Investigating the correlation between accruals and cash flows, the authors show that it
has diminished constantly and significantly since the 1960s. This decline is mainly
explained by increases in such non-timing-related accruals as one-time and nonoperating
items and frequency of losses.
The authors’ objective is to explore the temporal changes in the correlation between
the cash flow from operations and the accounting accruals. They find that non-timing-related
accruals offset the generally accepted negative relationship between cash flows and
accruals. The traditionally held view—that a strong cash flow/accrual relationship is
a measure of earnings quality—no longer holds. The attenuated correlation is partially
the result of economic shocks and the dominant impact of nonrecurring and nonoperating items
on cash flows and earnings.
How Is This Research Useful to Practitioners?
Researchers and academics understand that accruals serve to smooth temporary fluctuations
in cash flow. In the financial literature, the negative timing-related association between
accruals and cash flow is generally accepted. Because of the diminished relationship between
operating accruals and cash flows in recent years, practitioners may need to rethink their
reliance on the accrual/cash flow association as a measure of earnings quality.
Accounting principles have not changed significantly. Therefore, the smoothing role of
accruals has not changed. The authors explain 63% of the decline in the accrual/cash flow
correlation and find that temporal changes in the matching of expenses with revenue and the
growth of intangible-intensive industries contribute only slightly to the diminished
association of accruals and cash flows. The increased frequency and magnitude of
non-timing-related accruals (one-time and nonoperating items and loss years) explains the
decrease in the relative prominence of operating accruals’ smoothing effect on
According to the authors, nonoperating and one-time accounting events, as well as firm-year
losses, have become increasingly more prominent in the current business environment.
Analysts, academics, and investors who have traditionally focused on the information content
of financial statements, consistent reporting, and quality of earnings may need to
reevaluate the usefulness of accrual accounting and pay more attention to the one-time and
nonoperating revenue or expense items that sway earnings. Economic shocks and the resulting
losses may also indicate a need for changes in the traditional criteria for evaluation of
How Did the Authors Conduct This Research?
The authors regress total accruals on cash flows from operations for 217,164 firm-years
between 1964 and 2014. They find that the correlation between accruals and cash flows has
significantly diminished over the past 50 years. When total accruals are regressed on
contemporaneous operating cash flow, the negative accruals/cash flow relationship decreased
from $0.70 on $1.00 of cash flow in the 1960s to less than $0.10 since 2004. Furthermore,
the adjusted R2—the goodness-of-fit measure—has
decreased from 70% in the 1960s to near zero in more recent years. A similar decline in the
negative association of accruals and current cash flows is reported when accruals are
regressed on past and future operating cash flows.
The authors also examine firm-specific time-series regressions of accruals on cash flows.
They break the sample into four subperiods of time. The correlation between accruals and
cash flow diminishes progressively over time, and the R2 for the
firm-specific time-series model declines from a good fit of 84% in the 1964–75
subsample to 16% with the 2000–14 subsample, confirming the cross-sectional regression
Based on the findings of this study, the impact of one-time and nonoperating items
outweighs that of ordinary accruals on the smoothing of income.
Such one-time items as write-downs of goodwill or other assets—and understanding the
reasons behind them—should become a focus of financial analysis. The authors’
findings are relevant for research on acquisitions and subsequent one-time adjustments.
Analysts and researchers need to give attention to nonoperating items—for example,
investment-related activities of companies.