Harvard Governance Blog - Taxes and the Backdating of Stock Option Exercise Dates - 17 Dec 2008

(Editor’s note: This post comes from Shane Heitzman at the University of Rochester Simon Graduate School of Business, Dan Dhaliwal at the University of Arizona and Merle Erickson at the University of Chicago Graduate School of Business.)

http://blogs.law.harvard.edu/corpgov/2008/12/15/taxes-and-the-backdating-of-stock-option-exercise-dates/

In our paper “Taxes and the Backdating of Stock Option Exercise Dates”, which was recently accepted for publication at the Journal of Accounting and Economics,
we investigate the opportunistic timing of stock option exercises by
insiders. We focus on a group of exercises where there likely exists
both the incentive and the ability to backdate an option exercise:
exercises paid in cash where the insider holds the acquired shares.
Once the decision to exercise is made, insiders who plan to hold the
acquired shares have an unambiguous personal tax-based incentive to
exercise on the day with the lowest possible stock price. Unlike
exercises in which the acquired shares are sold immediately through a
broker, these exercise-and-hold transactions are often accomplished
in-house. Thus, we expect that opportunistic backdating, to the extent
it exists, is more likely to occur in exercise-and-hold transactions.
We find that exercise-and-hold transactions tend to occur at monthly
stock price lows. Before SOX, we find that 13.55% of exercise-and-hold
transactions by CEOs occurred on the day the stock was at its lowest
price during the month (i.e. suspect exercises). After SOX, only 7.20%
of CEO exercise-and-hold transactions occurred on that day.

Consistent with the prediction that backdating an exercise-and-hold
transaction is driven by personal tax considerations, we find that the
likelihood of a suspect exercise is increasing in the potential taxes
saved by the option holder from exercising on the day of the month with
the lowest closing price before SOX, but not after SOX. Finally,
suspect exercises are more likely in small firms. While this finding is
consistent with the conclusion that backdated exercises are more likely
when the firm has a relatively weaker internal control environment we
also find that the probability of a suspect exercise is not
consistently related to common proxies for corporate governance based
on the subsample of observations with available governance data.

We estimate that our sample of CEOs saved an average of $96 thousand
in taxes per exercise by exercising on the day of the month with the
lowest closing stock price. These tax savings make up only about 3.2%
of the total value of the options exercised, and are even smaller at
the median. Given that filing a false tax return can be a felony, can
result in individual level penalties in excess of $100 thousand (among
other costs), and can have significant adverse consequences for the
firm and shareholders, the tax savings realized by CEOs through suspect
option exercises seem remarkably modest. One likely explanation is that
insiders believed the probability of getting caught was low enough to
justify the risk. We also find that the firm’s foregone tax benefits
from suspect exercises are of similar magnitude to the taxes saved by
the CEO.

Finally, we find that suspect exercise-and-hold transactions are
more likely in firms with a higher likelihood of stock option grant
backdating. For the sample of exercise-and-hold transactions by
insiders of firms under scrutiny for option grant backdating as listed
in the Wall Street Journal’s “Options Scorecard”, we find
that 21.53% of exercises by CEOs and 18.41% of exercises by non-CEOs
occurred on the day the stock was at its lowest price during the month.
That is, insiders from firms with alleged grant backdating practices
were more likely to have a suspect exercise than other insiders. We
also find that the firm-specific odds of option grant backdating are
positively associated with the frequency of suspect exercises among
firms not mentioned in the Wall Street Journal’s list.
Together with the remainder of this study, our analysis provides
additional evidence on the magnitude and determinants of opportunistic
behavior associated with executive stock options.

The full paper is available for download here.