10-34
d. To test the relationship between the identified incentives and upward management of
CFO, the author develops a sample of companies with publicly available data from 1998-2008.
However, companies in regulated industries, and banks and financial institutions are not included
To understand if classification choices are employed as a manipulation method in response to
incentives, the author first obtains a sample consisting of 414 firm years of cash flow
restatements and a matched sample of firms without cash flow restatements. The author then
uses logistic regression to test the relationship between a restated cash flow statement (the
dependent variable) and the incentive factors (the independent variables)
The next classification test considers how firms classified a particular item, when given
discretion as to their choice. Prior to the mandatory expense of stock options, no guidance was
provided as to where the tax benefit from the exercise of employee stock options should be
classified. The author obtained a sample of firms from 1994 (when data is first available) to 2000
The timing test examines differences in the length of the last quarter’s cash conversion cycle
compared to the other quarters. If firms are delaying payments and accelerating collections
before the financial statement cutoff date (as a final effort to boost CFO), Q4’s cash conversion
e. This study has a number of limitations. First, the author uses an approach to determining
unexpected CFO from prior literature. This approach requires that firms have data available for a
ten year period, creating a natural bias towards more mature, stable firms. Further, it may be the
case that the approach for determining unexpected CFO may not have successfully captured all