The traditional approach for identifying earnings manipulation has been to construct a prediction model for earnings (or
its components) and then to treat deviations from these predictions as evidence of either deliberate misrepresentation or
low reporting quality. Two classic examples of this approach areJones (1991)andDechow and Dichev (2002). Jones (1991)
develops a measure of discretionary accruals based on the assumption that non-discretionary accruals are a deterministic,
linear function of the change in sales and the level of property, plant, and equipment, implying that anything unexplained
by the model represents discretionary accruals. In a similar fashion,Dechow and Dichev (2002)specify a deterministic
intertemporal decomposition of cash flows and then measure accruals quality as the estimation error in a regression of
changes in working capital on past, current, and future cash flows from operations. This approach is, however, limited in its
ability to separate earnings manipulation from operating volatility (for a discussion, seeDechow et al., 2010). Essentially, the
deterministic benchmarking approach classifies innovations to firm performance as misreporting and therefore potentially
leads to the excessive identification of earnings manipulation (for discussions, see Kasznik, 1999;McNichols, 2002).
To separate misreporting from operating volatility, we explore the time-series properties of reported earnings. We
demonstrate that several forms of misreporting produce serial correlation patterns in earnings that are difficult to
rationalize in the absence of misreporting. Specifically, we show that the residuals from a regression of reported earnings on
lagged reported earnings will have a negative second-order autocorrelation in the presence of misreporting. Empirically, we
find that the distribution of the second-order autocorrelation measure is asymmetric around zero with 74% of the
observations being negative and 27% being significantly negative. Assuming that misreporting is linear in the performance
shock, we find that firms in our sample subject to SEC AAERs have significantly higher estimates of manipulation intensity
and that our estimates of unmanipulated earnings are more highly correlated with contemporaneous returns and have
higher volatility than reported earnings.
There are, however, several important caveats to our methodology. First, we focus on the manipulation of earnings and
are indifferent about whether manipulation occurs through accruals or real activities that manifest in cash flows. Second,
we use an AR(1) specification to model persistence in unmanipulated earnings. It is important to point out that our AR(1)
specification for unmanipulated earnings could be misspecified. Third, our time-series specification imposes several
restrictions on the nature of the misreporting that can be identified. Specifically, it has to be a sustainable manipulation
strategy (i.e., it does not result in exponentially growing account balances). Fourth, our time-series strategy requires a
relatively long and stable series of firm-level observations to estimate our manipulation measure. Fifth, our methodology is
unlikely to capture manipulation strategies driven by motives other than masking performance shocks.
Overall, we do not claim that we identify the predominant form of earnings manipulation. However, it is important to
point out that misreporting arising due to motives unrelated to performance shocks is statistically similar to measurement error. Because measurement error produces serial correlations of the opposite sign than our prediction, such misreporting
would make it more difficult to identify manipulations of the form that we consider