One possible explanation for the weaker EBIT results on the PC hypothesis is that
using EBIT as a definition of operating profits might overstate free cash flow and, hence,
understate excess CAPX, which consequently leads to misclassifying some observations as
PC motive firms when they should be classified as AC motive firms. To shed light on this
issue, we perform a logit regression analysis similar to those reported in Table 6, but using
the PTI AC/PC motive partition in conjunction with using I ROS EBIT as the abnormal
profit measure. Results (not tabulated) are significantly stronger for the PC motive hypothesis—
the coefficients on I ROS EBIT * PC are positive and significant at p 0.05 in all
three deviation samples (the coefficient on I ROS EBIT * AC remains consistently negative
and significant at p 0.05 in all three deviation samples). These results are consistent with
potential misclassification of AC motive firms in the PC motive sample driving the weak
results on the PC hypothesis. Nevertheless, the PC results are not robust to some of the
sensitivity tests discussed in the next subsection. We therefore interpret our results as being
mixed with regard to the proprietary cost hypothesis.