a16z Podcast: The Economics of Expensive Medicines

Episode Summary

Executive Summary: Andrew Lo and Jorge Conde examine how gene/cell therapies and other high-value cures should be priced, financed, and insured, arguing that one-time treatments should be amortized like mortgages rather than paid upfront. They also explore why drug development is simultaneously getting less risky scientifically and more risky financially, and why biotech markets behave more like evolving biological systems than static efficient markets.

Main Topics: Pricing cures as lifetime value (Priority: 5/5): The discussion frames gene and cell therapies as cures that deliver a stream of future health benefits, analogous to buying a house rather than renting. This leads to the idea of amortizing payments over time instead of paying a single upfront lump sum. Insurance and payer design for one-time therapies (Priority: 5/5): Lo argues insurers should pay for cures, but current plan-switching dynamics create financing mismatches. He proposes drug mortgages that follow the patient across plans to preserve incentives and spread costs fairly. How to value a cure (Priority: 5/5): The speakers debate whether value should be based on QALYs, avoided costs, taxes paid, or broader societal effects. They favor objective cost-effectiveness frameworks such as NICE and ICER, while warning against cherry-picking metrics. Drug discovery risk is changing over time (Priority: 4/5): Lo explains that scientific progress can increase financial risk by making older approaches obsolete and creating more uncertainty around investment timing, even as actual path-to-approval success rates improve in some areas. Evidence from oncology, Alzheimer’s, and vaccines (Priority: 4/5): Specific therapeutic areas illustrate divergent economics: oncology success rates have risen recently, Alzheimer’s remains extremely difficult, and vaccines have decent success probabilities but weak commercial incentives. Markets as biological systems (Priority: 4/5): Lo argues that healthcare and capital markets should be understood through evolutionary biology rather than pure physics, because competition, adaptation, and feedback loops drive boom-bust patterns and rapid change.

Key Arguments: Cures should be financed over time because their value accrues over a patient’s lifetime, just as a mortgage matches housing payments to the life of the asset. Health insurers are the natural payer for cures, but patient mobility across plans creates a free-rider problem unless payment obligations can transfer with the patient. The Affordable Care Act’s preexisting-condition logic may eventually need to account for preexisting financial liabilities created by prior cure payments. Value of a cure should start with patient benefit, then be assessed with objective frameworks like QALYs and cost-effectiveness analysis, rather than ad hoc or politically convenient metrics. Drug development is financially riskier even as science improves, because advances can obsolete programs, funding is fragmented, and investors rarely stay through the full phase 1-to-NDA pathway. Combination therapies and platform technologies both raise and lower risk: they can improve efficacy and create reusable know-how, but they can also spread systemic risks across many programs. Therapeutic areas differ sharply in economics: oncology has seen improved recent success rates, Alzheimer’s remains a major failure zone, and vaccines suffer from weak incentives despite relatively high technical success rates. Biotech and financial markets behave like evolving ecosystems, where uncertainty, adaptation, and overreaction produce boom-bust cycles rather than smooth rational pricing.

Data Points: Gene therapy clinical trials registered: over 300 - Lo notes the scale of the pipeline and the coming wave of approvals. Expected gene therapy approvals in next 3 years: 5 to 10 - Lo predicts multiple approvals soon, increasing payer pressure. Luxturna list price: $850,000 for both eyes / $425,000 per eye - Used as an example of current gene therapy pricing. Potential hemophilia A therapy price: well north of $1.5 million - Illustrates future one-time therapy prices. Hemophilia annual treatment cost: $300,000 to $500,000 per year - Used to argue a high one-time price may still be economically rational. Sickle cell patients in the U.S.: 100,000 - Example of a larger-prevalence indication that could stress insurers. AMD patients in the UK: 600,000 - Illustrates scale if a gene therapy succeeds. AMD patients in the U.S.: 2 million - Used to show potential system-wide budget impact. Alzheimer’s patients: 5 million - Example of a high-prevalence indication that would force national reckoning if treated by gene therapy. Biopharma U.S. annual R&D spend: $75 billion - Sets the scale of drug discovery and development investment. Time to approve a new drug: 10 to 15 years - Average timeline from discovery to FDA approval. Total cost to bring one drug to market: about $2.5 billion - Historical estimate including failures along the way. Compound success rate: about 1 in 10,000 - Generalized survival statistic across the development gauntlet. Oncology success rate, 2000-2015: less than 5% - Historical path-by-path approval probability. Oncology success rate, last 5 years: about 10% to 15% - Recent improvement, especially for lead indications with biomarkers. Alzheimer’s success rate since 2003: 0% - No approved Alzheimer’s drugs since 2003. Vaccine success rate: north of 30% - Technical success is relatively high, but incentives remain weak. Approved drugs cited: 2,800 - Used in the discussion of whether existing drugs could be recombined to treat disease.

Pivotal Quotes: "we can also apply that to thinking about a mortgage for a drug. Drugs, drug mortgages." — Andrew Lo: Explains the core financing metaphor for one-time cures. "markets behave more like biological systems than physical devices." — Andrew Lo: Summarizes the evolutionary framing of market behavior. "As we learn more about delivering genes, we're going to actually reduce the risk across all gene therapies." — Andrew Lo: Describes how platform knowledge can lower risk while also creating shared systemic risk.

Implications: The healthcare system may need new payment infrastructure for curative therapies, including transferable multi-year financing. For biotech, investors and policymakers should expect both better science and more complex risk, especially as platform therapies and common delivery technologies scale.