In crafting its original Statement of Financial Accounting
Standards (SFAS) for reporting about a firm’s lines of
business, the Financial Accounting Standards Board (FASB)
gave managers considerable discretion regarding what lines
of business are reported as separate business “segments” and
which ones should be aggregated into a reportable segment.
A segment is similar tobut not necessarily synonymous
witha business division, in that it may include several of a
firm’s divisions. Within General Electric, for example, the
various parts of the company’s financing arm might constitute
one segment, while its jet aircraft manufacturing operations
could constitute another. Managers can use this latitude to
conceal key information from outsiders.
On the basis of readily available data, it is clear many managers
do choose to hide some profit information. Managers
may not want to fully disclose abnormally high profits in a
segment out of concern that such a move would invite competitors
interested in capturing some of those profits. This
motive is palatable to shareholders, who may appreciate a
manager’s disinclination to share the good news with rivals
even if it also is withheld from owners.
On the other hand, abnormally low profits may not be fully
disclosed because doing so would stir unwanted scrutiny from
shareholders. This motive would likely alienate shareholders,
who would be inclined to demand answers from managers
overseeing poorly performing business segments.
In their recent study “Segment Profitability and the Proprietary
and Agency Costs of Disclosure,” University of Chicago
Graduate School of Business professor Philip G. Berger and
Rebecca N. Hann of the University of Southern California’s
Leventhal School of Accounting set out to determine what
motivates managers to aggregate (and thus, to effectively
hide) line-of-business profit data.
Accounting Change Offers Insights
When the FASB proposed changing the Statement of Financial
Accounting Standards from SFAS 14 to SFAS 131 in the late
1990s to mandate more disaggregated segment disclosure, 86
percent of firms commenting on the recommended change
opposed this move based on the contention that it would put
them at a disadvantage vis--vis competitors.
Accordingly, most studies of disclosure have focused on the
proprietary cost of disclosing abnormal profitabilityrisking
the loss of enviable profits to rivals. Berger and Hann argue
that the self-interested agency cost motive of hiding abnormal
profits to avoid external scrutiny also merits examination.
According to earlier studies, when the adoption of SFAS
131 made it more difficult for managers to lump segments
together to hide abnormal profits, a greater “diversification
discount” was created. The diversification discount is the
comparison of a firm’s stock price relative to the value of a
number of accounting variables, such as its sales or total
assets. A greater diversification discount is consistent with
a lower stock price relative to those accounting variables.
This would seem to suggest a motive for managers to
combine segments under SFAS 14 to hide poorly performing
segments. When greater disclosure more fully reveals what
the authors call “the extent of value destruction at an underperforming
firm,” there is a greater threat that the underperforming
manager will face discipline in the form of heightened corporate
governance and control mechanisms.
In conducting their research, Berger and Hann assembled
a sample of 796 firms that each reported multiple segments
under SFAS 131, with a total of 2,310 business segments
reported by these firms. Only firms reporting multiple segments
under SFAS 131 were included in the sample because
the authors sought to determine whether segments with
abnormal profits could have been lumped together with
others in the firm under the old SFAS 14 reporting rules.
Berger and Hann compared the segment information actually
released under the old rules of SFAS 14 in 1997 with the
restatement of 1997 segment information for the same firms
under SFAS 131, which went into effect in 1998. To ensure
they only captured restatement effects relating to the switch to
SFAS 131, they eliminated all firms with other changes taking
place during the SFAS 131 adoption year, including discontinued
operations and changes in inventory accounting methods.
This comparison allowed the authors to identify “new”
business segments disclosed under SFAS 131 but not revealed
under SFAS 14. The appearance of these segments under the
new regime indicated they had previously been grouped with
other business segments in an attempt by managers to avoid
fully disclosing the individual segments’ abnormal profits.
“Our maintained assumption is that SFAS 131 reporting,
which has to correspond to the internal structure used at the
firm to allocate resources, provides a more accurate view of
the firm’s true lines of business than
did SFAS 14, which was based on
vaguely defined industry categorizations,”
says Berger.
The focus on abnormal segment
profits was one of the key innovations
of the study. “Previous papers
have tried to measure the overall
firm’s abnormal profit as a way to
determine whether the management
wanted to hide something,” says
Berger. “The problem with that is
there is nothing hidden about a
firm’s overall profit.”
Line-of-business revenues are
often available in the management discussion and analysis
(MD&A) section of a company’s 10-K, but earnings figures are
generally found only in segment disclosures. For that reason,
the dominant reason for managers to aggregate or lump segments
together may be to hide abnormal segment profits.
What Constitutes Abnormal?
Berger and Hann compared the profits of new segments identified
under SFAS 131 with the profits of old segments under
SFAS 14, classifying abnormal profits as segment rates of
return greater or less than those of their industry.
However, this alone was not enough to determine whether
managers were guided by the proprietary cost or agency cost
motives in choosing nondisclosure.
Focusing first on the agency cost motive, the authors
sought to determine if in diversified firms, the new segments
that had been identified in the SFAS 131 restatement had
received subsidies from other segments of the company.
“It wasn’t whether they were simply being subsidized, but
whether they were being inefficiently subsidized,” Berger says.
“The investments that the hidden segments made with these
transfers were not good investment opportunities. By continuing
to transfer funds to a hidden segment, the firm might
extend the hidden segment’s existence beyond the time that it
would have survived had it been forced to stand on its own.”
Berger and Hann identified firms where inefficient transfers
occurred as firms with the agency cost motive to conceal
segments. All other firms were classified as having the proprietary
cost motive for segment aggregation.
In the agency cost motive sample, new segments earned
substantially lower average abnormal profits than the old
segments, ranging from 10 percent to 18 percent lower. This
suggested that managers were indeed hiding poorly performing
segments when they had the opportunity to do so combined
with the agency cost motive. In the proprietary cost motive
sample, however, results were not nearly as striking, with only
mixed evidence that the new segments earned higher average
abnormal profits than the old segments.
The Shareholders Benefit
In finding that the new segments had lower abnormal profits
than the old segments where the agency cost motive was suspected,
Berger and Hann were left with the obvious conclusion.
“The main implication is that the agency cost motive for
disclosure decisions exists and is important,” Berger says.
“Managers might publicly claim, ‘I don’t want to give away this
information to competitors and suppliers.’ They won’t say,
‘I’m not giving out this information because I want to hide my
own underperformance.’ So a survey of managers or a review
of their lobbying claims is unlikely to reveal an agency cost
motive for their reporting choices. In contrast, from our
empirical evidence it appears they were using their discretion
to hide information that shareholders wouldn’t want them
to hide.”
In instances where the proprietary cost motive was suspected,
results were mixed.
“Our study doesn’t necessarily refute earlier studies that
documented the existence of the proprietary cost motive’s role
in disclosure decisions,” notes Berger. “We do, however, find
less evidence consistent with the proprietary cost motive than
prior studies have.”
Berger and Hann’s findings suggest that if managers had
been left with the amount of discretion afforded them under
SFAS 14, they would be inclined to use that discretion in a way
that would not be optimal for shareholders. Another implication
of the findings is that the change in segment disclosure
under SFAS 131 may be beneficial to shareholders. The new
standard reduces agency costs by forcing managers to give
shareholders more information about segments, even if they
are underperforming. However, shareholders also may be hurt
by the greater disclosure forced upon managers by SFAS 131,
because the additional information may benefit competitors.
“Overall, SFAS 131 segment reporting may be better for
shareholders than SFAS 14 reporting,” Berger says. “Even
though the firm may be giving more proprietary information
to competitors when forced to use the new standard, it’s also
true that the firm receives more information from its publicly
traded competitors. The net increase in the proprietary cost of
disclosure under the new standard is thus unlikely to be high.”
Berger adds: “The new standard appears to be lowering the
agency costs of the firm’s disclosure policy without increasing
the proprietary costs much. More information allows outsiders
to monitor what the firm is doing. The benefit of that
probably outweighs the net cost to the firm of giving more
information to competitors while also receiving more information
from competitors.”
What has been shown to be a better standard in the United
States may be embraced internationally as well. The International
Accounting Standards Board (IASB) has proposed
changing its segment-reporting standard from the current
approach (which is similar to that of SFAS 14) to a new
approach similar to SFAS 131. Such a change would permit
additional research on the ways in which widely varying
disclosure practices in a number of other countries reflect
proprietary and agency cost concerns.
“Segment Profitability and the Proprietary and Agency Costs of Disclosure.” Philip G. Berger and Rebecca N. Hann. Forthcoming in The Accounting Review, 2007.