Variable Description, Sample, and Descriptive Statistics
This section builds on discussion in the prior section of the analytical
properties of the persistence of accruals and cash flows by examining the
15 Resultsempirical properties of the accounting-based and time-specific measures of
these constructs. Our focus is on whether we draw different inferences about
the relative persistence of accruals and cash flows depending on whether
accounting-based or time-specific measures of these constructs are used. Because
prior studies focus on the aggregation of accrual and deferral transactions
in evaluating the persistence of the noncash component of income
(accruals), we follow the same framework for our evaluations. Specifically,
our tests focus on the relative persistence of the aggregate accruals measures
(AccrualsAcctg
t and AccrualsTime
t ) and the aggregate cash measures (CashAcctg
t
and CashTime
t ). Our empirical analyses require information on income and its
accrual and cash components; we follow Sloan’s [1996] definitions of these
variables. Specifically, income in year t, Incomet, is defined as operating income
after depreciation and amortization expense (Compustat #178), and
accounting-based accruals equals the change in operating working capital
accruals adjusted for depreciation