a hill: revenue climbs during the launch ramp, plateaus near peak, then desclines post-LOE. The NPV of this curve is highly sensitive to two parameters: the ramp speed (how quickly revenue reaches peak, which affects the value of early years when discounting is less severe) and the decline shape (how steeply revenue falls post-LOE, which determines the terminal revenue tail). Healthcare bankers model the revenue curve explicitly rather than simply applying peak sales to a flat projection, because the curve shape can change the asset's rNPV by 30-50% even when peak sales assumptions are held constant.

One of the most important drivers of peak sales (and the most common source of forecast outperformance) is label expansion into additional indications beyond the initial approval. A drug approved for second-line metastatic breast cancer may subsequently win approvals for first-line breast cancer, lung cancer, gastric cancer, and other tumor types, each adding incremental patient populations and revenue.

Multi-indication expansion creates a "stacking" effect where the peak sales of the overall franchise (all indications combined) far exceeds the peak sales of the initial indication alone. Keytruda (pembrolizumab) was initially approved for advanced melanoma with an initial indication peak sales estimate of approximately $3-5 billion. As of 2024, Keytruda is approved in 30+ indications across numerous tumor types and generates over $25 billion in annual revenue, a 5-8x outperformance versus the initial single-indication forecast.

Healthcare bankers model multi-indication expansion by assigning separate patient funnels, probability of success estimates, and revenue curves to each potential indication. The pipeline SOTP for a single drug can therefore include multiple "sub-assets," each representing a different indication at a different clinical stage with its own PoS and peak sales estimate. This granularity is essential for drugs with broad mechanism-of-action applicability (checkpoint inhibitors, ADCs, bispecific antibodies) where indication expansion is the primary value driver.

Common Peak Sales Estimation Mistakes

Several systematic errors plague peak sales estimation that healthcare bankers should be aware of:

Overestimating the addressable population. Epidemiological prevalence data often overstates the treatable population. Not all patients with a disease are diagnosed, not all diagnosed patients are treated, not all treated patients are eligible for a new therapy (some have contraindications, comorbidities, or are stable on current treatment), and not all eligible patients will switch from existing treatments. Each filter reduces the addressable population, sometimes dramatically. The gap between "epidemiological prevalence" and "commercially addressable patients" is routinely 60-80%.

Ignoring competitive dynamics. Peak sales models built assuming limited competition may prove wildly optimistic if three or four competitors launch within the same window. In oncology, where dozens of companies often target the same pathway, the competitive landscape can shift between the time a model is built (Phase II) and the time the drug launches (3-5 years later). Best practice is to build a competitive timeline mapping all known competitors by expected launch date and modeling dynamic market share rather than static assumptions.

Confusing peak sales with steady-state sales. Peak sales occur in a specific year; revenue before and after the peak is lower. The revenue curve shape (how quickly the drug ramps, how long the peak persists, and how steeply it declines post-LOE) matters enormously for the NPV calculation but is often given less attention than the peak number itself. Two drugs with identical $2 billion peak sales but different curve shapes (one with a 3-year ramp and steep decline, another with a 5-year ramp and gradual decline) will have materially different NPVs.

Underestimating gross-to-net deductions. Many peak sales models use list price (WAC) rather than net price. For drugs with significant Medicaid exposure, 340B exposure, or PBM rebate pressure, net price can be 30-50% below WAC. Modeling peak sales at WAC and then discounting in the rNPV overstates the revenue stream that actually reaches the company.

After building the pipeline SOTP, healthcare bankers cross-check it against market-based indicators to validate the bottom-up analysis:

Corporate Cost NPV

The present value of ongoing corporate overhead (G&A, management salaries, facilities, public company costs, and non-program-specific R&D) that the biotech will incur regardless of which pipeline programs succeed. For clinical-stage biotechs, corporate costs are typically $50-100 million per year and represent a drag on equity value of $300-600 million in NPV terms, depending on the discount rate and assumed timeline to profitability. Corporate cost NPV is sometimes overlooked in simple models but can represent 10-20% of a clinical-stage biotech's total valuation. When a pharma company acquires a biotech, corporate costs are largely eliminated (the acquirer already has corporate infrastructure), which is why acquirers' internal valuations of a biotech target are often $300-600 million higher than the standalone pipeline SOTP. This "corporate cost elimination" is a genuine synergy in biotech M&A and partially explains why acquisition premiums appear so high relative to standalone trading values.

The next article covers comparable company and transaction analysis for biotech, which provides the market-based cross-check for the rNPV-derived pipeline SOTP.