additional sensitivity tests – 182 firms.

Ordinary least squares regressions of the log of IPO net proceeds on explanatory variables denoting proxies for various investment strategies, a set of control variables and a set of sensitivity variables. The dependent variable in Model 5 and 6 is the log of IPO net proceeds whereas the dependent variable in Model 7 and 8 is the log of IPO net proceeds deflated to 1967 dollars. The control variables are gross spread, log of the number of employees, log of underwriter reputation, log of the number of shares locked up, hot issue and sector sales growth. Gross spread is measured as the difference betIen IPO gross proceeds and IPO net proceeds over the IPO gross proceeds. The number of employees is downloaded from SDC. The measure of underwriter reputation is based on a ranking table provided by Loughran and Ritter [5]. The “number of shares locked up” is downloaded from SDC. Hot issue is coded as 1 if the IPO occurs in 1999 or 2000 and 0 otherwise. Sector sales growth is the change in average annual sales for the genomic, biotechnology, medical devices and healthcare sector. EBITA is the earnings before interest, tax and amortization expenses in the year that the issuer go public. R&D expenditure is research and development expenditure and is downloaded from Compustat. EBITA-deflated refers to EBITA deflated to 1967 dollars. R&D expenditure-deflated refers to R&D expenditure deflated to 1967 dollars. The explanatory variables in model 5 and 6 are number of corporate elites with healthcare industry background while the explanatory variables in model 7 and 8 are percentage of corporate elites with healthcare industry background. Healthcare industry background refers to experience in selected areas such as home health care services, hospitals, outpatient care centers, nursing and residential care facilities and, HMOs and health/medical insurers. We report standardized beta coefficients for all variables except for constant. I report beta coefficient for the constant. T-statistics are in parentheses. Statistical significance at 10%, 5% and 1% are denoted respectively as *, ** and ***.