| 1. Exploring blockbuster economics. Pharmaceutical firms have relied on "blockbuster" drugs (annual sales over {dollar}1 B) to meet earnings targets. However, doing so means numerous medical needs in more modest markets remain unaddressed. To explore development incentives, we built a probabilistic model using published data. We found that a threshold of {dollar}444M, not {dollar}1 B, justifies investment in clinical trials. Sensitivity analysis revealed that sub-blockbuster drugs should still be profitable even with three times the risk as historical averages. We offer explanations for the gap between observed behavior and the model's predictions, and suggest a logic for modifying the regulatory environment to direct incentives towards apparently profitable but underserved markets.; 2. Pharmaceutical pricing and the innovation incentive. Growth in pharmaceutical expenditures has refueled interest in regulating pharmaceutical pricing to contain healthcare costs. We modeled the effects of price containment using a discrete-event simulation. Results showed that reduced pricing led to superlinear decreases in returns and portfolio size while increasing risk. Patent extension and improving success rates restored value and reduced risk; no degree of cost containment could produce these effects. Further, some scenarios with the base case valuation but with reduced pricing and patent extension had lower revenue, implying lower expenditures. We conclude that modifying the patent environment offers a lever in value restoration and risk reduction in the face of price controls.; 3. A dynamic programming model of pharmaceutical development decisions. This paper presents a dynamic programming model of pharmaceutical development decisions, in which the estimate of ultimate market value changes with the receipt of new clinical trial information. For intermediate market estimates, this model gives a higher project values than under constant conditions, thus shifting down the minimum value required for funding. The extra value derives from the option to abandon, and from the probability that future updates may increase the project value above the current cost-to-go. When the updated estimate is uniformly distributed, projects of intermediate value are shown to be polynomial of degree n+l, where n is the number of remaining updates. Numerical examples illustrate how taking optionality into explicit account can materially affect decision making. |
