A Neural Network Approach to Value R&D Compound American Exchange Option

Abstract In this paper we show as the neural network methodology, coupled with Least Squares Monte Carlo approach, can be very helpful in valuing R&D investment opportunities. As it is well known, projects are made a phased manner, commencement of subsequent phase being dependent on successful completion preceding phase. This known sequential and therefore considered compound options. addition, investments often involve considerable cost uncertainty so that they viewed an exchange option, i.e. swap uncertain for gross project value. Finally, production realized at any time before maturity date, after effects disappear. Consequently, American option. context, method powerful flexible tool capital budgeting decisions American-type But, using simulated values “targets”, implementation allows to extend results valuation abate waiting simulation.