Over the last two decades empirical evidence shows a significant increase in R&D investments for developing new drugs. However, such an increase has not been matched by a proportional growth in the number of R&D outputs, especially in terms of investigational new drugs and new drug approvals. As a consequence, a debate has developed to understand the reasons for such fall in the industry productivity of new drugs. In this paper we aim at contributing to such debate by proposing a new method to evaluate the innovative performance of a biotechnology and pharmaceutical firm, hinging on the portfolio dynamics of a firm. In particular, based on data taken from the Biopharma Insight database, we first estimate success probabilities of transition, for compounds belonging to different therapeutic areas, among clinical phases and the average duration time of a phase. We then forecast a firm portfolio composition at 2009, by projecting through our model its portfolio at 2004. Finally, as an indicator of firm performance we compute the difference between the actual (observed) portfolio at 2009 and the projected one. Positive differences would indicate that a firm performed better than the average industry and vice versa. We conclude by suggesting how further research should proceed to improve the method's reliability as an instrument that could be used to for identify and evaluate the impact of those factors that could contribute to reverse the productivity crisis.