pith. sign in

arxiv: 2212.07384 · v5 · submitted 2022-12-14 · 💰 econ.GN · q-fin.EC

Valuing Pharmaceutical Drug Innovations

Gaurab Aryal , Federico Ciliberto , Leland E. Farmer , Ekaterina Khmelnitskaya This is my paper

Pith reviewed 2026-05-24 10:37 UTC · model grok-4.3

classification 💰 econ.GN q-fin.EC
keywords pharmaceutical drugsdrug valuationevent studydiscounted cash flowdrug development costspreclinical valuetherapeutic areas
0
0 comments X

The pith

Stock price reactions to drug announcements allow estimation of average drug value at $2.16 billion for small firms.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper develops a method to estimate the market value of pharmaceutical drugs by combining event studies of stock price responses to development announcements with a discounted cash flow model. Using this approach, it finds that the average value of a drug developed by small firms is $2.16 billion. At the preclinical stage, the risk-adjusted present discounted net value averages $50 million, implying an expected development cost of $38 million from the discovery stage. These estimates vary by therapeutic area and can support policies such as drug buyouts and targeted interventions.

Core claim

We propose a methodology to estimate the market value of pharmaceutical drugs by combining the event study method with a discounted cash flow model that infers drug values from stock market responses to drug development announcements. The average value of a drug developed by small firms is $2.16 billion. At the preclinical stage the risk-adjusted and present discounted average net value of drugs is $50 million. The expected drug development cost at the start of the discovery stage is $38 million. Values and costs are also estimated for several therapeutic areas, with applications to policies that support drug development through buyouts and targeted preclinical interventions.

What carries the argument

The combination of an event study of stock price responses to announcements with a discounted cash flow model that infers drug values from those responses.

If this is right

  • Drug values and costs can be estimated separately for therapeutic areas such as neoplasm and infections.
  • The estimates can be applied to design policies that support drug development through buyouts.
  • Targeted preclinical interventions can be evaluated using the stage-specific net values.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same announcement-based approach could be tested on other high-uncertainty R&D sectors to recover comparable value estimates.
  • Linking the derived values to post-approval sales or licensing data would provide an external check on the stock-reaction method.
  • Policymakers could use the $38 million discovery-stage cost figure to calibrate the size of subsidies or buyout offers needed to shift entry decisions.

Load-bearing premise

Stock price reactions to drug development announcements primarily reflect the incremental economic value of the specific drug rather than confounding news, market movements, or other firm events.

What would settle it

If stock price changes around announcements, after market adjustment, show no systematic correlation with eventual clinical success, revenues, or realized drug values, the valuation estimates would not hold.

read the original abstract

We propose a methodology to estimate the market value of pharmaceutical drugs. Our approach combines the event study method with a discounted cash flow model that infers drug values from stock market responses to drug development announcements. We estimate the average value of a drug developed by small firms (those below the 95th percentile of market capitalization) to be \$2.16 billion. At the preclinical stage, the risk-adjusted and present discounted average net value of drugs is \$50 million. Leveraging these estimates, we also determine the expected drug development cost at the start of the discovery stage to be \$38 million. We estimate values and costs for several therapeutic areas (e.g., neoplasm, infections) and explore applying these estimates to design policies that support drug development through drug buyouts and targeted preclinical interventions.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The paper proposes a methodology combining event-study analysis of stock price reactions to drug development announcements with a discounted cash flow model to estimate the market value of pharmaceutical innovations. It reports an average value of $2.16 billion for drugs from small firms (below 95th percentile market capitalization), a risk-adjusted preclinical net value of $50 million, and an implied expected development cost of $38 million at the discovery stage, with breakdowns by therapeutic area and policy applications such as buyouts.

Significance. If the identification strategy holds, the estimates would provide a data-driven basis for valuing early-stage drug projects, calibrating R&D incentives, and evaluating targeted interventions, addressing a gap between clinical success rates and economic valuations in the pharmaceutical innovation literature.

major comments (2)
  1. [Methodology and Data sections] The central estimates rest on the event-study identification that abnormal returns around announcements isolate the incremental cash-flow value of the announced drug. For the subsample of small firms emphasized in the abstract, this assumption is particularly vulnerable to confounding firm-level news, multi-project portfolio effects, or liquidity shocks; the manuscript must detail (in the methodology and data sections) the exact announcement-window cleaning procedure, handling of multi-drug firms, and any robustness checks that exclude overlapping events.
  2. [Section on DCF implementation] The mapping from abnormal returns to drug values via the DCF model requires explicit assumptions on risk adjustment, discount rates, and the fraction of value capitalized at each development stage. Without these parameters shown to be identified from external data rather than calibrated to match the headline figures, the $50 million preclinical net value and $38 million discovery cost cannot be assessed for sensitivity.
minor comments (2)
  1. [Abstract] The abstract states precise numerical results but provides no information on sample size, number of announcements, or firm coverage; this should be added for transparency.
  2. [Results tables] Therapeutic-area breakdowns are mentioned but not shown in the provided abstract; ensure tables report standard errors and sample sizes for each category.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive feedback. The comments highlight areas where additional methodological detail will strengthen the paper. We address each point below and will revise accordingly.

read point-by-point responses

  1. Referee: [Methodology and Data sections] The central estimates rest on the event-study identification that abnormal returns around announcements isolate the incremental cash-flow value of the announced drug. For the subsample of small firms emphasized in the abstract, this assumption is particularly vulnerable to confounding firm-level news, multi-project portfolio effects, or liquidity shocks; the manuscript must detail (in the methodology and data sections) the exact announcement-window cleaning procedure, handling of multi-drug firms, and any robustness checks that exclude overlapping events.

    Authors: We agree that greater transparency on these procedures is warranted to support the identification strategy for the small-firm subsample. The revised manuscript will expand the Methodology and Data sections to describe the announcement-window cleaning procedure in full, the treatment of multi-drug firms, and the robustness checks that exclude overlapping events. revision: yes

  2. Referee: [Section on DCF implementation] The mapping from abnormal returns to drug values via the DCF model requires explicit assumptions on risk adjustment, discount rates, and the fraction of value capitalized at each development stage. Without these parameters shown to be identified from external data rather than calibrated to match the headline figures, the $50 million preclinical net value and $38 million discovery cost cannot be assessed for sensitivity.

    Authors: The risk-adjustment factors, discount rates, and stage-specific capitalization fractions are taken from external sources in the pharmaceutical valuation literature. The revision will add a dedicated subsection documenting these parameters with citations, together with sensitivity analyses that vary each assumption independently of the headline estimates. revision: yes

Circularity Check

0 steps flagged

No significant circularity; estimates inferred from external market reactions

full rationale

The paper's core estimates ($2.16B average drug value, $50M preclinical net value, $38M discovery cost) are obtained by combining event-study abnormal returns around development announcements with a discounted cash flow model. This constitutes inference from independent stock-price data rather than any self-definitional loop, fitted-input prediction, or self-citation chain that reduces the outputs to the inputs by construction. No equations or steps in the provided abstract or description exhibit the enumerated circularity patterns; the derivation remains self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The valuation approach rests on standard financial-economics assumptions about market incorporation of information and the applicability of discounted cash flow to uncertain drug revenues; no free parameters or invented entities are explicitly named in the abstract.

axioms (2)
  • domain assumption Stock markets are semi-strong efficient with respect to public drug-development announcements
    Required for event-study price reactions to isolate drug value
  • domain assumption Drug revenues and failure probabilities can be represented by a standard multi-stage discounted cash flow model
    Underlies the inference of net present values from observed price changes

pith-pipeline@v0.9.0 · 5666 in / 1245 out tokens · 35576 ms · 2026-05-24T10:37:11.594598+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

17 extracted references · 17 canonical work pages

  1. [1]

    Nonparametric Inference in Connection with Multiple Decrement M odels,

    Aalen, Odd , “Nonparametric Inference in Connection with Multiple Decrement M odels,” Scandinavian Journal of Statistics , 1976, 3 (1), 15–27. Adams, Christopher P. and Van V. Brantner , “Estimating the Cost of New Drug Development: Is It Really $802 Million?,” Health Affairs , 2006, 25 (2), 420–428. Austin, David H. , “An Event-Study Approach to Measuring...

    work page 1976

  2. [2]

    Causes of Clinical Failures Vary Widely by Therapeutic, Phase of Study.,

    DiMasi, Joseph A. , “Causes of Clinical Failures Vary Widely by Therapeutic, Phase of Study.,” Impact Report, 2013, 15 (5). , Henry G. Grabowski, and Ronald W. Hansen , “Innovation in the Pharmaceutical Industry: New Estimates of R&D Costs,” Journal of Health Economics , 2016, 47, 20–33. Dubois, Pierre and Margaret K. Kyle , “Are Cancer Drugs Worth the Pr...

    work page 2013

  3. [3]

    The Behavior of Stock-market Prices,

    Fama, Eugene F. , “The Behavior of Stock-market Prices,” Journal of Business , 1965, 38 (1), 34–105. , Lawrence Fisher, Michael C. Jensen, and Richard Roll , “The Adjustment of Stock Prices to New Information,” International Economic Review , 1969, 10 (1), 1–21. 41 Food and Drug Administration , “Compliance Program Guidance Manual 7346.832,” https://www.f...

    work page 1965

  4. [4]

    Market Value, R&D, and Patents,

    Giliches, Zvi, “Market Value, R&D, and Patents,” Economics Letters, 1981, 7 (2), 183–187. Gupta, Charu N. , “One Product, Many Patents: Imperfect Intellectual Propert y Rights in the Pharmaceutical Industry,” Working Paper UCLA Anderson ,

    work page 1981

  5. [5]

    Market Value and Patent Citations,

    Hall, Bronwyn H., Adam Jaffe, and Manuel Trajtenberg , “Market Value and Patent Citations,” RAND Journal of Economics , 2005, 36 (1), 16–38. Hay, Michael, David W. Thomas, John L. Craighead, Celia Econ omides, and Jesse Rosenthal , “Clinical Development Success Rates for Investigational Drugs,” Na- ture Biotechnology, 2014, 32 (1), 40–51. Howard, David H.,...

    work page 2005

  6. [6]

    Techno- logical Innovation, Resource Allocation, and Growth,

    Kogan, Leonid, Dimitris Papanikolaou, Amit Seru, and Noah S toffman, “Techno- logical Innovation, Resource Allocation, and Growth,” Quarterly Journal of Economics , 2017, 132 (2), 665–712. 42 Kremer, Michael, “Patent Buyouts: A Mechanism for Encouraging Innovation,” Quarterly Journal of Economics , 1998, 113 (4), 1137–1167. and Rachel Glennerster , Strong ...

    work page 2017

  7. [7]

    Advance Market Commitments: Insights from Theory and Experience,

    , Jonathan D. Levin, and Christopher M. Snyder , “Advance Market Commitments: Insights from Theory and Experience,” American Economic Review P&P , 2020, 110, 269–

    work page 2020

  8. [8]

    Designing Advance Market Commitments for New Vaccines,

    , , and , “Designing Advance Market Commitments for New Vaccines,” Management Science, 2022, 68 (7), 4786–4814. Krieger, Joshua, Danielle Li, and Dimitris Papanikolaou , “Missing Novelty in Drug Development,” The Review of Financial Studies , 2022, 35 (2), 636–679. Laffont, Jean-Jacques and Jean Tirole , “Using Cost Observation to Regulate Firms,” Journal ...

    work page 2022

  9. [9]

    Rational Expectations, Information Acquisition, and Competitiv e Bidding,

    Milgrom, Paul R. , “Rational Expectations, Information Acquisition, and Competitiv e Bidding,” Econometrica, 1981, 49 (4), 380–391. Missialos, Elias, Chantal M Morel, Suzanne Edwards, Julia B erenson, Marin Gemmill-Toyama, and David Brogan , “Policies and Incentives for Promoting Inno- vative Antibiotic Research,” Technical Report, European Observa tory o...

    work page 1981

  10. [10]

    Parsing Clinical Success Rates,

    43 Mullard, Asher , “Parsing Clinical Success Rates,” Nature Reviews Drug discovery , jun 2016, 15 (7),

    work page 2016

  11. [11]

    Lessons from 60 Years of Pharmaceutical Innovation,

    Munos, B , “Lessons from 60 Years of Pharmaceutical Innovation,” Nature Reviews Drug Discovery, 2009, 8, 959–968. Myerson, Roger B. , “Optimal Auction Design,” Mathematics of Operations Research , 1981, 6 (1), 58–73. Ng, Rick , Drugs: From Discovery to Approval , Wiley,

    work page 2009

  12. [12]

    Patents as Options: Some Estimates of Value of Holding European Patent Stocks,

    Pakes, Ariel , “Patents as Options: Some Estimates of Value of Holding European Patent Stocks,” Econometrica, 1986, 54 (4). Pammolli, F., L Magazzini, and M. Riccaboni , “The Productivity Crisis in Pharma- ceutical R&D,” Nature Reviews Drug Discovery , 2011, 10, 428–438. Paul, Steven M., Daniel S. Mytelka, Christopher T. Dunwiddi e, Charles C. Persinger, ...

    work page 1986

  13. [13]

    Proof That Properly Anticipated Prices Fluctuate Randomly,

    Samuelson, Paul , “Proof That Properly Anticipated Prices Fluctuate Randomly,” Indus- trial Management Review , 1965, 6, 41–49. Scannell, Jack W., Alex Blanckley, Helen Boldon, and Brian W arrington, “Diag- nosing the Decline in Pharmaceutical R&D Efficiency,” Nature Reviews Drug Discovery , 2012, 11, 191–200. 44 Schrag, Deborah , “The Price Tag on Progress...

    work page 1965

  14. [14]

    Key Cost Drivers of Pharmaceutical Clinical Trials in the United States,

    , Hui-Hsing Wong, Amber Jessup, and Trinidad Beleche , “Key Cost Drivers of Pharmaceutical Clinical Trials in the United States,” Clinical Trials, 2016, 13 (2), Epub. Shin, H. S. , “Disclosures and Asset Returns,” Econometrica, 2003, 71, 105–133. Singh, Manish, Roland Rocafort, Cathy Cai, Kien Wei Siah, an d Andrew W. Lo, “The Reaction of Sponsor Stock Pr...

    work page 2016

  15. [15]

    Entry and Competitive Structure in Deregulated Airline Markets: An Event Study Analysis of People Expr ess,

    Whinston, Michael D. and Scott C. Collins , “Entry and Competitive Structure in Deregulated Airline Markets: An Event Study Analysis of People Expr ess,” RAND Journal of Economics , 1992, 23 (4), 445–462. Wong, Chi Heem, Kien Wei Siah, and Andrew W. Lo , “Estimation of Clinical Trial Success Rates and Related Parameters,” Biostatistics, 2019, 20 (2), 273–...

    work page 1992

  16. [16]

    success”). There are “competing risks

    and all subse- quent development-stage time probabilities. In that case, we can d etermine the probability of time to success from discovery to FDA approval as a product of those probabilities , which we denote by P(·). Then we can estimate the expected discount rate using Monte Ca rlo simulation, i.e., E(δdisc→ ) = E(δτ) ≈ 1 L ∑L ℓ=1 δτℓ , where τℓ ∼ P(·...

    work page 1976

  17. [17]

    Next, let N0(t) and N2(t) be the number of drugs in the development stage and for which the firms applied for approval at time t, respectively

    The probability of transit ion from state 0 to state 2 by time t is given by W (t) = 1 − exp ( − ∫ t 0 − Q′ 2(y) Q0(y) dy ) . Next, let N0(t) and N2(t) be the number of drugs in the development stage and for which the firms applied for approval at time t, respectively. Suppose we partition [0 , t] into small intervals with at most one transition in each su...

    work page 1976