R&D Asset Launches
Forecasting the probabilities of success of a portfolio of R&D projects
Research and development refers to innovative activities undertaken by institutions in developing new services or products, or improving existing services or products.
In the United States, a typical ratio of research and development for an industrial company is about 3.5% of revenues; this measure is called “R&D intensity”. A high technology company, such as a computer manufacturer, might spend 7% or pharmaceutical companies such as Merck & Co. 14.1% or Novartis 15.1%. Anything over 15% is remarkable, and usually gains a reputation for being a high technology company such as engineering company Ericsson 24.9%, or Allergan a biotech company, tops the spending table with 43.4% investment.
Typical R&D intensive industries are prescription drugs or special chemicals, scientific instruments, and safety-critical systems in medicine, aeronautics or military weapons. Usually, these firms prosper in markets whose customers have extreme high technology needs. The extreme needs justify the high risk of failure. It is not difficult to find companies with high gross margins from 60% to 90% of revenues.
Take for example the cases of large pharmaceutical corporations. These companies heavily invest on dozens or hundreds of simultaneous research endeavors. Many of them will eventually fail since the inherent risk of developing complex molecules is no easy task. However, eventually, some of them will become successful drugs that will fill in the drugstores and hospital shelves, providing effective and carefully approved medications. These processes usually take many years.
Drug R&D (across all therapeutic areas) takes on average 14 years. Discovery research lasts for 4.5 years, preclinical testing continues for 1 year, the three clinical development phases take 1.5, 2.5 and 2.5 years, respectively, and the phase from submission to launch requires another 18 months.
We will build here an example model that calculates probabilities of such a R&D portfolio of projects. The model will be capable of answering questions about the likelihood of success of the portfolio as a whole. For example, what is the probability of having 3 successful product launches on any given year or range of years. What is the likelihood of not having a single successful launch before a certain date.