Back to Index96
Order hard copy editions of Frontiers of Entrepreneurship Research by mail
ESTIMATION OF RISK IN YOUNG VENTURES
Sanjeev S. Punjabi
Robert H. Keeley
College of Business
University of Colorado at Colorado Springs
P.O. Box 7150
Colorado Springs, Colorado 80908-7150
This study examines the problem of estimating the risk in young companies. Such estimates are important for setting values on companies, and are crucial for valuing their stock options--a matter being hotly debated today. Regardless of that debate's outcome, managers, employees and investors are increasingly interested in valuing employee stock options (even before a company is public) This study compiles a set of risk measures for technology-based companies that can assist in setting such values.
Two characteristic measures of risk are well established: "beta"--the risk that even a diversified investor can not avoid, and "volatility"--the standard deviation of an individual company's return in the stock market. Several questions arise in attempting to use any of these to describe the risk of a young company:
Is the risk measure reasonably constant over time?
Are risk measures the same across industries (for companies of similar sizes and ages)?
Are risk measures constant for companies of different ages (holding size constant)?
Can a risk measure be derived from some fundamental properties of a company (e.g. size, age, industry, cash flow, leverage)?
To the extent that the answers to the above questions are "Yes" the task of estimating risks, even for new, private companies may be fairly simple. But these questions have only been studied for large, mature companies. This study develops answers for young companies.
The study has two sub-samples: 190 young companies with IPO's between 1979 and 1986;
and 150 mature companies (selected to match the industry makeup of the young companies). One-half of the young companies were backed by venture capitalists and approximately 75 percent are from technology-based industries.
The first stage of the research estimated the risk of each company for as many years as possible during the period 1980-1989 using a regression method developed by Scholes & Williams. The resulting risk estimates were then regressed against potential explanatory variables (e.g. calendar year, age of company, young vs. old, industry, year of IPO) using a two-stage, error components model (developed by Nerlove). The error components model is well-suited for panel data because it allows the error term to incorporate a firm-specific component as well as a time component.
Although the simplest outcome (i.e. a "Yes" answer to the four questions
above) does not hold, risk measures are reasonably predictable based on a company's fundamental properties (e.g., industry, size, leverage). Young companies do not have substantially higher systematic risk (measured by "beta") than established companies, but their total risk is higher (the standard deviation is 57%/yr. for young companies and 41%/yr. for established firms). High tech companies have higher systematic risk but lower total risk than non-high tech companies. Both risk measures change from year to year for young companies. The change in total risk is small (though statistically significant) and can probably be ignored. The change in systematic risk is often as much as 0.2 against an average value of about 1.0.
This study sets values for the risk parameters of young companies that are central to valuing a company's equity and its stock options. Some of the findings, such as the fact the high tech companies have lower total risk than others, are counter-intuitive and raise questions about how risks as assessed in product markets translate into financial risks.