5. Conclusion
CFOs are in charge of corporate financial reporting and play an important role in the financial reporting system. This
paper investigates why CFOs become involved in material accounting manipulations. To address this research question, we
examine two possible explanations. CFOs might instigate accounting manipulations for immediate personal financial gain,
as reflected in their equity compensation. Alternatively, CFOs could manipulate the financial reports under pressure
from CEOs.
Using a comprehensive sample of material accounting manipulations disclosed between 1982 and 2005, we investigate
the costs and benefits associated with intentional financial misreporting for CFOs. We find that CFOs bear substantial legal
costs when involved in accounting manipulations. We also document that these CFO equity incentives (measured by payfor-performance
sensitivity) are not significantly different from those of CFOs of control firms. However, CEOs of the
manipulation firms have significantly higher equity incentives and power than CEOs of the control firms. Moreover, CFO
turnover is significantly higher within three years prior to the occurrences of material accounting manipulations for
manipulation firms than control firms, consistent with CFOs facing significant costs (loss of job) for saying no to CEOs.
Finally, our AAER content analyses suggest that CEOs of manipulation firms are more likely than CFOs to be described as
having orchestrated the manipulation and to be requested to disgorge financial gains from the manipulation. Taken
together, our findings suggest that CFOs are likely to become involved in material accounting manipulations because they
succumb to CEO pressure, rather than because they seek immediate financial benefit.
Some caveats are in order. First, we assume that CFOs of accounting manipulation firms are aware of or are involved in
misreporting. We believe this assumption is reasonable given that one of the main job responsibilities of CFOs is to watch
over the financial reporting process and make related decisions. However, in some unusual cases accounting
manipulations could occur without the knowledge of CFOs (e.g., CEOs collude with divisional managers to create fictitious
sales and hide the manipulation from CFOs). These cases are likely to add noise instead of introducing a systematic bias to
our empirical results. Second, we assume that the companies identified by the SEC have indeed manipulated financial
statements. This assumption seems reasonable given that the SEC spends effort and resources to establish evidence for the
alleged manipulations. However, the SEC likely does not identify all the companies with accounting manipulations; as a
result, some of our control firms might have ‘‘undetected’’ manipulations. This issue would be a concern if the SEC
systematically pursues companies with characteristics examined and found significant in our empirical tests, but we are
not aware of any evidence supporting this possibility.
While subject to these caveats, our paper contributes to the understanding of CFOs’ incentives when they face
accounting manipulation decisions. Our findings suggest that CFOs are typically not the instigator of accounting
manipulations. Instead, it appears that CEOs, especially powerful CEOs with high equity incentives, exert significant
influence over CFOs’ financial reporting decisions. In other words, the CFO’s role as watchdog over financial reports is
compromised by the pressure from CEOs. Overall, the findings of this study suggest a corporate governance failure for the
accounting manipulation firms, and have important implications for current corporate governance reform. While
researchers, practitioners, and regulators have generally concluded that stock-based compensation has provided managers
with incentives to misstate accounting numbers, our results indicate that re-designing compensation packages for CFOs is
not necessarily the only remedy. Improving CFO independence by alleviating the pressure of CEOs on CFOs could be critical
to improving financial reporting quality. One possible way to achieve this would be to have boards or