Review of Accounting Studies, 4, 145–162 (1999)
1999 Kluwer Academic Publishers, Boston. Manufactured in The Netherlands.
On Transitory Earnings
JAMES A. OHLSON
New York University, Leonard N. Stern School of Business, 40 West 4th Street, New York, NY 10012
Abstract. The paper develops a concept of transitory earnings and contrasts this source of earnings to “core”
(or recurring) earnings. It is shown that any two of the following three attributes of transitory earnings imply
the third: (i) forecasting irrelevance with respect to next-period aggregate earnings, (ii) value irrelevance, and
(iii) unpredictability. The paper makes the case that the current “dirty surplus” items make sense, especially if
one expands the valuation perspective to also allow for agency considerations.
This paper considers concepts of “transitory earnings,” and analyzes how this source of
earnings differs from other income items. Traditional accounting theory, the policy precept
dirty surplus, and practical ﬁnancial statement analysis all recognize that some earnings
sources can be characterized as transitory. It is well established that such earnings need to
be separated, if not completely eliminated, from the income statement. Casual observation
suggests a number of speciﬁc attributes associated with transitory earnings. First, they
are unpredictable, at least in the sense that current transitory earnings do not inﬂuence
subsequent transitory earnings. Second, current transitory earnings are irrelevant when one
forecasts totalearnings for the subsequent year. Third, the line item transitory earnings plays
no informational role when one estimates the present value of a ﬁrm’s expected dividends.
After having formalized these ideas, the basic analytical result shows the following: any
two of the aforementioned three attributes imply the third; further, any one of the three
attributes taken individually leads to no implications concerning the remaining two. The
analysis uses Ohlson’s (1995) framework, but with earnings split into two components (i.e.,
core earnings plus transitory earnings equal net comprehensive earnings).
The paper goes beyond ﬁtting together various attributes of transitory earnings. Key parts
of the analyses validate the idea that the information inherent in transitory earnings has
much in common with dividends. From an information perspective, the effect of dividends
on valuation and earnings forecasting is basically captured via its effect on book value.
The same is true for transitory earnings. As a consequence, there is no loss of information
for purposes of valuation and earnings forecasting if one nets transitory earnings against
dividends. It is shown that core earnings, end-of-period book value, and start-of-period
book value summarize the relevant accounting information. From this information one can
only infer the net of transitory earnings minus dividends, not the two components. Stark
(1997) makes similar points in a slightly different model.
Having established informational similarities between transitory earnings and dividends,
a key difference is noted. In the relation that explains market returns in terms of the “new”
information, a cent of transitory earnings (deﬂated by start-of-period market value) equals
a cent of return. Unexpected dividends, in sharp contrast, have no effect on the return