A framework for the valuation of biotechnology companies

This thesis explores valuation of biotechnoogy companies, and attempts to establish a framework for valuation based on underlaying technologies. The particular problems associated with biotech valuation are: most of the company's assets are in the form of intellectual capital, sa long product development time of 7-10 years, a short product shelf life and a high product failure rate.
The analysis begins with a study of the effectiveness of 2 ratios: the price earnings ratio and revenue multiple, in selecting a portfolio of stocks which outperform an index. Portfolios constrructed on the basis of most and least attractive price earnings ratios/ revenue multiples did not significantly outperform the benchmark Amex Bioteh Index over 1,2 or 3 year time frame even after adjustment for risk and volatility. One explanation for this could be that the range of biotechnology companies selected purely on the basis of ratios were too wide in terms of technology platforms and business models to be meaningful. Successful selection of outperforming biotech companies may possibly be improved by classification of the companies.
In the remainder of the study, the biotech companies are classified in accordance to underlaying technologies in a framework similar to that used for semiconductor comapanies and analyses performed to explore the characteristics of each category. The main types of technology used for classiffication (Traore and Rose (2003)) are: DNA based technologies, Biochemistry/Immunochemistry based technologies and Bioprocessing based technologies. Other categories included are Conglomerates, Computer Software, Contract Research Organisations and Medical Instruments.
Average revenue was lowest for DNA types companies, in comparison to the other types. As can be expected, revenue for conglomerates was substantially higher than the rest, with statistically significant differences. Average P & L for all of the 3 main category companies were negative. Average betas observed for DNA companies wer highest, and conglomerates lowest, and these differences were statistically significant. These observations are consisent with or upstream/downstream model in that products of DNA companies take longer to come to the market, the risk of these products being profitable ar higher and therefore DNA companies can be expected to have higherbetas than conglomerates which offer a diversified range of products. Average number of employees ranged from 220 for DNA companies to 6230 for conglomerates, and differences were statistically significant.
We then selected stocks from the 3 main technology classifications using as criteria, the most and least attractive price earnings ratios and price-revenue ratios. Overall results were mixed, but there was evidence that selection of biotechnology stocks following classification had more success in building outperforming portfolios than when price earnings ratio alone is used as selection criteria. One portfolio formed from each of the 3 main technology classifications showed statistically significant superior returns to the benchmark. Price/revenue used as selection criteria showed less clear trends. Further investigation showed that the CAPM is not an effective model in explaining cross sectional stock returns. There was low correlation between stock returns and the variables-revenue, R & D/ Revenue, P & L, beta, number of employees. This observation held even after analysis of the companies in their technology classifications. Further investigations into defining the risk/reward profile of each technology classification will allow portfolio managers to be more specific about their preferred risk exposure.