Review of Quantitative Finance and Accounting, 17: 187–211, 2001
2001 Kluwer Academic Publishers. Manufactured in The Netherlands.
The Adjustment Process of Accruals: Empirical
Evidence and Implication for Accrual Research
EMEKA T. NWAEZE
School of Business, Rutgers University, Camden
Abstract. This study analyzes the intertemporal behavior of accruals and uses the results to offer some perspec-
tives on models of accruals. Separate adjustment rates are estimated for total and managed accruals and further
compared to assess the inﬂuence of managers on accrual adjustment. Analysis is further performed to test whether
there is an industry-speciﬁc adjustment rate for either total or managed accruals. The results show that both the
total and managed accruals follow a partial adjustment process rather than a pure-adjustment or random-walk pro-
cess. Nonetheless, managed accruals exhibit longer adjustment length than total accruals in each industry sample.
In addition, total and managed accruals appear to follow certain adjustment patterns that are industry-speciﬁc,
consistent with the view that common industry practices and norms lead to benchmark adjustment rates to which
industry ﬁrms conform. Based on the results showing that accruals exhibit lagged and industry-speciﬁc adjust-
ments, an extension of the modiﬁed Jones model is proposed. The extended model incorporates lagged accruals
and an industry accrual index in an attempt to exploit the lagged-adjustment and industry effects and, in that
context, improve the temporal classiﬁcation of accrual innovations. Simple tests show that the extended model
achieves substantial improvements in speciﬁcation over the modiﬁed Jones model.
Key words: accrual adjustment process, managed accruals, model speciﬁcation
JEL Classiﬁcation: C22, M41
The role of accruals in ﬁnancial reporting has been the subject of much debate. At issue are
the properties of accrual that impact the information content of reported income. Several
prior studies link the information import of accrual components to the persistence of the
components (Sloan, 1996; Xie, 2000) and to their non-discretionary/discretionary quality
(Subramanyam, 1996). Despite numerous research in this regard, there is a view that empir-
ical analyses of the relevance of accrual components suffer from the limitations of existing
accrual models. One common argument is that the existing accrual models do not reliably
Address correspondence to: Emeka T. Nwaeze, School of Business, Rutgers University, Camden, New Jersey
08102, Fax: (609) 225-6231.