i've recently done a work on a RO valuation of a mobile base station launch. i have one (:-)) word: it's VEERY untrivial. to my mind, Monte Carlo is the only realistic technique if you don't want to get to the level of "assume in period 1 you produce 2 widgets etc." for instance, MC allows you to parametrize the underlying stochastic process, which BS doesn't (and in the real world our sales do grow faster than risk-free interest rate). in fact, BS is theoretically unacceptable in most of the cases (unless there are spanning assets or effective securitization is possible). i advise combining the MC with a custom-structured backward induction process. Also: in practice, if you want to value something like a "real option to introduce new product", in most of the cases you won't be effectively able to say how many of such products the company can introduce, let alone their projected cash flows. The latter might be very specific, but it's actually extrapolatable to the whole classes of ROs.in my case, I was eventually able to approach the real option to delay, but I used a sort of "sensitivity analysis" on the side of the parameters for volatility and growth trends.Final advise: check out the following site and specifically for Dias, Rocha (2001, i guess) for REALISTIC RO valuationshttp://
www.puc-rio.br/marco.ind/