3.2.2 Industry Adjusted Returns
It is possible that the abnormal returns we document are not caused by man-
agerial ownership but rather by an unequal industry distribution of firms
with high managerial ownership and firms with low managerial ownership.
As managerial ownership indeed is not equally distributed across all indus-
tries, adjusting firm returns by industry returns will capture some of the
positive effect of managerial ownership on stock returns. However, we still
expect positive albeit smaller abnormal returns for our managerial owner-
ship portfolios even after industry adjustment, because not all firms in the
respective industries will be characterized by high CEO ownership. Results
for the industry-adjusted portfolios based on Fama-French industry classi-
fications are presented in Panel B of Table 4.9 For 5% and 10% managerial
ownership portfolios we still find statistically significant abnormal returns
of 7.9% p.a. and 11.5% p.a., respectively. This supports the idea that indus-
tries capture some of the managerial ownership effect on returns. However,
even after taking this effect into account, we can still document significant
abnormal returns.
9We carry out the industry adjustment by subtracting the industry return from each
individual firm return before constructing our portfolios. Technically, the managerial own-
ership portfolios then consist of the same stocks as before. Additionally, for each firm there
is an industry hedge term. It essentially is a short position in a portfolio consisting of all
stocks in the same industry as the firm and which is equal in size to the stock’s weight in
the portfolio. It is sometimes argued that adjusting returns based on the Fama and French
(1993) industry classification of firms might lead to misleading results, especially during
the U.S. tech bubble. Thus, we use an alternative industry adjustment based on two-digit
SIC-codes. Results (not presented) are very similar to those obtained using the Fama and
French (1993) industry classification.
12