median leverage ratio is 4.21% higher for adopting firms than for their industry
counterparts; it is 14.91% higher than industry competitors during the postadop-
tion period. During both periods, shark repellent-adopting firms display payout
ratios that are significantly higher than those for industry counterparts.
In sum, based upon nonparametric results shown in Table 9.3, it is fair to
conclude that shark repellent-adopting firms tend to display higher than typical
levels of investment in plant and equipment and R&D. Based upon results shown
in Table 9.4, it also is fair to deduce that shark repellent-adopting firms display
relatively high dividend payout and leverage (debt-to-equity) ratios. None of
these findings is consistent with the concept of management entrenchment and
the use of shark repellents by managers of firms that underinvest in risky or long-
term projects. The fact that shark repellent-adopting firms tend to invest heavily
in long-term or risky projects, pay out relatively high amounts of free cash flow,
and employ significant leverage is consistent with the shareholder-interest
hypothesis.
Industry-adjusted medians reported in Tables 9.1–9.4 are interesting because
they offer useful insight concerning adopting firm performance and financial
policy in the pre- and postadoption periods. For a direct test of changes in firm
performance and financial policy resulting from shark repellent adoptions, it is
necessary to analyze changes in firm performance and financial policy decisions
in the period surrounding shark repellent adoptions.
IX. REGRESSION MODEL SPECIFICATION
Following Healy, Palepu, and Ruback (1992), simple cross-sectional regressions
are estimated to capture statistical differences between postadoption industry-
adjusted ratios and corresponding pre-adoption ratios. These simple regression
equations take the form
(9.17)
where IAR
i,post
is the median industry-adjusted firm performance or financial
policy ratio for the 5-year postadoption period, and IAR
i,pre
is similar median
industry-adjusted ratios for the 5-year pre-adoption period as defined in equation.
In this simple regression, the estimated slope coefficient, β, captures the
correlation between pre- and postadoption firm performance and financial policy
decisions. More important is the size and statistical significance of α, the inter-
cept term. The estimated intercept coefficient α is independent of pre-adoption
performance and reflects significant changes in the industry-adjusted relative
position of adopting versus nonadopting firms in the postadoption period. Using
this approach, significant changes in relative firm performance or financial policy
decisions in the postadoption period are reflected by statistically significant
intercept coefficient estimates.
IAR IAR u
i post i pre,,
,=+ +
αβ
REGRESSION MODEL SPECIFICATION 223

Get Tech Stock Valuation now with O’Reilly online learning.

O’Reilly members experience live online training, plus books, videos, and digital content from 200+ publishers.