For a very general family of continuous time-dependent cost-spending and reward functions we propose modeling the cost-spending and Net Present Value as stochastic processes by applying time-to-event analysis methods. We show how this can be achieved by requesting no more information than necessary to calculate the Pearson Index. We analyze and discuss the impact of the success and survival probabilities, the cost-spending and reward patterns on the expected cost, and the budgeting and the profitability of individual projects as well as the portfolio as a whole. We also provide some advice on how the cost and profitability performances can be improved.
