Coping with uncertainties: Monte Carlo simulation

SKIM has been using Monte Carlo simulation to help clients with forecasting the uptake of branded generics, biosimilars and targeted medicine – areas of increasing complexity and uncertainty. On behalf of SKIM, our experts Dirk Huisman and John Ashraf presented this approach at EphMRA's 2012 Conference in Paris. They talked about the principles of Monte Carlo Simulation and how it can be applied to forecast the uptake of your product.

Did you miss our presentation or would you like to review the slides? Browse through the slides in SlideShare below or contact us for more!

About our presentation: Managing uncertainty in the forecast of the uptake of branded generics, biosimilars and targeted medicine

Ever-increasing complexity and uncertainty are a reality for the healthcare industry's commercial and strategy teams. In spite of these elements, however, decision makers must develop initiatives that help build a successful business. These strategic activities, which include commercialization plans and product launches, rely heavily on a plethora of data and business intelligence including, among many others, forecasts.

In the context of forecasting, 'increased complexity and more uncertainty' means that parameters used to build the forecast interact more; their accuracy is lower and not all conditions that may influence the forecast are known. Combined, these features result in an outcome that is impacted by an element of unpredictability. In market research we are used to dealing with statistical uncertainty, but coping with data uncertainty (i.e., uncertain product profiles, uncertain access data, uncertain competitive actions, conflicting data sources, etc) calls for a different strategy. One approach is to use Monte Carlo simulation.

The principles of Monte Carlo simulations

The basic principle of Monte Carlo simulation is to assign a random value, limited to an appropriate range, to each parameter. The full forecast is then generated using this random value and the process repeated thousands of times using different random numbers. It is crucial, therefore, to limit each parameter to the appropriate range and take into account the most likely value (i.e., the mean) and the probability of the other values within the range. This 'qualification' depends on other data sources (e.g., results of conjoint, insights from KOLs, desktop research, historical data, etc.) and is subjective.

In addition to the process of 'qualifying' the random values, it is essential to define action standards, essentially thresholds for taking action, at the outset. As these forecasts involve uncertainties, probabilities are attached to outcomes: for example, there is a 70% chance of reaching revenue of $500 million and a 90% chance of reaching $300 million. The associated action standard for this outcome is, for example, to in-license the product if the likelihood of revenue of $500 million is more than 80%. In this case, therefore, the decision would be a no-go. Action standards are based on the level of acceptable risk to make strategic choices such as a go/no go decision.

The challenges of Monte Carlo simulations are how to best and most accurately qualify parameters and decide on action standards at the outset of a research study. Both of these involve input from and a partnership between the market researchers and business intelligence managers.

Ultimately, Monte Carlo simulations are a way to obtain more accurate forecasts in areas of great complexity and uncertainty. As with a traditional forecast, however, the goal remains to deliver an estimate of market share and sales revenue.

If you are unable to view the slides on this page or you'd like to download the presentation file,
click here to view our slides on slideshare.

About European Pharmaceutical Market Research Association (EphMRA)

The EphMRA annual conference of 2012, 'The 360 Future', focused on looking at issues from all angles and possibilities, and examining the challenges we face in healthcare market research. The conference took place on June 19-21, 2012 in Paris, France.

Want to know more? Don't hesitate to give us a call or drop an email!