Corporate decision making in pharmaceutical licensing has become increasingly complex. As the number of pharmaceutical companies constricts and the competition for effective products grows, the pressure on marketing and licensing managers to make successful decisions increases. At stake are the efficiency of each of our companies and the hundreds of millions of dollars used to ensure the competitive edge of the research, development, and marketing engines.

While good decision making often translates into tremendous profits and market opportunities, inappropriate decisions can result in a considerable loss of resources. Managers are increasingly required to quantify the risks inherent in their negotiations, and this requires better, though not necessarily more complex, analysis tools.

In this paper, we will merge the tools of these two meetings and demonstrate how simulation software can provide a solid foundation for quantifying and understanding risk. Desktop simulation software such as Decisioneering's Crystal Ball® can dramatically improve the development of spreadsheet models used for negotiation and forecasting. Crystal Ball is an Excel-based forecasting tool that performs Monte Carlo analysis (a technique for simulating real-world situations involving elements of uncertainty) on your existing spreadsheet models.

The Benefits of Monte Carlo Simulation

Monte Carlo simulation is a technique that uses random number generation to simulate reality. The benefits of a simulation modeling approach are: (1) an understanding of the probability of specific outcomes (e.g. will your profit exceed 20%?); (2) the ability to pinpoint and test the driving variables within a model (e.g. what factors most affect the NPV?); (3) a far more flexible model; and (4) clear summary charts and reports.

One of the problems associated with traditional spreadsheet models is that for variables that are uncertain, you are forced to supply a single, best-guess value. For example, you enter 10% for the 2001 sales growth even though you know it may vary between 0 and 20%. If you want to examine the effect of a 20% sales growth, you can change the variable and review the results. However this manual "what-if" analysis becomes tedious with multiple uncertain variables. Instead, most managers opt for best case, worst case, and most likely case evaluations, which still lack any sense of probability of occurrence.

With Crystal Ball, you have the ability to replace each uncertain variable with a probability distribution, a function that represents a range of values and the likelihood of occurrence over that the range. In the example above, you could use a normal distribution (the classic bell curve) for sales growth with a range of 0% to 20% and a mean value of 10%. In effect, you are describing the sales growth as mostly likely to be at or near 10%, with a decreasing but allowable possibility of nearing 0% or 20%.

Monte Carlo simulation uses these distributions, referred to as 'assumptions', to automate the complex "what-if" process and generate realistic random values. Instead of modeling a single uncertain outcome, you can quickly generate thousands of possible scenarios, view the result statistics, and evaluate your risks (e.g. in 95% of 1000 scenarios, your net profit only exceeded 10%). Because Crystal Ball imparts the ability to quantify your risks, it can be a crucial tool for a successful negotiation.

Adapting the Negotiation Model for Simulation

For this paper, we have used the negotiation model from the 1997 Canadian Health Care Licensing Association (CHLA) workshop on negotiation. The complete workbook used at the workshop is available at from the CHLA at CHLA@pathcom.com. With this simple model, we will show how the accuracy and variability of spreadsheet model assumptions can markedly influence the outcome of a negotiation. An excerpt of the spreadsheet is presented below (Exhibit 1).

Exhibit 1.

This spreadsheet models negotiations between BigPharm Inc., a buyer, and BioTech Wonder, a seller of a pharmaceutical product. To come to an agreement on the value of the product to either party, both groups use historical information (sales, promotional samples, etc.) and forecasts/estimates of future (e.g. earnings, tax rates, internal cost allocations, discount rates, royalty rates, and up-front licensing fees) (Exhibit 1).

Clearly, both the buyer and the seller enter the negotiations with different expectations. For example, the buyer may seek to optimize sales revenue or minimize tax, while the seller may seek to maximize the up-front fee. In a normal valuation or negotiation exercise, the parties would not share all of their information and expectations. Each firm would estimate the uncertain variables and follow up with a scenario analysis using a high, medium, and low number for each of the variables. If three variables were recognized as uncertain, then each firm would produce nine versions of the model (three possibilities for each of the three variables), and negotiations would be based on some combination of these scenarios.

For this spreadsheet, we will treat three of the variables as uncertain: 1997 Units Sold, Royalty Rate, and Up Front Licensing Fee for Technology. For the Up Front Licensing Fee with single-value estimation, we might select $18 million. With Crystal Ball, this variable becomes a triangular distribution with a range between $10-20 million and the most likely value at $18 million, as shown in Exhibit 1.

This distribution now represents a range of possible outcomes rather than a single, questionable outcome. The other two assumptions are similarly defined. The 1997 Units Sold variable ranges between 30 and 35 million with a most likely value of 30 million, and the Royalty Rate ranges between 1 and 5% with a mean value of 3%. By defining these variables as assumptions, the likelihood of making a realistic and successful negotiation increases.

Simulating the Negotiation Model and Interpreting the Results

Running the model without these assumptions might lead the buyer to conclude that the product could be licensed for any combination of royalty rates and up-front fees that give an IRR of 60% (e.g. 14% and $20M, or 1% and 35M). However, neither of these solutions might be acceptable to the product seller.

Instead, when we simulate the spreadsheet with Crystal Ball, we can examine the probability of results that occur in 5000 trials, or simulated outcomes. After running 5000 trials, we can view these results in a histogram called a frequency chart (Exhibit 2). The NPV forecast values range from a $15 to $50 million, with a mean value (50% certainty) of $38 million.

Exhibit 2.

We can also view the IRR forecast, to see the certainty of producing an IRR over a certain level, say 60%. When we enter 60% in the lower range field, we can see that in almost 80% of the trails - nearly 4000 outcomes - the IRR was greater than the 60% level (Exhibit 3).

Exhibit 3

With Crystal Ball, we can also examine which of the assumptions that drive this uncertainty. As Crystal Ball runs its simulation, the sensitivity analysis function collects and stores the assumption and forecast values, eventually compiling coefficients that describe the relationships, both direct and inverse, of the forecasts and assumptions. In the Sensitivity Chart below (Exhibit 4), we can see that the Royalty Rate is responsible for roughly 92% of the variance in the NPV forecast. 1997 Units sold has no real impact.

Exhibit 4.

We can see from the Sensitivity Chart that the main influences of NPV are the Royalty Rate (92%) and the Up Front Fee (8%). This would imply that the Buyer must closely manage even small changes in Royalty Rate to Biotech Wonder. For example, even a modest increase in the mean Up Front Fee of $500,000 would decrease the mean NPV by $1 million; however, a decrease in the mean Royalty Rate of 1% would improve the mean NPV by potentially $3 million with a decrease in the standard deviation of the results (a tighter range of the values around the mean). The implications for negotiations (and optimizing NPV) might include using a graded royalty rate based on sales and increasing the Up Front Fee by means of either cash or purchase of stock in Biotech Wonder.

Conclusion

The obvious benefits of Monte Carlo simulation are savings in time and resources. Crystal Ball eliminates the need to run, test, and present multiple spreadsheets, and the software lets negotiators estimate and quantify potential outcomes. Less obvious, but far more important, is the improvement in negotiating position and more convincing arguments for a product or technology valuation, a licensing negotiation, or the discussion of situation outcomes in the face of market competition or economic uncertainty.

Certainly, this is a very simplified example adapted for the purposes of demonstration; however, the tool can be adapted to any number of variables and assumptions associated with finance, marketing or clinical/research risk. Depending on the amount of information available on each of the potential variables in the deal under analysis, the approach can be adapted to improve the certainty of the forecast and the associated model.

Additional information on Crystal Ball is available at http://www.decisioneering.com or via phone at 800-289-2550 (U.S. and Canada) or 303-534-1515 (international). Decisioneering publishes a free, biweekly online newsletter that features risk analysis articles, application stories, and software tips, and interested parties can subscribe at the Decisioneering web site. Discussion groups on this tool meet at http://clubs.yahoo.com/clubs/cbug. Software training sessions are offered throughout the U.S. and Canada, and more information is available at http://www.decisioneering.com/training.html.

The Authors

Marcus Brady, Director, Corporate Development at Neurochem Inc.

Mr. Brady oversees Neurochem's corporate alliance initiatives with synergistic, fully integrated drug firms. Drawing upon an extensive background in technology licensing, venture capital, contract negotiation, intellectual property protection, research and development, and corporation alliance formation, he manages and expedites the commercialization of Neurochem technologies.

Prior to joining Neurochem in August 1994, he led the pharmaceutical and medical business development and marketing activities of the National Research Council, Canada's largest research and development organization. In that role, he established corporate relationships, collaborative research ventures and licensing agreements with large and small biotechnology and pharmaceutical companies

Mr. Brady holds an M.Sc. in Medical Physiology from the Health Sciences Centre of the University of Ottawa, Ontario, Canada, and a B.Sc. in Biochemistry from McGill University, Montreal, Quebec, Canada. He is an executive member of the Canadian Health Care Licensing Association.

Lawrence Goldman, Web Marketing Manager, Decisioneering, Inc.

Mr. Goldman currently directs Decisioneering's Web site development and Internet marketing strategies. During his tenure at the company, he has held positions as program manager, software trainer, and advanced technical support technician. Prior to 1997, he served as a technician in the mining software industry and as an exploration mining geologist. Mr. Goldman holds a M.S. in Geology from the University of Cincinnati, Cincinnati, Ohio, and B.A from Cornell University, Ithaca, New York.

To make any comments on this article, or to ask a question of the author, please contact the publisher. If you would like to submit an article please subscribe to our PL Intelligence service.

The opinions expressed in the articles published in this section do not necessarily reflect those of Pharmalicensing or UTEK Corporation. No actions including proposals to or agreements with other companies should be taken by any reader without obtaining specific business or legal advice. Neither the publisher nor the authors accept any liability for any actions or activities undertaken by any reader or other third party as a consequence of these articles or for any errors or omissions therein.