Optimal management of an R&D budget

This paper characterizes the optimal allocation of a budget between several stages of an R&D project. When intermediate stages are profitable, the optimal allocation involves greater expenditures in early stages. Increasing the budget does not necessarily increase expenditures in every stage but does increase the budget that remains unspent at each stage. A simple necessary and sufficient condition is given for the optimality of 'bold play' - the spending of all remaining budget on the current stage. If profits are realized only after all stages are successfully completed, decreasing returns in the innovation production function implies that the optimal strategy is to spread the budget evenly between the stages of the R&D project.