Optimal ordering decision for assemblers facing multiple supply uncertainties

We study the optimal ordering decision for an assembler who assemble two components into a final product. Each component faces one type of supply uncertainty---the standard component has uncertain production capacity, the non-standard component has random production yield. The profit-maximizing firm orders components before supply uncertainties are realized to meet a deterministic demand for the final product. We characterize the optimal ordering decision and find that: (a) component l's order amount should be exactly a fixed fraction of component 2's order amount, the fraction depends on component l's ordering cost and component 2's random yield, but not on component l's uncertain capacity, (b) an increase of the standard component's uncertain capacity leads to an increase in both components' optimal ordering amounts, but surprisingly, an increase of the non-standard component's random yield does not necessarily lead to a monotone change in both components' optimal ordering amounts.