2 Tax Motivated Loss Shifting

THE ACCOUNTING REVIEW American Accounting Association
Vol. 88, No. 5 DOI: 10.2308/accr-50496

2013
pp. 16571682
Tax-Motivated Loss Shifting
Merle M. Erickson
The University of Chicago
Shane M. Heitzman
University of Rochester
X. Frank Zhang
Yale University
ABSTRACT: This paper examines the implications of tax loss carryback incentives for
corporate reporting decisions and capital market behavior. During the 1981 through 2010
sample period, we find that firms increase losses in order to claim a cash refund of recent
tax payments before the option to do so expires, and we estimate that firms with tax
refund-based incentives accelerate about $64.7 billion in losses. Tax-motivated loss
shifting is reflected in both recurring and nonrecurring items and is more evident for
financially constrained firms. Analysts do not generally incorporate tax-motivated loss
shifting into their earnings forecasts, resulting in more negative analyst forecast errors
for firms with tax-based incentives than for firms without. Holding earnings surprises
constant, however, investors react less negatively to losses reported by firms with tax
loss carryback incentives.
Keywords: taxes; net operating losses; liquidity; analyst forecasts; capital markets.
Data Availability: Data are available from sources identied in the paper.
I. INTRODUCTION
T
ax rules give the rm an option to obtain a cash refund of recently paid taxes by reporting a
tax loss. This cash infusion can be particularly valuable for nancially constrained rms. In
fact, the purported liquidity benets of these tax loss carrybacks played a central role in
at least two attempts to stimulate the business sector during the past decade, and refunds of
We appreciate comments from John Harry Evans III (senior editor), two anonymous referees, and workshop participants
at the National University of Singapore, Singapore Management University, Yale University, and the Iowa Tax Readings
Group. Professor Erickson appreciates the nancial support of the Booth School of Business.
Editors note: Accepted by John Harry Evans III.
Submitted: January 2012
Accepted: April 2013
Published Online: April 2013
1657
corporate tax payments by the federal government exceeded $95 billion in both 2009 and 2010.
1
In
this paper, we address the accounting and economic consequences of allowing rms to carry back
tax losses. Specically, we investigate whether rms increase their reported tax losses to obtain
cash refunds of prior tax payments, whether nancial analysts anticipate the effect of these tax
incentives on reported earnings, and how investors respond to the earnings news for rms with
tax-based incentives to report losses.
U.S. tax law limits corporate tax refunds to the taxes paid in the most recent (currently two) years.
Thus, for taxes paid on prots in year t2, year t is generally the last year the rmcan claim a refund of
those taxes. To reduce the rms expected tax liability, and therefore increase the rms after-tax cash
ow, the manager will often have an incentive to report tax losses in year t to maximize the cash refund
of taxes paid on income in t2. When the rm exhausts its capacity to carry a loss back to get a refund,
or delays recognizing a tax loss altogether, liquidity benets disappear and the economic value of the
tax loss is discounted because of uncertainty and the time value of money.
2
Our study complements and extends other research that focuses on corporate and market
responses to the Tax Reform Act of 1986 (TRA 86). The signicant decline in corporate marginal
tax rates resulting from TRA 86 created strong incentives for managers to accelerate losses (Scholes
et al. 1992; Guenther 1994; Maydew 1997) so that the losses would apply under the higher tax rates
and thus reduce current taxes by a larger amount. Tax rate shifts of that magnitude are infrequent, so
evidence on the inuence of loss-shifting incentives when statutory tax rates are relatively constant,
as in the last 25 years in the U.S., is important. A rm experiencing a negative economic shock will
often be in a position to decide between accelerating the recognition of a loss for a certain and
immediate cash refund or deferring the recognition of the loss for an uncertain and discounted tax
benet. These negative shocks can be idiosyncratic or systematic, as in times of recession, and thus
tax-based incentives for accelerating tax losses represent a pervasive issue for managers, policy
makers, and capital market participants.
Each year we identify a set of rms in a position to recognize a tax loss in the current period
that would provide a cash refund of prior tax payments that otherwise would be lost after year-end.
These are rms that (1) paid income taxes on prots in the earliest carryback year that have not yet
been refunded, and (2) are expected to report a loss in the current year. The rst condition ensures
that there is actually a cash refund available. The second condition narrows our focus to those rms
with expected benets from tax-motivated loss shifting.
We nd that rms with tax-motivated loss-shifting incentives report signicantly larger losses
than a comparison sample of rms that are also expected to report a loss but do not have access to a
potential cash tax refund. The loss equates to 11.25 percent lower reported earnings for the average
loss rm with carryback incentives versus comparable rms and implies that incentives for
tax-motivated loss shifting play an important role in reported earnings. These accelerated losses
1
A comment letter from Financial Executives International to congressional leaders dated January 13, 2009
stated, Extending the carryback period to ve years will enhance liquidity of businesses with current losses,
while helping to insulate against future losses. This provides companies with more capital to make investments
that will help move the economy forward. In a speech on the oor of the Senate on November 3, 2009, Senator
Patty Murray of Washington State argued that the Worker, Homeownership, and Business Assistance Act of
2009 would also provide a critical boost to businesses . . . by extending their ability to carry back losses theyve
suffered in 2008 or 2009. This tax provision will provide badly needed capital to help companies avoid layoffs,
expand their operations, and create jobs. See also Leone (2010).
2
The uncertainty in the tax loss carryforward is driven by uncertainty about future taxable income and the
possibility that a future ownership change will put a binding constraint on the rms ability to use these
carryforwards. With respect to the latter, a number of rms have adopted so-called NOL poison pills designed
to prevent the constraint triggered by Section 382 of the Internal Revenue Code, which limits tax loss usage
following a change in ownership (Erickson and Heitzman 2010). Section 382 concerns increase the incentive to
accelerate tax losses.
1658 Erickson, Heitzman, and Zhang
The Accounting Review
September 2013
reduce operating income and increase one-time losses. In addition, we nd that nancially
constrained rms more aggressively accelerate losses to obtain cash refunds. This result supports
the idea that these rms view tax refunds as an important source of cash and implies that liquidity
benets outweigh the incremental direct and indirect costs of accelerating the tax loss.
We then investigate whether and how nancial analysts account for these incentives when
forecasting earnings. A large body of literature argues that analysts are sophisticated and that their
forecasts are a reasonable proxy for market expectations.
3
However, studies based on nancial
reporting and tax law events suggest that analysts often do not fully utilize the information in tax
footnote disclosures (Amir and Sougiannis 1999) and have difculty incorporating the predictable
effects of tax law changes (Chen and Schoderbek 2000; Plumlee 2003; Shane and Stock 2006). We
expand upon these studies by not constraining our analysis to one-time events. Although managers
respond to tax-loss reporting incentives in predictable ways, we nd that analysts do not incorporate
these incentives into their earnings forecasts.
4
Analysts forecasts are signicantly less precise for rms
with tax incentives to shift losses than for comparable loss rms without these tax incentives. These
results are not driven by systematic differences in the measurement of forecasted and actual earnings.
Finally, we evaluate how investors react to the earnings of rms with incentives to accelerate
tax losses. Although these rms tend to have negative earnings surprises based on past earnings and
analyst forecasts, the markets reaction to the news is less negative for those rms with incentives to
accelerate tax losses. This result implies that investors recognize the value inherent in tax-motivated
loss shifting. The results suggest that tax-motivated losses represent positive decisions by managers
to increase rm value in the presence of potential contracting and nancial reporting consequences.
To put the main results of our paper in context, consider homebuilder Lennar Corporation.
5
In
scal years 2005 and 2006, current tax expense gures indicate that Lennar incurred federal and
state income tax liabilities of $805 million and $547 million, respectively. Thus, 2007 was the last
year Lennar could claim a refund for the $805 million of taxes paid in 2005. Faced with a continued
decline in the homebuilding sector, Lennar took several actions to generate tax losses at the end of
2007, which was two years before the temporary extension of the loss carryback window. For
example, in the nal month of the 2007 scal year, Lennar sold land to a partnership they formed
with Morgan Stanley, generating an $800 million tax loss for Lennar (Lennar 10-K). Lennar also
wrote off $530 million in deposits and pre-acquisition costs for building lots the company decided
not to pursue. According to the CEO of Lennar, As a by-product of our strategic fourth quarter
3
See Brown et al. (1986), OBrien (1987), and Doyle et al. (2006). During the earnings season, Wall Street
typically compares a rms reported earnings with analysts most recent forecasts to assess earnings surprises.
Prior research concludes that similar patterns between analyst forecast errors and stock returns also support this
assumption. For example, Sloan (1996) shows that high-accrual rms have lower future stock returns, whereas
Bradshaw et al. (2001) nd that high-accrual rms have overly optimistic earnings forecasts. Zhang (2006a)
shows that investors underreact more to new information in cases of greater information uncertainty, whereas
Zhang (2006b) nds a similar pattern for analysts.
4
The results are consistent with the idea that analysts either do not have the ability to incorporate the tax-
motivated loss shifting or that analysts understand such corporate reporting behavior but choose not to
incorporate it in their earnings forecasts. Our evidence is more consistent with the rst view. In Table 4, we nd
analyst forecast errors are more negative for rms with carryback incentives. In the Analyst Revisions of Future
Earnings section, we nd that analysts gradually incorporate the effect of NOL-related loss shifting in their
forecasts of current years earnings while keeping next years forecasts relatively unchanged.
5
While Lennars numbers may be unusually large, such strategic choices and corporate decisions are likely to
apply to other companies, a phenomenon that underpins the key point of the paper. For example, one week
following the passage of the ve-year carryback window on November 6, 2009, William Lyon Homes revealed
in the companys 10-Q ling that, In considering ways to maximize such tax refund, the Company is
determining whether to elect to defer certain cancellation of indebtedness income generated from its repurchase
of Senior Notes during 2009. The company subsequently recorded losses through the sale of assets, and by the
end of scal year 2009 expected a federal income tax refund of $101.8 million. This would nearly double the
rms cash balance. See also Leone (2010).
Tax-Motivated Loss Shifting 1659
September 2013
moves, we have generated losses that have resulted in the receipt of a cash tax refund of $852
million subsequent to the close of the quarter (press release, January 24, 2008).
6
The week before
earnings were released, analysts forecasted losses ranging from $0.00 to $4.45 per share, with an
average expected loss of $1.84 per share. Lennar surprised analysts by reporting a loss of nearly
$7.92 per share. With 159.9 million shares outstanding, this translates to an unexpected accounting
loss of nearly $972 million. Nevertheless, the market reacted positively to the news, with Lennars
three-day stock return around the earnings announcement date exceeding 25 percent, reecting a
$493 million increase in market value.
Our studys rst contribution is to provide evidence that rms accelerate tax losses to obtain
cash inows through refunds of prior tax payments, even when statutory tax rates are constant. We
show that this tax-motivated loss shifting is reected in both recurring and nonrecurring items and
is more pronounced for nancially constrained rms. To our knowledge, this paper is one of the
rst to link liquidity demands to tax and nancial reporting decisions. Second, we provide a broader
examination of the capital market implications of tax loss carryback incentives. We nd that
analysts do not incorporate tax loss carryback incentives into their earnings forecasts, yet
stockholders react less negatively to reported losses when rms have tax incentives to accelerate
losses. Taken together, our evidence provides new insight on how managers and capital market
participants incorporate tax-based incentives to accelerate losses into their decision-making. Our
evidence is also relevant to understanding the growing importance of tax losses on rm value, as tax
losses are becoming increasingly important for scal and corporate policy decisions (Graham and
Kim 2009; Erickson and Heitzman 2010). Overall, our evidence suggests that the tax-based
incentive to accelerate reported losses plays a material and persistent role in corporate reporting
decisions and capital market activities.
Section II next reviews prior literature and develops hypotheses. Section III describes the data
and provides summary statistics. Section IV presents the results. Section V provides a variety of
additional analyses, and Section VI concludes.
II. PRIOR LITERATURE AND HYPOTHESIS DEVELOPMENT
Managers have incentives to reduce the rms total tax liability over the life of the rm because
a dollar less paid to the tax authority is a dollar more for shareholders. Moreover, rms face
asymmetric tax treatment of prots and losses because taxes are paid immediately on prots, but
taxes are not necessarily refunded on losses. This means that a rm with zero expected pretax
income will still have a positive expected tax liability that is increasing in income uncertainty
(Scholes et al. 2008, 172). Thus, Graham and Smith (1999) show that rms facing such a convex
tax schedule have incentives to hedge in order to reduce expected tax liabilities.
Net operating loss provisions contained in the tax code partially mitigate this asymmetry. To
illustrate, a rm that paid taxes in the recent past can claim a refund of those taxes in a year in which
it reports a loss. This is achieved through tax loss carryback rules that allow the rm to use its tax
loss in the current year to reduce taxable income in a prior tax year (i.e., a carryback of the current
tax loss for a refund of prior tax payments), starting with the earliest tax year of the carryback
period. Put differently, when the rm pays taxes on income, it effectively gets an option to claim a
refund of those taxes that expires after the length of the carryback window, T. This option gives
some rms an incentive to accelerate their losses to generate cash ow through a refund of prior tax
6
These tax losses were also reected in Lennars nancial statements, albeit in a somewhat different form. The
transaction was treated as a sale for tax purposes, but not for GAAP purposes, because Lennar retained 50
percent of the voting rights in the partnership. In Lennars nancial statements, the decline in the value of these
assets was recorded, but as an asset impairment rather than as a loss on sale.
September 2013
payments because otherwise the option will expire. For example, assume the loss carryback period
(T) is two years. If a rm paid taxes on prots in year t2 and those taxes have not been refunded
through losses in year t1, then the rm will have an incentive to accelerate losses into year t in
order to maximize a refund of taxes paid in year t2. If the rm did not report taxable income in a
recent year, or there is no tax to be refunded, then it can carry the current years tax loss forward and
use it to reduce taxable income in a future year. For example, a startup rm with accumulated losses
cannot claim a refund of taxes they have not paid. Current tax losses are carried forward to reduce
taxable income if and when the rm becomes protable. If the rm does not generate enough future
taxable income to use the loss before the carryforward period expires, then the loss expires unused.
An accelerated tax loss reduces current year taxes, and if the loss is large enough, leads to a
cash refund. But for the protable rm this strategy will only increase the tax liability in the
following period. Because consistently protable rms face more symmetric tax treatment, they
derive fewer benets from accelerating losses (Graham and Smith 1999). Thus, the present value of
the cash tax benet from accelerating a loss is strongest if the rm is already in a tax loss position
(that is, before considering the tax incentives to report additional losses) and does not expect to
immediately return to protability. These rms obtain immediate and certain tax benets by
accelerating the recognition of the loss to recoup prior tax payments and face uncertain and
discounted tax benets if they do not. The benet of accelerating those losses to generate cash ows
is therefore stronger when the rm expects losses in future periods as well.
Shifting a tax loss to generate a refund requires the manager to alter real decisions, reporting
decisions, or both. Since the incremental benet of the refund must be weighed against the
incremental cost to shareholders, the expected tax benets could go unclaimed if accelerating a loss:
(1) involves costly real actionssuch as disposing of productive assets or deferring sales, (2)
generates nancial reporting costssuch as violating a debt covenant, or (3) increases the taxing
authoritys scrutiny of the rms tax positions.
7
Thus, whether the incentive to accelerate a tax loss
has a material effect on corporate reporting and capital market activity is an open question.
The incentives for tax-motivated loss shifting increase when the marginal tax rate during the
carryback window exceeds the expected marginal tax rate during current and future periods. Prior
research examines rms reporting behavior around TRA86, which reduced the top statutory corporate
tax rate from 46 percent in 1986 to 34 percent in 1988 (Scholes et al. 1992; Guenther 1994; Maydew
1997; Shane and Stock 2006). Among all rms that report tax losses during a ten-year window,
Maydew (1997) nds that rms appear to report larger losses when the relative tax benet of the
carryback, measured as difference between tax rates in the current and carryback years, is greater. He
nds that loss-shifting actions are evident in both operating income and nonrecurring losses.
We extend this research to a general setting in which statutory tax rates are effectively constant.
Maydew (1997) suggests that rms facing losses always have incentives to increase their refund of
prior years taxes for at least two reasons. First, the cash ows from tax refunds are certain, whereas
expected cash ows from operations are not. Second, the time value of money provides an incentive
to defer income and accelerate deductions. The ability to claim a certain refund of cash taxes paid is
permanently lost when the carryback window closes, substantially reducing the present value of the
tax loss.
8
We extend these points by emphasizing that a tax refund represents real cash inows that
7
There is an extensive literature that analyzes how managers consider the tradeoff between taxes (benets and
costs) and GAAP accounting effects. See for example Matsunaga et al. (1992), Engel et al. (1999), Shackelford
et al. (2010), and Hanlon and Heitzman (2010).
8
Consider the change in the NOL carryback period in 20082010. The extension of the carryback window from
two to ve years was motivated by a desire to provide a refund of prior cash taxes paid to rms in a difcult
economic climate (Graham and Kim 2009). Such tax refunds were unavailable to rms prior to the law changes
because the carryback period prevented rms from claiming refunds of taxes paid in years outside the then-
current two-year carryback window.
September 2013
provide liquidity. The liquidity motivation is likely to be more relevant for loss rms facing
difculty raising external capital. In essence, even in periods of constant statutory tax rates, rms
will have incentives to accelerate losses.
9
This leads to the following prediction:
H1: Reported earnings are decreasing in tax loss carryback incentives.
In testing this hypothesis, we explicitly consider variation in the costs and benets to rms that
execute this strategy. For rms that are persistently protable and consistently pay taxes, the
benets of accelerating a loss to claim a tax refund are either unavailable or too small to make a
difference. Thus, more powerful tests of the hypothesized behavior require comparing rms with
tax-motivated loss-shifting incentives, which we operationalize as rms that paid taxes during the
earliest year of the carryback window and expect to have losses in the current year, to rms without
such incentives, which are then rms with similar expected losses but no tax payments to recoup.
Our tests assume that the loss reported to the tax authority is also reected in GAAP earnings, so the
incremental tax benets from increasing a tax loss should be weighed against the incremental costs
of increasing an accounting loss, such as violating a debt covenant.
If tax rules create incentives for rms to increase reported losses, then a natural question is
whether analysts anticipate corporate responses to these incentives when forecasting earnings.
Analysts are often viewed as sophisticated users of accounting information, but their forecasts may
ignore carryback-based incentives if the tax disclosures are complex or provide noisy signals of true
tax status (Chen and Schoderbek 2000; Dhaliwal et al. 2004). For example, Amir and Sougiannis
(1999) nd that analysts do not incorporate the information in deferred taxes contained in the SFAS
No. 109 disclosure, while Chen and Schoderbek (2000) document that analyst forecasts do not
incorporate the predictable earnings effect from the revaluation of deferred tax assets and liabilities
following a one percentage point increase in the top marginal corporate tax rate in 1993 (from 34
percent to 35 percent). Based on capital market responses to TRA 86, Plumlee (2003) nds that
analysts do not incorporate the impact of tax incentives and tax disclosures, particularly for more
complex tax issues.
10
Shane and Stock (2006) show that analysts do not incorporate income
shifting induced by the 1986 tax rate change when forecasting earnings.
While the existing evidence is informative, it is limited to a handful of events that radically
changed tax law or GAAP disclosures of tax circumstances. Even if analysts ignore the impact of
one-time macro events, the tax benets of accelerating losses are recurring and apply to a signicant
set of rms every year. This provides a stronger case for analysts to forecast such information. This
reasoning leads to the following hypothesis:
H2: Analyst forecasts do not incorporate tax loss carryback incentives.
Finally, we examine the equity markets reaction to tax-motivated loss shifting. If analysts
do not incorporate tax-motivated loss shifting in their earnings forecasts, then their estimates
9
Like Maydews (1997) focus on declining statutory marginal tax rates, our setting can potentially be interpreted
as a decline in expected marginal tax rates, holding statutory tax rates largely constant. In other words, a rm that
accelerates a loss to claim a refund of taxes paid in a prior year essentially secures a tax benet at an
undiscounted statutory tax rate. But if management decides to wait to report the loss, then it is less likely the rm
will be able to carry the loss back for an immediate refund, and instead the rm would have to carry the loss
forward to offset income in some future period, reducing the effective tax benet of the deduction. Direct
estimates of the marginal tax rate, such as the simulated tax rates of Shevlin (1990), Graham (1996), and Blouin
et al. (2010), are not well suited to our analysis for several reasons. First, they provide an estimate of the
expected marginal tax rate for current year prots and losses, but not the expected marginal tax rate in future
years. Second, they do not directly address the dollar amount of potential refunds from carrybacks. Third, these
measures are endogenous to the reporting decisions we are analyzing.
10
Moreover, Outslay and McGill (2002) provide evidence that the tax disclosures of some rms are quite difcult
to interpret and understand.
September 2013
are likely to be systematically optimistic. Such optimism would lead to negative earnings
surprises as reected in analysts forecast errors, measured as actual earnings minus forecasted
earnings. However, unexpected losses that generate cash tax refunds arguably create value
by reducing the present value of taxes paid and providing liquidity, so the markets response to
losses motivated by cash tax benets should be tempered relative to the rm whose losses are
not. Prior evidence suggests that investors do not understand the implications of tax incentives
on reported earnings (Shane and Stock 2006). However, there is some evidence that
informative disclosure of the tax-based reasons behind the earnings surprise leads to more
efcient market responses (Chen and Schoderbek 2000). We predict that managers have
incentives to disclose the transitory nature of tax-motivated losses to investors through
nancial reporting disclosures, conference calls, and future earnings guidance. This leads to the
nal hypothesis:
H3: The market reacts less negatively to earnings surprises of rms with tax-motivated loss-
shifting incentives.
III. SAMPLE DATA AND DESCRIPTIVE STATISTICS
Each year, we identify rms that have the incentive to accelerate tax losses to obtain a tax
refund by analyzing the time-series of estimated taxable income.
11
We rst calculate tax loss
carryback capacity (NOLC), which is an estimate of refundable income taxes paid in the earliest
year of the carryback period in year t and is described in Appendix B.
12
The length of the carryback
period ranges from two to ve years over our sample period. Unrefunded tax payments in the
earliest carryback year will expire if the rm does not claim a refund in year t. Thus, our main test
variable (D_NOL) for the period is an indicator variable equal to 1 if in year t the rm has
unrefunded tax payments on income in the earliest carryback year and analysts expect the rm to
report a loss in year t. This approach identies a set of rms that will lose the ability to claim a
refund of taxes paid in a prior year (D_NOL) and the corresponding amount of potentially
refundable taxes (NOLC).
Because tax returns are unobservable, we follow prior research and rely on GAAP earnings
numbers to identify tax-motivated loss shifting (Scholes et al. 1992; Guenther 1994; Maydew 1997;
Shane and Stock 2006). Thus, we focus on forecasts and realizations of GAAP earnings and assume
that book earnings reect the underlying tax loss reporting. We expect GAAP earnings numbers to
11
An alternative research design is to focus on tax law changes, as in Maydew (1997), who compares rm-years
with loss carrybacks during a period immediately after TRA 86 to rm-years with loss carrybacks in other
periods. This alternative setting differs from ours in two primary ways. First, Maydew (1997) tests the additional
tax incentives introduced by TRA 86, whereas we are interested in the general phenomenon whereby rms
always have incentives to shift income and carryback losses. Second, Maydew (1997) conducts an ex post
analysis using rm-years with loss carrybacks. We do not require rms to have loss carrybacks, because a
variable based on realized loss carrybacks would have a look-ahead bias in our capital market tests (that is, we
are interested in capital market responses to expected tax loss shifting, which is measured prior to the return
window). Instead, we focus on an ex ante variable and predict whether rms and the capital market behave in
certain ways. Maydew (1997) carefully controls for the look-ahead bias issue because he takes rm-years with
loss carrybacks in other periods as the benchmark and examines the impact of additional tax incentives
introduced by TRA 86.
12
Ideally, we would incorporate the rms actual net operating loss carryforwards to calculate tax status. NOL
carryforwards can arise from domestic, foreign, and local tax jurisdictions, but we do not have access to rms
tax returns, and information on the source of the NOL carryforwards is inconsistently disclosed in nancial
reports. Moreover, Compustat provides a single data item for NOL carryforwards, and this has been shown to
have signicant measurement problems (Mills et al. 2003).
September 2013
be a noisy proxy for the reported tax numbers, and therefore this approach works against nding
evidence consistent with our predictions.
13
The ex ante nature of the D_NOL measure of the tax incentives to shift losses is important when
drawing inferences about capital market behavior. All information to construct D_NOL is available at
the beginning of year t. In this way, we test whether measures of tax loss carryback incentives are
associated with subsequent actions of managers, analysts, and investors. Because we focus on a single
year of tax payments and require that the rm have an expected loss for the year, D_NOL is a
conservative measure of the incentive to shift losses into the current year. We compare rms with
incentives to accelerate tax losses (D_NOL 1) to a set of rms also expected to report a loss, but
without cash taxes available for refund. To do this, rst we dene an indicator variable D_NEG equal
to 1 if analysts forecast a loss eight months before the scal year-end, independent of tax carryback
opportunities. Because D_NOL is an interaction between D_NEG and the indicator for potential tax
refunds, the coefcient on D_NEG is the estimated effect for expected loss rms without carryback
opportunities, while the coefcient on D_NOL represents the incremental effect for expected loss rms
with carryback opportunities.
We focus on the managers nancial reporting decisions to understand analysts and investors
reaction to tax-motivated loss shifting. We include all rm-year observations with non-missing
earnings or analysts earnings forecasts, resulting in a nal sample of 99,564 rm-year observations
from 1981 to 2010. Table 1 provides summary statistics of the primary variables used in this study.
Annual earnings, as a percentage of stock price, average 8.4 percent with a median of 4.4 percent.
Unexpected earnings, based on the difference between reported earnings and analysts forecasts
eight months prior to year-end (scaled by stock price), has a mean of 3.1 percent and median of
0.4 percent, respectively. With regard to nancial variables, the average rm in our sample has a
market value (MV) of $2.967 billion and a book-to-market equity ratio (BM) of 0.576. The average
book leverage ratio for sample rms is 22.2 percent, while EBITDA averages 11.2 percent of total
assets. All nancial variables are measured at the end of year t1. On average, about 10.8 percent of
rms are expected to report losses in our 1981 through 2010 sample period (D_NEG 1). Of all
such rms with expected losses, the subset of rms with the incentive to accelerate tax losses under
our denition (D_NOL 1) account for 2.6 percent of all rm-year observations, compared with
7.8 percent (10.8 percent 2.6 percent) for loss rms without these incentives.
IV. EMPIRICAL EVIDENCE
Evidence of Tax-Motivated Loss Shifting Based on a Time-Series Earnings Model
To test whether rms with tax-motivated loss-shifting incentives increase losses, we estimate
the following regression that includes a set of control variables:
E
t
b
0
b
1
D NOL b
2
D NEG b
3
E
t1
b
4
E
t2
b
5
LOSS
t1
b
6
LOSS
t2
b
7
E
t1
LOSS
t1
b
8
E
t2
LOSS
t2
b
9
LogMV
t1
b
10
ACC
t1
b
11
RET
t1
e
t
; 1
13
Tax reporting rules generally start with nancial reporting rules, but there are signicant exceptions. First, the
GAAP nancial statements will tend to consolidate more entities than the rms U.S. tax return. GAAP
effectively requires the rm to consolidate all entities owned at least 50 percent, whereas the U.S. tax return
ignores foreign subsidiaries and those where the rm owns less than 80 percent. Second, tax rules often require
an actual transaction to record a loss. For example, an impairment of an assets value would show up as an
expense on the GAAP nancial statements but would be absent from the rms tax returns, and hence the
calculation of taxable income, until the rm actually sells the asset. Maydew (1997) nds that income shifting
around the 1986 Tax Act was concentrated in gross margin and SG&A numbers.
September 2013
TABLE 1
Descriptive Statistics
Panel A: Univariate Statistics
Variable Mean Std. Dev. Min. Q1 Median Q3 Max.
E
t
0.084 0.768 15.620 0.008 0.044 0.073 0.445
FE
t
0.031 0.117 1.556 0.028 0.004 0.005 0.617
RET
t
0.135 0.676 0.997 0.211 0.060 0.346 22.881
ARET
t
0.000 0.050 0.664 0.022 0.000 0.023 0.978
D_NOL 0.026 0.160 0.000 0.000 0.000 0.000 1.000
D_NEG 0.108 0.310 0.000 0.000 0.000 0.000 1.000
MV
t1
2.967 13.941 0.000 0.099 0.328 1.239 508.329
BM
t1
0.576 0.476 1.321 0.278 0.479 0.754 5.699
RET
t1
0.192 0.823 0.996 0.189 0.083 0.383 53.660
ACC
t1
0.032 0.090 0.407 0.076 0.036 0.005 0.372
LEV
t1
0.222 0.201 0.000 0.044 0.186 0.345 0.954
EBITDA
t1
0.112 0.178 1.175 0.074 0.134 0.197 0.538
COV
t
7.090 6.990 1.000 2.000 5.000 10.000 50.000
Panel B: Correlation Matrix
E
t
FE
t
RET
t
D_NOL D_NEG MV
t1
BM
t1
COV
t1
E
t
1
FE
t
0.38 1
RET
t
0.15 0.16 1
D_NOL 0.11 0.09 0.02 1
D_NEG 0.24 0.06 0.00 0.43 1
MV
t1
0.03 0.05 0.02 0.02 0.06 1
BM
t1
0.11 0.19 0.10 0.09 0.01 0.09 1
COV
t
0.07 0.11 0.00 0.04 0.13 0.31 0.14 1
RET
t1
0.11 0.15 0.06 0.07 0.06 0.01 0.19 0.00
E
t
is annual Compustat earnings before extraordinary items per share scaled by stock price at the rms scal year-end.
Forecast error (FE) is calculated as the difference between the actual earnings and the median analyst forecast made eight
months prior to scal year-end, scaled by stock price at the forecast date. RET
t
is year ts annual returns starting eight
months prior to scal year-end. ARET
t
is average three-day earnings announcement return in year t, measured as raw
returns minus value-weighted market returns over the three-day [1, 1] period, where day 0 is the earnings
announcement date. NOLC is the net operating loss carryback capacity limit (see Appendix B for variable measurement).
D_NOL is a dummy variable with the value of 1 if NOLC is positive for a rm in an expected loss position, and 0
otherwise, indicating the rms incentives to carry back NOLs and to claim tax refunds. D_NEG is a dummy variable
with the value of 1 if analysts forecasts of year ts earnings are negative, where forecasts were made eight months prior
to scal year-end. MV is a rms market value of equity at the end of year t1. BM is the book-to-market ratio at the end
of year t1. COV is the number of analysts covering the rm at the forecast date. ACC is total accruals scaled by average
assets. LEV is the leverage ratio. EBITDA is earnings before interest, taxes, depreciation, and amortization scaled by
average total assets. Please see Appendix A for detailed variable denitions. The sample includes 99,564 rm-year
observations with non-missing observations of FE or annual earnings from 1981 to 2010. All variables except for
D_NOL, D_NEG, COV, and return variables are winsorized at the 1st and 99th percentiles.
September 2013
where:
E the reported earnings per share scaled by the rms stock price at scal year-end;
D_NOLa dummy variable indicating a rms incentive to carry back tax losses to claim a tax
refund;
D_NEGa dummy variable that takes a value of 1 if analysts are forecasting negative annual
earnings for year t four months after year t begins;
E
t1
and E
t2
reported earnings for years t1 and t2, respectively;
14
LOSS
t1
(LOSS
t2
) a dummy variable taking a value of 1 if E
t1
(E
t2
) is negative, and 0
otherwise;
MV a rms market value of equity at the beginning of the year;
ACC total accruals scaled by average assets; and
RET the 12-month buy-and-hold return beginning eight months before a rms scal year-
end.
Earnings variables and ACC are winsorized at the 1st and 99th percentiles.
Table 2 presents results for three progressively richer regression models, with D_NOL as the
main variable of interest. We expect the coefcient on D_NOL to be negative, reecting the
hypothesized tax-motivated loss shifting. Because expected earnings for D_NOL rms are negative
by denition, we include D_NEG in the regression to ensure that the coefcient on the D_NOL
variable does not pick up any difference in earnings surprises between prot and loss rms (Hayn
1995). Because D_NOL is implicitly an interaction between D_NEG and an indicator for carryback
potential, the coefcient on D_NOL reects the incremental loss reported by rms with an option to
use losses to obtain a cash refund of taxes paid in the earliest carryback year.
Across the three specications of Equation (1) in Table 2, the coefcient on D_NOL is
signicantly negative. For example, in column (1), the coefcient on D_NOL is 0.073 (t 2.61),
indicating that earnings are about 7.3 percent of prior-year-end market value lower in years when
the rm has a tax-based incentive to accelerate losses.
15
The magnitude of D_NOL is also
economically signicant. As the coefcient on D_NEG is 0.157 (t 7.90), we interpret the
results to mean that reported earnings are 46 percent (0.073/0.157) lower for loss rms with tax-
based incentives to accelerate losses than for loss rms without those incentives. This result holds
across specications with additional controls in columns (2) and (3). Overall, the results in Table 2
provide consistent support for the prediction that rms engage in tax-motivated loss shifting even
when statutory tax rates are constant.
Firms have a number of ways to accelerate losses, such as frontloading expenses, writing down
inventory, and selling assets at a loss. To provide more evidence on the drivers of reported losses
for rms with tax loss carryback incentives, we analyze recurring and nonrecurring items during the
event years, and compare them to both previous and subsequent years. As the literature offers no
well-established model for earnings components, we adopt a simple random walk model and use
both previous and subsequent years as the benchmark.
14
Following the capital markets research, we scale earnings and earnings surprises by stock price. In this way, the
deator is stock price for both stock returns and earnings surprises in the return regressions.
15
The economic magnitude is large for two reasons. First, we use annual earnings, and some rms report large
losses relative to their market values. Second, we winsorize the data and do not introduce any ad hoc cutoff of
data based on reported earnings. If we delete observations with earnings scaled by market value in the top and
bottom 1 percent, the coefcient on D_NOL becomes 0.049 ( p , 0.01). If we delete observations with an
absolute value of earnings scaled by market value above 1, the coefcient on D_NOL becomes 0.026 ( p ,
0.01). Although the magnitude of the coefcient drops as we drop more observations with extreme values, the t-
statistics and p-values change very little. In all our tables, we follow the general practice in the literature of
winsorizing the variables at 1 percent and 99 percent.
September 2013
Panel A of Table 3 documents rm performance in the D_NOL years relative to previous or
subsequent years. Using the previous year as the benchmark, we nd that loss rms with carryback
incentives have lower unexpected earnings than loss rms without such incentives, with the mean
and median differences of 11.3 percent (t 3.17) and 0.5 percent (t 1.97), respectively.
Unexpected operating income is also signicantly lower for D_NOL rms, but unexpected special
items are similar to those of loss rms without carryback incentives. Because D_NOL rms are
presumed to accelerate tax losses from future periods, it is also useful to compare the reported
numbers to earnings in the subsequent year. When we do this in the lower half of Panel A, earnings,
operating income, and special items all show signicant differences between loss rms with
TABLE 2
Analysis of Reported Earnings Based on Time-Series Models
(1) (2) (3)
Intercept 0.053 0.005 0.080
(2.64) (0.94) (5.52)
D_NOL 0.073 0.079 0.084
(2.61) (3.06) (3.14)
D_NEG 0.157 0.129 0.114
(7.90) (7.32) (5.94)
E
t1
0.948 0.408 0.594
(5.10) (6.87) (6.15)
E
t2
0.130 0.142 0.308
(2.35) (1.12) (1.55)
LOSS
t1
0.064 0.035
(3.89) (2.35)
LOSS
t2
0.030 0.039
(2.42) (2.75)
E
t1
LOSS
t1
0.555 0.270
(2.74) (1.99)
E
t2
LOSS
t2
0.292 0.549
(1.52) (1.89)
ln(MV
t1
) 0.011
(7.60)
ACC
t1
0.159
(4.16)
RET
t1
0.094
(3.86)
Adj. R
2
0.252 0.258 0.281
This table reports multivariate regression results. The dependent variable is E
t
, the annual Compustat earnings before
extraordinary items per share scaled by stock price at a rms scal year-end. LOSS
t1
(LOSS
t2
) is a dummy variable
with the value of 1 if E
t1
(E
t2
) is negative, and 0 otherwise. NOLC is the net operating loss carryback capacity limit
(see Appendix B for variable measurement). D_NOL is a dummy variable with the value of 1 if NOLC is positive for a
rm in an expected loss position, and 0 otherwise, indicating the rms incentives to carry back tax losses to claim a tax
refund. D_NEG is a dummy variable with the value of 1 if analysts forecasts of year ts earnings are negative, where
forecasts were made eight months prior to scal year-end. MV is a rms market value of equity at the end of year t1.
ACC is total accruals scaled by average assets. RET is the 12-month buy-and-hold return starting from eight months
before a rms scal year-end. Please see Appendix A for detailed variable denitions. The sample includes 85,577 rm-
year observations with non-missing earnings variables (E
t
, E
t1
, and E
t2
) from 1981 to 2010. t-statistics in parentheses
are based on the Fama-MacBeth regression approach. Earnings variables and ACC are winsorized at the 1st and 99th
percentiles.
September 2013
TABLE 3
Analysis of Recurring and Nonrecurring Items
Panel A: Performance Relative to Previous or Subsequent Year
Loss Firms with
Carryback Incentives
(D_NOL 1)
Loss Firms without
Carryback Incentives
(D_NOL 0)
Difference
Mean Median Mean Median
Mean
(t-stat.)
Median
(Z-stat.)
Performance relative to previous year (t1)
Earnings 0.181 0.013 0.068 0.008 0.113*** 0.005**
[(E
t
E
t1
)/P] (3.17) (1.97)
Operating earnings 0.077 0.009 0.016 0.003 0.093*** 0.006***
[(OPINC
t
OPINC
t1
)/P] (5.63) (2.74)
Special items 0.013 0.000 0.051 0.000 0.038 0.00
[(SI
t
SI
t1
)/P] (1.68) (1.82)
Performance relative to subsequent year (t1)
Earnings 0.347 0.046 0.268 0.012 0.080** 0.034***
[(E
t
E
t1
)/P] (2.14) (9.96)
Operating earnings 0.275 0.049 0.162 0.016 0.113*** 0.033***
[(OPINC
t
OPINC
t1
)/P] (3.00) (10.41)
Special items 0.170 0.001 0.072 0.000 0.098*** 0.001***
[(SI
t
SI
t1
)/P] (4.97) (6.88)
Panel B: Logistic Model of Negative One-Time Items in Event Years
Negative Special Items Loss on Sale of PP&E and Investment
Coefcient
(p-value) Marginal Effects
Coefcient
(p-value) Marginal Effects
Intercept 1.847 1.756
(, 0.01) (, 0.01)
D_NOL 0.622 14.23% 0.247 5.01%
(, 0.01) (, 0.01)
D_NEG 0.294 6.65% 0.268 5.41%
(, 0.01) (, 0.01)
E
t1
0.411 9.41% 0.083 1.67%
(, 0.01) (0.011)
E
t2
0.406 9.32% 0.103 2.03%
(, 0.01) (0.009)
LOSS
t1
0.519 11.76% 0.032 0.63%
(, 0.01) (0.338)
LOSS
t2
0.287 6.49% 0.047 1.00%
(, 0.01) (0.145)
ln(MV
t1
) 0.221 5.03% 0.134 2.71%
(, 0.01) (, 0.01)
ACC
t1
0.347 8.00% 0.759 15.55%
(, 0.01) (, 0.01)
RET
t1
0.198 4.53% 0.035 0.71%
(, 0.01) (0.002)
(continued on next page)
September 2013
carryback incentives and loss rms without. In general, the magnitude of the loss is larger when we
use the subsequent year rather than the previous year as the benchmark, especially with medians.
This result indicates that the earnings changes are transitory, and is consistent with D_NOL rms
shifting expenses and losses from the subsequent year to the current year. Finally, the magnitude of
the difference in operating income is similar to that of the earnings difference, especially for
medians, and this is driven by the infrequency and variability of special items.
In Table 3, Panel B, we further examine whether rms with tax incentives are more likely to
report negative one-time items. We focus on negative special items and the loss on the sale of
PP&E and investment under the premise that rms can sell assets at a loss to recapture taxes paid in
previous years. Both logistic models show signicant coefcients on D_NOL, and the marginal
effects suggest that rms with carryback incentives are 14.23 percent more likely to report a
negative special item and 5.01 percent more likely to report a loss from an asset sale. As D_NOL is
implicitly an interaction between D_NEG and an indicator for carryback potential, these effects are
incremental to what we observe for loss rms without carryback incentives (D_NEG 1 and
D_NOL 0). Finally, the coefcients on control variables are largely signicant with expected
signs. Protable rms and rms with high past returns are less likely to report negative one-time
items, whereas past loss rms and rms with high accruals tend to report negative one-time items.
In sum, we nd strong evidence that rms increase losses in order to claim tax refunds. Both
core earnings and one-time items are signicantly lower in D_NOL years than in adjacent years,
suggesting that rms use both recurring and non-recurring items to shift losses.
Do Analysts Incorporate Tax-Motivated Loss Shifting in Their Earnings Forecasts?
If analysts anticipate tax-motivated loss shifting and incorporate these incentives into their
forecasts, then we expect to nd no difference in analyst forecast errorsactual earnings less the
analysts forecast of earningsbetween rms with tax incentives to accelerate losses (D_NOL1)
and those without (D_NOL0). But if analysts ignore or do not fully incorporate information about
taxes, then their forecast errors should be more negative for rms with tax incentives to accelerate
losses. We conduct both univariate and multivariate analyses on analysts forecast errors. In
univariate analyses, Panel A of Table 4 shows that the average analyst forecast error for rms with a
tax-motivated loss-shifting incentive (D_NOL rms), scaled by stock price, is 9.5 percent (p ,
0.01). For all rms without this carryback incentive, including rms with expected positive earnings
TABLE 3 (continued)
**, *** Indicate signicant differences at the 5 percent and 1 percent levels, respectively.
Panel A reports nancial performance relative to previous and subsequent years, respectively. Loss rms are dened as
rms with negative analyst earnings forecasts made eight months prior to scal year-end. Earnings are earnings before
extraordinary items. Operating income (OPINC) is operating income after depreciation and amortization. P is the market
value of equity. There are 2,210 and 8,068 rm-year observations for loss rms with and without carryback incentives,
respectively. Panel B reports the results of estimating a logistic model for negative one-time items. The data on gain/loss
on sale of PP&E and investment start in 1995. D_NEG is a dummy variable with the value of 1 if analysts forecasts of
year ts earnings are negative, where forecasts were made eight months prior to scal year-end. E is annual Compustat
earnings before extraordinary items per share scaled by stock price at a rms scal year-end. LOSS
t1
(LOSS
t2
) is a
dummy variable with the value of 1 if E
t1
(E
t2
) is negative, and 0 otherwise. NOLC is the net operating loss carryback
capacity limit (see Appendix B for variable measurement). D_NOL is a dummy variable with the value of 1 if NOLC is
positive for a rm in an expected loss position, and 0 otherwise, indicating the rms tax incentive to accelerate losses.
MV is a rms market value of equity at the end of year t1. ACC is total accruals scaled by average assets. RET is the
12-month buy-and-hold return starting from eight months before a rms scal year-end. The sample includes 85,577
rm-year observations with non-missing earnings variables from 1981 to 2010.
September 2013
TABLE 4
Analyst Forecast Error Properties
Panel A: Analyst Forecast Errors for Firms with and without Tax-Motivated Loss-Shifting
Incentives
Loss Firms with
Tax Incentives
(D_NOL 1)
Other Firms
(D_NOL 0)
Difference
(t-stat.)
Loss Firms without
Tax Incentives
(D_NEG 1, D_NOL 0)
Difference
(t-stat.)
A B A B C A C
Entire Sample
Forecast error 0.095 0.029 0.066
(27.85)
0.042 0.053
(9.91)
19871990
Forecast error 0.204 0.055 0.149
(10.37)
0.143 0.071
(3.24)
Panel B: Multivariate Regressions of Analyst Forecast Errors (FE
t
)
(1) (2)
Intercept 0.078 0.048
(10.82) (11.06)
FE
t1
0.426
(10.88)
D_NOL 0.048 0.027
(6.25) (3.81)
D_NEG 0.015 0.004
(1.86) (0.72)
ln(MV
t1
) 0.009 0.005
(8.92) (7.60)
ln(COV
t
) 0.004 0.002
(2.86) (1.87)
BM
t1
0.023 0.018
(5.42) (4.38)
RET
t1
0.035 0.027
(6.23) (5.45)
ACC
t1
0.006 0.058
(0.87) (8.96)
Adj. R
2
0.112 0.180
Forecast error, FE, is calculated as the difference between I/B/E/S actual earnings and the median analyst forecast made
eight months prior to a rms scal year-end, scaled by stock price at the forecast date. NOLC is the net operating loss
carryback capacity limit (see Appendix B for variable measurement). D_NOL is a dummy variable with the value of 1 if
NOLC is positive for a rm in an expected loss position, and 0 otherwise, indicating the rms tax incentive to accelerate
losses. D_NEG is a dummy variable with the value of 1 if analysts forecasts of year ts earnings are negative, where
forecasts were made eight months prior to scal year-end. MV is a rms market value of equity at the end of year t1.
COV is the number of analysts covering the rm at the forecast date. BM is the book-to-market ratio at the end of year
t1. RET is the 12-month buy-and-hold return starting from eight months before a rms scal year-end. ACC is total
accruals scaled by average assets. Please see Appendix A for detailed variable denitions. The sample includes 83,875
rm-year observations with non-missing FE from 1981 to 2010. In all columns, t-statistics in parentheses are based on
the Fama-MacBeth approach. FE, BM, and ACC are winsorized at the 1st and 99th percentiles.
September 2013
surprises, analyst forecast errors are lower, averaging 2.9 percent.
16
The resulting difference in
analyst forecast errors of 6.6 percent of stock price (p , 0.01) is highly signicant. We also
compare the forecast errors in D_NOL rms to expected loss rms without carryback incentives and
nd that D_NOL rms have signicantly greater negative forecast errors, differing by 5.3 percent of
stock price (p , 0.01) between the two groups of rms. Overall, these results are consistent with
the conclusion that analysts do not fully incorporate tax-motivated loss-shifting behavior into their
earnings forecasts.
To control for other factors that could be correlated with both the tax incentives to shift losses
and the forecast error, we estimate the following equation:
FE
t
b
0
b
1
FE
t1
b
2
D NOL b
3
D NEG b
4
LogMV
t1
b
5
LogCOV
t
b
6
BM
t1
b
7
ACC
t1
b
8
RET
t1
e
t
; 2
where FE is analyst forecast error. If rms increase losses in year t to capture tax refunds and
analysts do not anticipate this behavior, then we expect a negative coefcient on D_NOL, where
again the coefcient on D_NOL represents the effect that is incremental to the association between
expected losses and realized forecast errors estimated by the coefcient on D_NEG. Following
previous research, we control for other determinants of earnings forecast bias. The literature argues
that analysts have incentives to issue optimistic forecasts for rms with poor information
environments to obtain access to management (Francis and Philbrick 1993; Lim 2001). We include
rm size (MV) and analyst coverage (COV) to capture a rms information environment (Atiase
1987; Zhang 2006b). We also include the book-to-market ratio (BM) to control for growth
opportunities, as growth rms have stronger incentives to meet or beat earnings targets, suggesting
a negative coefcient on BM. Prior literature documents evidence indicating that analysts are not
fully rational (see Kothari [2001] for a review). To allow for this possibility, we include RET and
FE
t1
to control for analyst underreaction to the information contained in stock returns and forecast
errors in the prior period. We further include accruals to control for the possibility that analysts do
not fully understand the implication of accruals for future earnings (Bradshaw et al. 2001; Teoh and
Wong 2002).
Table 4, Panel B summarizes the results of the forecast error regressions with and without the
lagged dependent variable as a control. In column (1), the coefcient on D_NEG is only marginally
signicant (coeff. 0.015, t 1.86), providing weak evidence that analysts are surprised by
negative earnings news even when they expect the rm to report a loss. Turning to our variable of
interest, the coefcient on D_NOL is 0.048 (t 6.25), which is much larger than the coefcient
on D_NEG, and indicates that analysts fail to anticipate the additional losses reported by rms with
tax incentives to shift income and that analysts forecast errors are predictable ex ante. We add the
lagged forecast error as a control in column (2) to control for unobservable rm effects, and nd
that the coefcient on D_NOL drops to 0.027 but remains highly signicant (t 3.81).
Economically, analysts forecast errors, scaled by stock price, are about 2.7 percentage points larger
for loss rms with carryback incentives than for rms without. This translates into about $64.7
billion of unexpected loss shifting for rms with incentives to use those losses to obtain cash tax
refunds between 1981 and 2010.
17
16
The observed negative forecast errors for D_NOL 0 rms is consistent with the optimistic bias in analysts
earnings forecasts documented in the prior literature (OBrien 1987). Analysts forecast errors are reliably
negative when the forecast horizon is long, and turn to zero or positive immediately before earnings
announcements (Richardson et al. 2004).
17
We estimate the $64.7 billion as the 2.7 percent multiplied by the average sample rms market value of $1.204
billion and then multiply that by 1,990 rm-year observations with tax incentives in Model (2).
September 2013
Equity Market Response to Tax-Motivated Loss Shifting
Finally, we examine the reaction of equity investors to tax-motivated loss shifting. If investors
naively follow analysts earnings forecasts, which tend to miss the tax incentives for accelerating
losses, then the market price will not fully reect the expected tax benets and the market reaction
will be determined by analysts forecast errors. Alternatively, if investor reactions are determined
by the revision in expected future cash ows and discount rates, then the positive cash ow effects
associated with accelerating a tax loss should lead to more positive (or less negative) market
reactions for rms with the tax-based incentives to shift losses, independent of the size of the
forecast error.
We consider both annual returns and earnings announcement returns and report the results in
Table 5. The rst two columns are based on annual returns, and we estimate them with and without
the analyst forecast error as a control. In the full model, we control for analyst forecast errors, rms
with expected losses at the beginning of the period using D_NEG, as well as size, book-to-market,
and prior returns. Column (1) shows that the partial correlation between D_NOL and annual returns
is positive, albeit insignicant, suggesting that the market reaction to negative earnings
announcements by D_NOL rms is similar to those by non-D_NOL rms, even though D_NOL
rms report much larger losses as shown in Tables 2 and 4. Since D_NOL is negatively correlated
with the forecast error, we add analyst forecast errors as a control variable in column (2), and nd a
positive and signicant coefcient on D_NOL (coeff. 0.074, t 2.17), implying that the annual
stock return of a rm with tax-motivated loss-shifting incentives is about 7.4 percent greater than a
loss rm without such incentives, holding analyst forecast errors constant.
In columns (3) and (4), we use abnormal return during the earnings announcement window as
the dependent variable. This allows us to focus on the importance of the earnings release in
conveying information about cash ow effects of the tax loss. Similar to the annual return
regressions, we nd a positive yet insignicant coefcient on D_NOL without controlling for
analyst forecast errors, and this turns positive and signicant once we include analyst forecast errors
(coeff. 0.006, t 3.23). Overall, the results suggest that investors understand either the tax
benets of the loss carryback, the transitory nature of the accelerated losses claimed by carryback
rms, or both.
In summary, we nd that the market reaction to earnings reported by rms with incentives to
accelerate tax losses is less negative than the market reaction to earnings by rms without such
incentives. This occurs despite the fact that rms with tax incentives to shift losses actually report
larger losses using both prior earnings and analyst forecasts as a benchmark, and is consistent with
these losses being transitory, increasing cash ows and providing liquidity.
18
V. ADDITIONAL ANALYSIS AND SENSITIVITY CHECKS
Cross-Sectional Variation in Tax-Motivated Loss Shifting
In Table 4, we provide some evidence of the time-series variations in rms incentives to
engage in tax-motivated loss shiftingrms have particularly strong tax incentives to shift losses
around TRA 86. As Maydew (1997) shows, TRA 86 reduced the corporate tax rate from 46 percent
in 1986 to 34 percent in 1988, providing rms with stronger incentives to accelerate losses between
18
Our market reaction results are in contrast to those in Shane and Stock (2006), who show that investors do not
understand income shifting around one-time tax rate change around 1986. We attribute the different results
between these two research settings to the recurring nature of loss carrybacks and better rm communication
with the market in the 1990s and 2000s. Management guidance was virtually non-existent in the 1980s.
September 2013
1987 and 1990 in order to generate a refund of taxes paid at the higher rate. In this section, we
explore the cross-sectional variations in rms incentives to carry back tax losses.
While rms have a variety of incentives to carry back losses, we specically test whether the
demand for liquidity drives the rms decision to increase cash ow through a tax refund. This
implicitly assumes that the incremental cost of external funds is higher than the cost of funds
obtained through a tax refund. Empirically, we measure nancial constraints (FC) using a rms
debt minus cash and short-term investment scaled by total assets. We then add FC, measured at
prior year-end, and its interaction term with D_NOL (D_NOL FC) to Equations (1) and (2) as
follows:
E
t
b
0
b
1
D NOL b
2
FC
t1
b
3
D NOL FC
t1
b
4
D NEG b
5
D NEG FC
t1
b
6
E
t1
b
7
E
t2
b
8
LOSS
t1
b
9
LOSS
t2
b
10
LOSS
t1
E
t1
b
11
LOSS
t2
E
t2
b
12
LogMV
t1
b
13
ACC
t1
b
14
RET
t1
e
t
: 3
TABLE 5
Analysis of Stock Returns
Dependent Variable
RET
t
(1)
RET
t
(2)
ARET
t
(3)
ARET
t
(4)
Intercept 0.153 0.277 0.0026 0.0055
(2.07) (3.68) (1.67) (3.54)
FE
t
1.554 0.1109
(10.72) (8.73)
D_NOL 0.003 0.074 0.0014 0.0058
(0.09) (2.17) (0.71) (3.23)
D_NEG 0.018 0.004 0.0086 0.0070
(0.39) (0.09) (6.61) (5.09)
ln(MV
t1
) 0.007 0.023 0.0001 0.0008
(0.99) (3.04) (1.22) (4.10)
BM
t1
0.032 0.063 0.0031 0.0050
(1.60) (3.19) (4.52) (6.67)
RET
t1
0.016 0.064 0.001 0.0029
(0.47) (1.95) (0.24) (6.43)
Adj. R
2
0.043 0.113 0.008 0.045
RET
t
is year ts annual returns starting from eight months prior to scal year-end. ARET
t
is the average three-day earnings
announcement return in year t, measured as raw returns minus value-weighted market returns over the three-day [1, 1]
period, where day 0 is the earnings announcement date. FE
t
is the forecast error for year t, calculated as the difference
between I/B/E/S actual earnings and the median analyst forecast made eight months prior to a rms scal year-end,
scaled by stock price at the forecast date. NOLC is the net operating loss carryback capacity limit (see Appendix B for
variable measurement). D_NOL is a dummy variable with the value of 1 if NOLC is positive for a rm in an expected
loss position, and 0 otherwise, indicating the rms incentives to carry back NOLs and to claim tax refunds. D_NEG is a
dummy variable with the value of 1 if analysts forecasts of year ts earnings are negative, where forecasts were made
eight months prior to scal year-end. MV is a rms market value of equity at the end of year t1. BM is the book-to-
market ratio at the end of year t1. RET
t
is the 12-month buy-and-hold return starting from eight months before a rms
scal year-end. Please see Appendix A for detailed variable denitions. The sample includes 83,875 rm-quarter
observations with non-missing observations of FE and return variables from 1981 to 2010. In all columns, t-statistics in
parentheses are based on the Fama-MacBeth regression approach. FE and BM are winsorized at the 1st and 99th
percentiles.
September 2013
FE
t
b
0
b
1
D NOL b
2
FC
t1
b
3
D NOL FC
t1
b
4
D NEG b
5
D NEG FC
t1
b
6
LogMV
t1
b
7
LogCOV
t
b
8
BM
t1
b
9
ACC
t1
b
10
RET
t1
e
t
: 4
Our main prediction is that the coefcients on the interaction term D_NOL FC should be
signicantly negative in Equations (3) and (4), suggesting that more nancially constrained rms
have stronger loss-shifting incentives. In Table 6, we nd this is indeed the case. In the regression
of future earnings, the coefcient on the interaction term D_NOL FC is 0.105 (t 1.97). In the
regression of analyst forecast errors, the coefcient on the interaction term D_NOL FC is 0.028
(t 2.24). If we further add the prior-period forecast errors as an additional control variable to
Equation (4), then the interaction term D_NOL FC becomes marginally signicant (coeff.
0.018, t 1.65). The results provide some support for the role of liquidity benets because
nancially constrained rms appear more likely to accelerate losses to access cash from tax refunds,
and this behavior results in more negative earnings and analyst forecast errors for those rms.
Analyst Revisions of Future Earnings
Section IV shows that analysts do not fully incorporate tax-related losses in their earnings
forecasts. It is possible that some analysts may learn about such tax-motivated loss shifting over
time, for example when the rm announces earnings or an asset sale, and revise their earnings
forecasts accordingly. If tax-related losses represent one-time attempts to shift income, then a
testable implication is that current-year earnings forecasts should fall, but future earnings forecasts
should be relatively unchanged, suggesting a smaller correlation in analysts forecast revisions
between the current and subsequent years earnings. To test this prediction, we analyze the
following model:
REV
t1
b
0
b
1
D NOL b
2
D NEG b
3
REV
t
b
4
REV
t
D NOL b
5
REV
t
D NEG
b
6
LogMV
t1
b
7
LogCOV
t
b
8
BM
t1
b
9
RET
t1
b
10
ACC
t1
e
t1
;
5
where REV
t
measures analyst forecast revision for year ts earnings from eight months before year
ts scal year-end to one month after, and REV
t1
measures analyst forecast revision for year t1s
earnings during the same period. We expect that on average b
3
will be positive (revisions are
positively correlated), b
5
will be negative (less correlated for negative earnings shocks), and b
4
will
be negative (revisions of t1 forecasts are even less sensitive to current revisions given tax
incentives to report losses).
The results are consistent with our predictions. Untabulated results show a positive coefcient
on REV
t
(b
3
0.558, t 22.50) and a negative coefcient on REV
t
D_NEG (b
5
0.114, t
4.44). More importantly, the coefcient on REV
t
D_NOL is signicantly negative (b
4
0.068,
t 2.05). The results suggest that revisions of future years earnings are less sensitive to revisions
in current-year earnings for rms with tax loss carryback incentives, consistent with the idea that
tax-related losses reect one-time attempts to shift income. In other words, to the extent that
analysts revise their current-year forecasts downward to reect better information about the rms
tax-based incentive to accelerate losses into that year, they are less likely to revise forecasts of
future earnings in the same direction.
Other Robustness Checks
Our main analysis is based on an indicator variable version of NOLC, the capacity for tax loss
carrybacks. In untabulated analysis, we incorporate the magnitude of tax loss carrybacks by
replacing D_NOL with NOLC and re-estimate the regressions. The results continue to hold. For
example, the coefcient on NOLC in Equation (2) is 0.795 (t 3.54), while the coefcient on
September 2013
TABLE 6
Analysis of Financial Constraints
Panel A: Regression of Earnings (E
t
)
Intercept 0.118
(9.46)
D_NOL 0.060
(2.48)
D_NEG 0.133
(4.47)
FC
t1
0.055
(5.52)
D_NOL FC
t1
0.105
(1.97)
D_NEG FC
t1
0.241
(3.65)
E
t1
0.509
(7.79)
E
t2
0.038
(0.45)
LOSS
t1
0.029
(2.49)
LOSS
t2
0.029
(3.23)
E
t1
LOSS
t1
0.174
(2.43)
E
t2
LOSS
t2
0.133
(1.20)
ln(MV
t1
) 0.016
(10.77)
ACC
t1
0.108
(4.58)
RET
t1
0.079
(5.94)
Adj. R
2
0.259
Panel B: Regression of Analyst Forecast Errors (FE
t
)
Intercept 0.082
(14.21)
D_NOL 0.030
(4.86)
D_NEG 0.014
(2.02)
FC
t1
0.021
(6.40)
D_NOL FC
t1
0.028
(2.24)
D_NEG FC
t1
0.062
(2.85)
(continued on next page)
September 2013
NOLC in Equation (3) is 0.151 (t 2.29). One caveat of the NOLC regressions is that actual
losses may not be linearly correlated with loss carryback capacity.
We use annual observations in the main analysis on the premise that tax reporting is done
annually. As a robustness check, we rerun our analysis based on quarterly observations by
assigning D_NOL in year t1 to four quarterly earnings surprises in year t. A caveat of the quarterly
approach is that four quarterly observations within a rm-year are not independent. As an
alternative, we execute our analyses using only the fourth-quarter observations. In both
specications, the tenor of the results is unchanged (e.g., signicantly negative coefcients on
D_NOL in the regressions of earnings and analyst forecast errors, and positive coefcients on
D_NOL in the return regressions). We also examine how the coefcient pattern varies across
quarters and nd that our results are driven by the third and fourth scal quarters, consistent with
the idea that rms are more likely to shift losses in later quarters as they receive more precise
information about expected tax and accounting prots, loss-shifting opportunities, and liquidity
needs.
VI. CONCLUSION
We identify rms with an incentive to exercise an option to obtain a cash refund of prior tax
payments that would otherwise expire at the end of the year. Over the 19812010 period with
relatively stable statutory tax rates, rms appear to accelerate reported losses to generate cash ows
from tax refunds, and do so to a greater extent when they are nancially constrained. This evidence
suggests that such tax-based incentives have pervasive effects on reporting decisions across time
even when statutory tax rates are stable and thus have broad implications for capital markets
research.
TABLE 6 (continued)
ln(MV
t1
) 0.010
(11.18)
ln(COV
t
) 0.003
(2.70)
BM
t1
0.018
(6.29)
RET
t1
0.029
(7.27)
ACC
t1
0.013
(2.42)
Adj. R
2
0.128
Panels A and B report regression results for earnings (E
t
) and analyst forecast errors (FE
t
), respectively. E is annual
Compustat earnings before extraordinary items scaled by stock price at a rms scal year-end. FE is the difference
between I/B/E/S actual earnings and the median analyst forecast made eight months prior to a rms scal year-end,
scaled by stock price at the forecast date. D_NOL is a dummy variable with the value of 1 if a rm has carryback
incentives, and 0 otherwise. D_NEG is a dummy variable with the value of 1 if analysts forecasts of year ts earnings are
negative, where forecasts were made in the fourth month after scal year-end. FC is nancial constraint measured as a
rms debt minus cash and short-term investment scaled by total assets. LOSS
t1
(LOSS
t2
) is a dummy variable with the
value of 1 if E
t1
(E
t2
) is negative, and 0 otherwise. MV is a rms market value of equity at the end of year t1. ACC is
total accruals scaled by average assets. RET is the 12-month buy-and-hold return starting from eight months before a
rms scal year-end. COV is the number of analysts covering the rm at the forecast date. Please see Appendix A for
detailed variable denitions. The sample includes 85,577 rm-year observations with non-missing earnings variables (E
t
,
E
t1
, and E
t2
) in Panel A and 83,875 rm-year observations with non-missing FE in Panel B from 1981 to 2010.
Earnings variables FE, BM, and ACC are winsorized at the 1st and 99th percentiles, and t-statistics in parentheses are
based on the Fama-MacBeth regression approach.
September 2013
Our evidence also suggests that analysts do not incorporate these tax-motivated loss-shifting
incentives into their earnings forecasts because we nd that analyst forecast errors are signicantly
more negative for rms with incentives to accelerate tax losses. This evidence contrasts with the
view that analyst forecast errors proxy for earnings surprises, and reinforces the evidence on
analysts disregard for the information in the rms tax disclosures. Despite the relatively larger
unexpected losses for rms with tax incentives to accelerate losses, the stock market reacts in a way
that suggests investors value tax-motivated loss shifting. If we control for the size of the reported
loss, then the stock returns of rms with tax-motivated loss-shifting incentives are signicantly less
negative than similar loss rms without such incentives.
Our evidence is relevant to understanding the growing importance of tax losses on rm value.
Recent innovations in takeover defenses are designed specically to protect the value of tax losses,
and within the last decade, Congress has argued that the temporary liberalization of tax loss
carryback rules injects needed liquidity into the business sector. Taken together, our evidence
suggests that the incentive to carry back tax losses plays a material and persistent role in corporate
reporting decisions and capital markets activities.
REFERENCES
Amir, E., and T. Sougiannis. 1999. Analysts interpretation and investors valuation of tax carryforwards.
Contemporary Accounting Research 16: 133.
Atiase, R. 1987. Market implications of predisclosure information: Size and exchange effects. Journal of
Accounting Research 25: 168176.
Blouin, J., J. Core, and W. Guay. 2010. Have the tax benets of debt been overestimated? Journal of
Financial Economics 98 (2): 195213.
Bradshaw, M., S. Richardson, and R. Sloan. 2001. Do analysts and auditors use information in accruals?
Journal of Accounting Research 39: 4574.
Brown, L., R. Hagerman, P. Grifn, and M. Zmijewski. 1986. Security analyst superiority relative to
univariate time-series models in forecasting quarterly earnings. Journal of Accounting and
Economics 9: 6187.
Chen, K., and M. Schoderbek. 2000. The 1993 tax rate increase and deferred tax adjustments: A test of
functional xation. Journal of Accounting Research 39: 2344.
Dhaliwal, D., C. Gleason, and L. Mills. 2004. Last-chance earnings management: Using the tax expense to
achieve analysts forecasts. Contemporary Accounting Research 21: 431459.
Doyle, J., R. Lundholm, and M. Soliman. 2006. The extreme future stock returns following extreme
earnings surprises. Journal of Accounting Research 44: 849887.
Engel, E., M. Erickson, and E. Maydew. 1999. Debt-equity hybrid securities. Journal of Accounting
Research 37: 249274.
Erickson, M., and S. Heitzman. 2010. NOL poison pills: Selectica v. Versata et al. Tax Notes (June 23).
Francis, J., and D. Philbrick. 1993. Analysts decisions as products of a multi-task environment. Journal of
Graham, J. 1996. Proxies for the corporate marginal tax rate. Journal of Financial Economics 42: 187221.
Graham, J., and C. Smith. 1999. Tax incentives to hedge. Journal of Finance 54: 22412262.
Graham, J., and H. Kim. 2009. The effects of the length of the tax-loss carryback period on tax receipts and
corporate marginal tax rates. National Tax Journal 52: 413427.
Guenther, D. 1994. Earnings management in response to corporate tax rate changes: Evidence from the
1986 Tax Reform Act. The Accounting Review 69: 230243.
Hanlon, H., and S. Heitzman. 2010. A review of tax research. Journal of Accounting and Economics 50:
127178.
Hanlon, M. 2003. What can we infer about a rms taxable income from its nancial statements? National
Tax Journal 56: 831863.
September 2013
Hayn, C. 1995. The information content of losses. Journal of Accounting and Economics 20: 125153.
Kothari, S. P. 2001. Capital markets research in accounting. Journal of Accounting and Economics 31: 105
231.
Leone, M. 2010. Your loss is your gain. Available at: http://www.cfo.com/article.cfm/14476845/c_
14477052
Lim, T. 2001. Rationality and analysts forecast bias. The Journal of Finance 56: 369385.
Matsunaga, S., T. Shevlin, and D. Shores. 1992. Disqualifying dispositions of incentive stock options: Tax
benets versus nancial report costs. Journal of Accounting Research 30 (Supplement): 3768.
Maydew, E. 1997. Tax-induced earnings management by rms with net operating losses. Journal of
Mills, L., K. Newberry, and G. Novack. 2003. How well do Compustat NOL data identify rms with U.S.
tax return loss carryovers? The Journal of the American Taxation Association 25: 117.
OBrien, P. 1987. Analysts forecasts as earnings expectations. Journal of Accounting and Economics 10
(1): 5383.
Outslay, E., and G. McGill. 2002. Did Enron pay taxes? Using accounting information to decipher tax
status. Tax Notes (August 19).
Plumlee, M. 2003. The effect of information complexity on analysts use of that information. The
Accounting Review 78: 275296.
Richardson, S., S. Teoh, and P. Wysocki. 2004. The walkdown to beatable analyst forecasts: The roles of
equity issuance and insider trading incentives. Contemporary Accounting Research 21: 885924.
Scholes, M., M. Wolfson, M. Erickson, E. Maydew, and T. Shevlin. 2008. Taxes and Business Strategy. 4th
edition. Upper Saddle River, NJ: Prentice Hall.
Scholes, M., P. Wilson, and M. Wolfson. 1992. Firms responses to anticipated reductions in tax rates: The
Tax Reform Act of 1986. Journal of Accounting Research 30 (Supplement): 161185.
Shackelford, D., J. Slemrod, and J. Sallee. 2010. Financial Reporting, Tax, and Real Decisions: Toward a
Unifying Framework. Working paper, University of Michigan.
Shane, P., and T. Stock. 2006. Security analyst and stock market efciency in anticipating tax-motivated
income shifting. The Accounting Review 81: 227250.
Shevlin, T. 1990. Estimating corporate marginal tax rates with asymmetric treatment of gains and losses.
Journal of the American Taxation Association 11: 5167.
Sloan, R. 1996. Do stock prices reect information in accruals and cash ows about future earnings? The
Accounting Review 71: 289315.
Teoh, S. H., and T. J. Wong. 2002. Why new issues and high-accrual rms underperform: The role of
analysts credulity. Review of Financial Studies 15: 869900.
Zhang, F. X. 2006a. Information uncertainty and stock returns. Journal of Finance 61: 105137.
Zhang, F. X. 2006b. Information uncertainty and analyst forecast behavior. Contemporary Accounting
Research 23: 565590.
September 2013
APPENDIX A
Variable Denitions
E
t
E is annual Compustat earnings before extraordinary items per share scaled by stock price
at a rms scal year-end;
FE
t
analyst forecast error, dened as (E F)/P
t1
, where E is reported annual earnings per
share from I/B/E/S, F is the median analysts earnings-per-share forecast made eight
months prior to scal year-end, and P
t1
is stock price at the forecast date;
NOLC
t1
net operating loss carryback capacity (see the estimation process in Appendix B for
details), indicating a rms tax paid in the earliest year of the NOL carryback window
that has not been refunded yet;
D_NEG a dummy variable with the value of 1 if analysts forecasts of year ts earnings are
negative, where forecasts were made eight months prior to scal year-end;
D_NOL a dummy variable with the value of 1 if the rm has an expected loss (D_NEG 1) and
the net operating loss carryback capacity is positive (NOLC . 0), and 0 otherwise;
MV
t1
the market value of equity at the end of year t1;
COV
t
the number of analysts following the company at the forecast date (eight months prior to
scal year-end);
BM
t1
the book-to-market ratio, measured as the book value of equity divided by its market
value, at the end of year t1;
RET
t
the 12-month buy-and-hold return starting from eight months before a rms scal year-
end;
ARET
t
average earnings announcement return of four quarterly earnings announcements for year t.
Quarterly earnings announcement return is dened as raw returns minus value-weighted
market returns over the three-day [1, 1] period, where day 0 is the earnings
announcement date;
ACC
t1
total accruals in year t1, measured as (DCA DCash) (DCL DSTD DTP)
DEPEXP scaled by average total assets, where DCA change in current assets, DCash
change in cash and cash equivalents, DCL change in current liabilities, DSTD
change in debt in current liabilities, and DEPEXP depreciation and amortization
expense;
FC
t1
nancial constraint, measured as total debt minus cash and short-term investment scaled by
total assets, at the end of year t1;
LEV
t1
the leverage ratio, measured as total debt divided by total assets, at the end of year t1;
and
EBITDA
t1
the ratio of earnings before interest, taxes, depreciation, and amortization scaled by average
total assets, measured in year t1.
September 2013
APPENDIX B
Variable Measurement
Timeline of the Variables (using December Fiscal Year-End as an Example)
Estimating Net Operating Loss Carryback Capacity (NOLC)
NOLC is meant to capture tax paid in the earliest year of the NOL carryback window that has
not been refunded by the beginning of year t. We estimate net operating loss carryback capacity
(NOLC) depending on the time period of the loss. Before 1997, rms could carry back NOLs up to
three years and carry forward NOLs up to 15 years. Since 1997, rms can carry back NOLs up to
two years and carry forward NOLs up to 20 years.
The Post-1997 Period (Fiscal Year-End of August 1998 and Later)
Firms can carry back NOLs up to two years:
Our main test is whether rms have incentives to report larger losses in year t based on whether
the rm can carry back losses against income in the earliest carryback year (t2). We dene taxable
income (TI) as current tax expense divided by the top statutory tax rate. We calculate NOLC as
follows:
NOLC MAX 0; TI
t2
MIN0; TI
t1
MAX0; TI
t3
f g: A1
In words, A1 denes the capacity to carry back losses (NOLC) in year t as the amount of taxable
prots in t2 that has not been used to claim a refund from tax losses in year t1. If the rm did not
generate a taxable prot in t2 (TI
t2
0), then there is no tax available to refund, and the outside
maximization function ensures that NOLC will simply equal 0. But if the rm has taxable prots in
t2 (TI
t2
. 0), then we must rst determine how much of those taxable prots in t2 were offset
through tax losses in year t1. This adjustment is given by MIN[0,TI
t1
MAX(0,TI
t3
)] in
Equation (A1). If the rm has a taxable prot in year t1, then this adjustment equals 0 and the full
amount of t2 taxes paid is available for refund. If the rm had taxable losses in year t1 (TI
t1
,
0), then we must ask whether the rm had taxable income in t3. A loss in year t1 will rst be
carried back against prots in year t3, and the prots available to be offset by year t1 losses are
presented by MAX(0,TI
t3
). If the year t1 loss exceeds the year t3 income (if any), then the
September 2013
excess loss is taken against t2 income, and this is the adjustment given by MIN[0,TI
t1

MAX(0,TI
t3
)].
To illustrate the potential values of NOLC as a function of prior taxable prot and loss patterns,
consider the cases in the following table:
Case
Taxable
Prots/Losses
in Earliest
Carryback Year
TI
t2
Taxable
Prot/Loss
in Year t1
TI
t1
Taxable Prots
in Year t3
Available to Offset
Year t1 Losses
MAX(0,TI
t3
)
Amount of Year
t1 Loss Not Covered
by Prots in t3
MIN[0,TI
t1
MAX(0,TI
t3
)]
Max. Amount
of Loss in Year
t that can be
Carried Back
to Obtain Refund
of t2 Taxes
NOLC
(1) (2) (3) (4) (5)
1. 10 0 20 0 10
2. 5 5 10 0 5
3. 10 10 10 0 10
4. 10 20 10 10 0
5. 10 15 10 5 5
6. 10 0 0 0 0
Cases 1 through 5 represent rms with taxes paid on prots in the earliest carryback year. Cases 1
and 2 represent rms with no tax loss in t1 (and hence no offset of t2 prots). Case 3 provides an
example of a rm whose tax loss in year t1 was offset entirely against prots in year t3,
preserving the entire capacity of t2 prots to offset losses in t. Case 4 represents the case where a
loss in t consumed both t3 and t2 prots, while in Case 5, only a portion of the prots in t2 are
used up to offset losses in t1. The remainder in Case 5 is the amount available for use against
losses in t. The nal case simply illustrates that rms that paid no taxes in t2 (or reported a loss)
will have no capacity for carryback of t losses regardless of what happens in year t1 or t3.
The Pre-1997 Period
Firms can carry back NOLs up to three years.
Here, we calculate NOLC as follows:
NOLC MAX
0; TI
t3
MIN
h
0; TI
t2
MAX0; TI
t5
MAX0; TI
t4
i
MIN
h
0; TI
t1
MAX
0; TI
t4
MIN
0; TI
t2
MAX0; TI
t5
: A2
Due to the dynamics involved with three carryback years, the formula is more complex. But the
intuition and results are similar as in A1. Numerical examples using this formula are available from
the authors upon request, but can be easily replicated.
September 2013
D_NOL
D_NOL is a dummy variable that takes a value of 1 if NOLC is positive for a rm in an
expected loss position, and 0 otherwise. Firms with expected losses are those with negative
analysts consensus (median) forecasts made eight months prior to scal year-end (D_NEG 1).
Our NOLC (and hence D_NOL) measure is subject to a potential sources of measurement error.
First, taxable income includes both operating income/loss and capital gain/loss, which have
different carryback and carryforward rules. We do not make such a distinction because the
carryback periods are largely similar and because we can only estimate taxable income. Second,
measurement errors arise because the average tax rate is not equal to the top statutory tax rate owing
to progressive tax schedules. Third, current tax expense does not reect the tax benets from the
exercise of non-qualied employee stock options (reected as a reduction in current tax liability)
and is reported net of tax cushions and tax credits (Hanlon 2003). Fourth, we do not incorporate the
magnitude of expected losses related to previous years prots, as analysts earnings forecasts are a
noisy proxy for a rms tax prots/losses. Incorporating the magnitude of expected losses would
make our NOLC formula too complicated. Fifth, rms were allowed to carryback NOLs for ve
years in 2001/2002 and 2008/2009. We do not incorporate the ve-year carry backs in our measure,
as the ve-year carryback period was often not known until a later date. In addition, a ve-year
carryback would make our measure extremely complicated. Finally, we do not incorporate the
information of net operating loss carryforwards on the balance sheet because of jurisdictional
differences and the impact of NOLs acquired in mergers and acquisitions. Moreover, research by
Mills et al. (2003) raises a number of additional concerns using Compustat data to infer net
operating loss positions. We hand-checked a number of cases and nd that many rms were able to
claim tax refunds when reporting non-missing NOL carryforwards in their nancial statements.
While we can certainly rene our NOLC measure along the directions mentioned above,
incorporating such issues would make our measure unduly complicated. We believe that such
measurement errors introduce noise to our measure and, if anything, are likely to bias against
nding any signicant results. With that in mind, we prefer a relatively simple measure as used in
the paper.
Alternative Specications of NOLC and D_NOL
We consider the following alternative specications of NOLC and D_NOL. In the rst
alternative specication, we replace expected losses proxied by analysts forecasts with those based
on a time-series earnings models in which we regress current years earnings (from Compustat) on
last years earnings and earnings in the year before. Then we use coefcient estimates based on
historical data and the most recent set of earnings data to calculate forecasted earnings for the next
year. D_NOL is equal to 1 if NOLC is positive and forecasted earnings are negative. In the second
alternative specication, we drop the requirement of expected losses. Rather, we require taxable
income to be positive in year t2 and to be negative in year t1 in the post-1997 period. In the pre-
1997 period, we require taxable income to be positive in year t3 and to be negative in year t2 and
t1. In both cases, we require that tax paid in the rst year of the NOL carryback window has not
been fully refunded. We nd robust empirical results across these two alternative specications
(results not tabulated).
September 2013

2 Tax Motivated Loss Shifting

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

2 Tax Motivated Loss Shifting

Uploaded by

Copyright:

Available Formats

THE ACCOUNTING REVIEW American Accounting Association

Vol. 88, No. 5 DOI: 10.2308/accr-50496

You might also like