The Impact of Post-Merger Accounting Integration on Long-Term M&A Success Tom Adams [email protected]Youree Kim [email protected]Todd Kravet [email protected]University of Connecticut School of Business 2100 Hillside Road, Unit 1041A Storrs, CT 06269-1041 Preliminary and Incomplete Current Draft: July 2018 We thank Jenny Luchs, Sarah Parsons, and workshop participants at the University of Connecticut for helpful comments. We also thank the University of Connecticut for financial support.
54
Embed
The Impact of Post-Merger Accounting Integration on Long-Term M&A Success › wp-content › uploads › 2018 › 01 › Kim-Y… · The Impact of Post-Merger Accounting Integration
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
The Impact of Post-Merger Accounting Integration on Long-Term M&A Success
Current Draft: July 2018 We thank Jenny Luchs, Sarah Parsons, and workshop participants at the University of Connecticut for helpful comments. We also thank the University of Connecticut for financial support.
The Impact of Post-Merger Accounting Integration on Long-Term M&A Success
ABSTRACT: M&As involve a substantial amount of accounting work from due diligence to integration accounting. In this study, we investigate whether accounting-related integration issues during the immediate post-merger period are associated with internal information quality and the combined entity’s long-term M&A success. We expect that firms with greater accounting-related integration issues, which includes purchase price allocation and the integration of accounting systems, will experience poorer internal information and long-term post-acquisition operating performance. We first document that our inverse measures of accounting-related integration quality, abnormally high audit fees and audit report lags in acquisition years, are positively associated with management EPS guidance error. We then document that our measures are negatively (positively) associated with long-term changes in acquirer cash flows (post-M&A goodwill impairments). Our results are consistent with integration quality, in the period immediately following deal completion, affecting both the quality of information produced from the acquirer and target’s newly integrated accounting system as well as long-term post-M&A financial outcomes. JEL codes: G34, M42 Keywords: Mergers & Acquisitions, Accounting Integration, Audit Fees, Audit Report Lags, Management Guidance, Goodwill Impairments
1
I. INTRODUCTION
The integration process immediately after mergers and acquisitions are completed is
expected to be important in realizing the synergies from combing the acquirer and target firm.
Once the acquisition is completed, the acquirer then starts integrating accounting systems of the
two entities, focusing on setting up common controls for various operational segments such as
compliance, reporting, or procurement, in order to operate as one company. Internal information
produced from the merged accounting system of the acquirer and target provides feedback to
management about the progress of the integration and is used to make decisions about the
integration. We expect difficulties with the integration that are related to accounting systems to
be particularly important because these issues likely result in worse internal information
available to managers and worse post-acquisition outcomes.
In this study, we examine whether accounting-related integration issues are associated
with firms’ post-merger internal information quality and their long-term acquisition outcomes.
Accounting-related integration issues are complications in the integration of a target into their
acquirer that affect the accounting system of the combined entity and can include operational
integration issues that affect financial reporting or issues directly stemming from the combining
of accounting systems and applying acquisition accounting under SFAS 141, Business
Combination (FASB, 2001a).1 We posit that quality of information produced from the
accounting system is important during the integration period as this is a critical time for setting
the stage for the combined entity’s operation (Haspeslagh et al. 1991). In line with this notion,
Feldman and Spratt (1998) argue that the first 100 days (i.e., approximately one quarter of a
1 The accounting standards effective during our sample period are SFAS 141, Business Combinations, and SFAS 142, Goodwill and Other Intangible Assets. (FASB 2001a, 2001b). SFAS 141 is revised and replaced by SFAS 141R after 2009 (FASB 2007).
2
year) is when all the critical actions should be launched, as this is likely to be the outer boundary
of the patience of the related parties of the combined firm such as customers and investors. We
expect accounting-related integration issues to adversely affect firms’ internal information and,
thus, disclosure quality. We also expect accounting-related integrations issues to be associated
with less successful acquisitions for two reasons. First, underlying operational integration issues
that affect accounting systems can cause acquirers not to realize expected synergies. Second,
poorer internal information stemming from accounting-related integration issues results in worse
managerial decision-making with respect to the integration. Quality of accounting during this
early post-merger period can be important because management is critically dependent on the
accounting information as this information provides data-driven input to their integration
decisions.
We use two measures of accounting-related integration issues based on abnormal audit
fees and abnormal audit report lags in the fiscal year the acquisition is completed. These
measures are based on prior audit research (Bamber et al. 1993; Knechel and Payne 2001;
Ashbaugh et al. 2003; Krishnan and Yang 2009) and capture instances where, during the
accounting integration period, either (1) the acquirer paid higher than expected audit fees or (2)
the acquirer’s auditor took longer than expected to sign its audit opinion. We argue that higher
than expected fees and/or longer than expected audit report lags capture difficulties experienced
by the acquirer in integrating the target company’s financial reporting and internal control
systems into its own. These two measures have a positive but modest correlation indicating they
capture different aspects of accounting-related integration issues. These also capture ex-post
actual integration issues that allow us to test whether acquirers and investors can predict which
acquisitions are likely to have integration issues.
3
We first test our argument that accounting-related integration issues decrease the quality
of accounting information produced from the newly integrated financial reporting and internal
control systems. We proxy for the quality of accounting information using management guidance
error (Rogers and Stocken 2005; Feng et al. 2009). Consistent with our predictions, we find that
both our proxies of accounting-related integration issues are positively associated with
management guidance error. This association is consistent with accounting-related integration
issues leading to worse internal accounting information and, thus higher management EPS
guidance error. Because accounting-related integration issues lead to lower internal information
quality we expect it also to result in worse decision-making during the integration and, thus
poorer post-acquisition outcomes.
We next test whether accounting-related integration issues are associated with long-term
post-acquisition outcomes. Specifically, we examine changes in acquirer cash flows and
goodwill impairments for the three-year period following the acquisition. We expect that when
post-acquisition accounting integration goes well (poorly), the newly integrated financial
reporting and internal control systems will produce higher (lower) quality information, which
will then be used by management to make better post-acquisition operational decisions. If good
(bad) information is used in these decisions, we expect to observe good (bad) long-term
outcomes. Underlying operational integration issues can also result in accounting integration
issues and cause poorer long-term outcomes. Consistent with our argument, we find that our
measures of accounting-related integration issues, abnormal audit fees and abnormal audit report
lags in the fiscal year of the acquisition, are negatively associated with post-acquisition changes
in cash flows. We also find that accounting-related integrations issues are positively associated
with goodwill impairments indicating that acquiring managers are more likely to overpay in
4
acquisitions with integration issues. This result suggests that acquiring managers are not able to
foresee these integration issues because otherwise they would have lowered the purchase price
and avoided an impairment. Overall, our empirical results are consistent with accounting-related
integration issues affecting both the quality of information produced from the newly integrated
firm’s accounting system and its long-term outcomes.
Prior research investigates whether various pre-acquisition acquirer accounting-related
characteristics are associated with acquisition performance based on acquirer announcement
returns (e.g., Biddle and Hilary 2006; McNichols and Stubben 2008; Biddle et al. 2009; Francis
and Martin 2010).2 A majority of these studies are based on the argument that acquisitions are
settings where managers’ incentives can diverge from shareholders resulting in agency costs but
that higher-quality or more informative accounting can increase monitoring and, thus decrease
agency costs. Our tests examine ex-post integration, rather than ex-ante acquirer accounting
characteristics expected to affect monitoring, and our arguments relate to the importance of
successful integration of accounting systems in the combining of business. Importantly, our
findings that ex-post accounting-related integration issues are positively (negatively) associated
with managerial forecasts errors (post-acquisition profitability) are incremental to the effect of
ex-ante acquirer characteristics, such as acquirers’ internal control weaknesses and earnings
quality. In further tests, we find that acquirer announcement returns cannot predict accounting-
related integration issues suggesting that while investors are aware that acquirer characteristics
2 Another branch of research looking at the pre-merger accounting quality at the target firm gauges its overall influence on M&A deal characteristics and on the profits to shareholders of the merging parties (Raman et al. 2013; Skaife and Wangerin 2013; Marquardt and Zur 2015; McNichols and Stubben 2015; Chen et al. 2017). For example, Marquardt and Zur (2015) find that target firm accounting quality influences the structure of the deal as well as the speed and the likelihood of deal completion. Skaife and Wangerin (2013) find that acquirers offer higher premiums for targets with low quality financial reporting, and that low quality financial reporting increases the likelihood of deal renegotiation. Chen et al. (2017) find that target firm accounting comparability is positively associated acquisition profitability.
5
are associated with acquisition outcomes there is no evidence investors are able to foresee ex-
post accounting-related integration issues. Overall, the results suggest that accounting-related
integration issues have an important association with firms’ post-acquisition information quality
and acquisition profitability but these integration issues are difficult for investors and acquiring
managers to predict ex-ante.
This study makes several contributions. First, we contribute to the literature examining
the importance of accounting quality in acquisitions. While prior research finds that ex-ante
acquirer and target characteristics are associated with acquirer announcement returns we focus
on ex-post realized integration issues and find that accounting-related integration issues are
important for the success of acquisitions but are difficult to predict by acquirers and investors.
Further, our ex-post measures of accounting integration are incrementally associated with
acquisition outcomes relative to ex-ante acquirer accounting characteristics. Second, we extend
literature examining the importance of the integration process in acquisitions. Prior literature
suffers from noisy measures of integration issues that are often measured ex-ante, such as
diversifying and foreign acquisitions. We develop an alternative measure of integration issues
based on the integration of accounting systems measured ex-post in the fiscal year the acquisition
is completed. Consistent with the long-standing argument that the acquisition integration process
is important, we find that the integration process is associated with ex-post acquisition
performance. However, we also find that investors are not able to predict at announcement
whether there will be accounting-based integration issues suggesting that it is difficult to foresee
which business combinations will encounter integration difficulties.
Lastly, we extend to the literature investigating the role of audit fees and audit report
lags. Investors are not likely to find the information about the amount of audit and other fees
6
disclosed in the proxy statement or 10-K useful, because while audit fee negotiation might have
information useful to investors, they do not learn the actual audit fees paid to the auditor until
disclosed in the following year’s definitive proxy statement (Hackenbrack et al. 2014). Also,
investors think that the auditor’s opinions and disclosures are often boilerplate and produce a
“largely uninformative pass-fail report” (Harris 2017). Therefore, in the context of acquisitions,
our study can show a new role of audit fees and audit report lags as predictors for future long-
term operating performance of the combined entity, which may be particularly useful to
investors.
The rest of the paper is organized as follows. In section II we discuss related literature
and develop our hypotheses. In section III we discuss research design. In section IV we discuss
sample selection and results. In section V we discuss the results of path analyses. In section VI
we provide the results from additional analyses. Finally, in section VII we provide concluding
remarks.
II. RELATED LITERATURE AND HYPOTHESES DEVELOPMENT
Accounting in Mergers and Acquisitions
Prior studies suggest that accounting plays an important role and involves a large volume
of work over the course of an acquisition (e.g., Lajoux and Elson 2000; Wangerin 2012; Shalev
et al. 2013; Marquardt and Zur 2015). When the acquisition process begins, the acquirer works
on the firm valuation to determine the price of the target and conducts preliminary due diligence.
In the due diligence phase of a deal, the review of financial statements is considered the “single
most important aspect of due diligence” (Lajoux and Elson 2000). After the acquisition
agreement is signed, a public disclosure of the acquisition and transactional due diligence follow.
7
Subsequent to the deal closure, the acquirers must allocate the purchase price to all separately
identifiable classes of assets acquired and liabilities assumed based on estimated fair values as of
the acquisition date according to SFAS 141, Business Combinations. Also, under SFAS 142,
Goodwill and Other Intangible Assets, any unallocated purchase price is recorded as goodwill to
reflect expected synergies and other future economic benefits resulting from the business
combination. Finally, once the purchase price allocation is done, the acquirer conducts more
detailed work on accounting integration, setting up centralized information flow for various
business segments such as compliance, planning and analysis, and reporting.
Quality of Accounting and its Impact on the M&A Process
Prior research investigating the role of accounting information in acquisitions often
focuses on the pre-merger accounting quality and its impact on deal outcomes. Previous studies
linking accounting quality to investment efficiency often look at the acquirer’s pre-acquisition
accounting quality (Biddle and Hilary 2006; McNichols and Stubben 2008; Biddle et al. 2009;
Francis and Martin 2010; Goodman et al. 2014; Kravet 2014; Harp and Barnes 2017). These
studies are generally based on the argument that acquirers with higher accounting quality will
make better investment decisions because their investment decisions will be reported more
accurately and transparently, which facilitates better decision-making and monitoring. Therefore,
pre-acquisition accounting quality represents an acquirer characteristic that is expected to be
associated with investment decisions. Biddle and Hilary (2006) relate accounting quality to
investment cash-flow sensitivity, implying that firms with poor accounting quality are more
likely to rely on internally generated cash flows rather than external financing to fund
investments. Also, McNichols and Stubben (2008) find that firms that manipulated earnings tend
to overinvest during the misreporting period, and Francis and Martin (2010) report that firms
8
with more conservative accounting make more profitable acquisitions. Harp and Barnes (2017)
find that problems in an acquirer’s internal control environment have negative operational
implications for acquisition performance. In general, these findings suggest that acquirers’ ex-
ante accounting quality improves investment efficiency by reducing information asymmetry and
increasing monitoring. Our study is interested in ex-post integration issues with respect to
accounting systems, which is not necessarily due to or related to agency problems between
managers and other stakeholders. Furthermore, our study differs from this literature to the extent
that integration issues do not arise because of acquirer characteristics but from idiosyncratic
factors related to the specific deal.
Another stream of literature focuses on the pre-merger quality of accounting at the target
firm by investigating (1) the relation between the quality of financial information provided by the
target and the profits to shareholders of the merging parties and (2) the target financial reporting
quality and its impact on the overall acquisition process (Raman et al. 2013; Skaife and
Wangerin 2013; Marquardt and Zur 2015; McNichols and Stubben 2015; Chen et al. 2018).
These studies are limited to acquisitions where the target is a public firm and publicly available
data is available. For example, McNichols and Stubben (2015) examine the relation between
target firm accounting quality and acquirer (target) profit from an acquisition and document a
positive (negative) relation between these two components. Marquardt and Zur (2015) contend
and find that high quality accounting information at the target firm reduces the costs of the
acquisition process, increases the likelihood of deal completion, and predict that target
accounting quality is negatively (positively) associated with the likelihood of an auction
(negotiation). These studies investigate the implications of target firms’ accounting information
quality from the acquirers’ perspective. They focus on how a target firm’s accounting quality can
9
influence the acquiring firm’s investment decisions. Overall, this line of research suggests that
the target firm accounting quality affects deal structures in mitigating adverse selection risk and
influences shareholder values for both the target and the acquirer.
We extend this literature examining the relation between pre-acquisition accounting
quality and acquisition decisions by considering how and whether ex-post accounting-related
integration issues for the immediate post-merger period, which includes both purchase price
allocation and accounting system integration, is associated with internal information quality and
long-term success of acquisitions. By looking at the ex-post integration period our study captures
a period that is likely critical for the combined entity’s future performance and that cannot be
completely predicted from acquirer and target pre-acquisition characteristics, thereby giving a
more complete picture of the importance of accounting integration in acquisitions. We especially
focus on operating performance measures rather than short-term measures such as cumulative
abnormal returns surrounding the deal announcement date, because it is not clear if integration
issues can be predicted at acquisition announcements and synergies expected from the deal can
only be fully realized by better operating performance on a long-term basis.
Few studies look at the accounting for the immediate post-merger period and its
association with the combined entity (Bens et al. 2012; Shalev et al. 2013; Paugam et al. 2015).
Shalev et al. (2015) find that CEOs whose compensation packages rely more on earnings-based
bonuses are more likely to over allocate the purchase price to goodwill, the largest asset recorded
post-merger. Paugam et al. (2015) finds that abnormal goodwill allocation is associated with a
negative market reaction when the purchase price allocation is disclosed and a higher likelihood
of future goodwill impairment. As noted here, despite the expectation that post-acquisition
accounting integration is important, this line of literature often focuses on purchase price
10
allocation only and ignores the overall integration of accounting systems. Studies involving the
quality of both purchase price allocation and integration accounting have been relatively
unexplored by the literature, especially in terms of long-term operating performance.
Accounting for Immediate Post-Merger Period and Management Guidance
The period immediately following deal completion is a critical time for the newly merged
firm. During this period, many things can go wrong. For example, the acquirer must perform a
purchase price allocation, whereby the transaction value is allocated to acquired assets and
liabilities (ASC 805). This process is complex and requires significant managerial judgment
(Laux and Leuz 2009). Managers often make mistakes; acquisition-related restatements comprise
8-17 percent of all financial restatements (Scholz 2008, 2014). In addition to performing a
purchase price allocation, an acquirer must integrate the target’s accounting and internal control
systems into its own. Recognizing these complexities, GAAP provides acquirers with a one-year
grace period, the measurement period, during which the acquirer can make adjustments to
with a one-year exemption from testing acquiree internal controls (SEC 2004). Thus, both the
FASB and SEC formally recognize the accounting and internal control complexity of the post-
merger period.
For these (and other) reasons, acquisitions are not always successful. High quality
integration, particularly in the period immediately following acquisition completion (e.g., the
first 100 days), is paramount for long-term acquisition success (Feldman and Spratt 1998;
Homburg and Bucerius 2006). An often overlooked, but a vitally important part of post-merger
integration, is the melding of the acquirer’s and acquiree’s accounting and internal control
systems (Deloitte 2018). Information produced from these systems is used, by managers, to make
11
operational decisions, prepare financial statements, and to make both mandatory and voluntary
disclosures (Feng et al. 2009). If the integration of the acquirer’s and acquiree’s accounting and
internal control systems goes well (poorly), we expect that the newly integrated systems will
produce high (low) quality information.
We proxy for the quality of information produced from a firm’s post-acquisition
accounting and internal control systems with management earnings forecast error. Firm
managers often voluntarily disclose earnings forecasts to manage expectations, reduce
information asymmetry between the firm and its stakeholders, and to reduce litigation risk
(Rogers and Stocken 2005; Hirst et al. 2008; Feng et al. 2009). Many attributes of management
guidance have been studied, including its determinants, its characteristics, and its consequences
(Hirst et al. 2008). However, the attribute most relevant to our prediction is guidance error. The
accuracy of management forecasts depend largely upon the inputs to the forecasts (Feng et al.
2009). Thus, we posit that if integration quality is high (low), the firm’s systems will produce
high (low) quality information, which will lead to more (less) accurate guidance. Stated formally
in the alternative form, our first hypothesis is:
H1: Accounting-related integration issues are negatively associated with the accuracy of post-acquisition management guidance.
Accounting for the Immediate Post-Merger Period and the Acquisition Success
If poorer accounting-related integration results in lower quality internal accounting
information then we expect poorer managerial decision-making during the integration that is
based on accounting information. We therefore posit that great accounting-related integration
issues during the immediate post-merger period contributes to poorer acquisition outcomes,
because it produces less accurate and less useful accounting information and increases
information asymmetry between the acquired and the acquirer during probably the most critical
12
period for the combined firm (Haspeslagh et al. 1991; Angwin 2004). It is also possible that
accounting-related integration issues can stem from broader integration issues related to
combining operations. If operational integration issues lead to lower than expected synergies,
then we expect also expect these type of integration issues to lead to worse acquisition outcomes.
Overall, we expect greater accounting-related integration issues to be associated with worse
acquisition outcomes.
However, if managers can anticipate integration issues then they can either not pursue the
deal or adjust the purchase price they offer to reflect the expected integration costs. If managers
forgo deals with high integration costs or pay a lower premium then we would not expect an
association between accounting-related integration issues and acquisition outcomes.
Given these arguments, we expect worse (better) quality of accounting for the immediate
post-merger period to be associated with worse (better) long-term success of acquisitions. As
such our second hypothesis, is as follows (alternative form):
H2: Accounting-related integration issues are negatively associated with the long-term success of acquisitions.
III. RESEARCH DESIGN
Abnormal Audit Fees and Audit Report Lags as Measures of Integration Issues
We develop two ex-post measures of realized accounting-related integration issues that
we expect to capture difficulty in accounting for the acquisition transaction and in combining of
accounting systems. Thus, the accounting quality measures which can reflect managers’ use of
discretion more would not be an effective proxy for the M&A setting. Therefore, we use two
alternative measures of accounting quality, abnormal audit fees and abnormal audit report lags,
which should be less influenced by firm risks or management’s earnings manipulation (Bamber
13
et al. 1993; Knechel and Payne 2001; Ashbaugh et al. 2003; Krishnan and Yang 2009; Hribar et
al. 2014). Both abnormal audit fees and abnormal audit report lags are inverse measures of
accounting-related integration quality.
We expect accounting-related integration difficulties to directly affect auditors in their
audit of the combined firm’s financial reporting. Difficulty in accounting for the acquisition or
combining accounting systems results in increased effort and audit risk for auditors as it is more
difficult to provide assurance that the combined financial statements are consistent with GAAP.
Prior literature finds that increased auditor effort and risk results in higher audit fees and longer
audits (Bamber et al. 1993; Knechel and Payne 2001; Ashbaugh et al. 2003; Krishnan and Yang
2009). We use abnormal audits fees in the fiscal year the acquisition is completed to capture
instances where, during the accounting integration period, the acquirer paid higher than expected
audit fees. We use abnormal audit report lag in the fiscal year the acquisition is completed to
capture instances where, during the integration period, the acquirer’s auditor took longer than
expected to sign its audit opinion. We argue that higher than expected fees and/or longer than
expected audit report lags capture difficulties experienced by the acquirer in integrating the
target company’s financial reporting and internal control systems into its own. We do not use
measures based on accounting numbers (e.g., accrual quality) because they are correlated with
firms’ operating risk or can be easily manipulated by managers (Marquardt and Zur 2015; Bens
et al. 2012).
Abnormal Audit Fees and Audit Report Lags and Management Guidance Error
We estimate the following model using OLS to test the association between management
EPS guidance error and abnormal audit fees and abnormal audit report lags.
GUIDANCE_ERROR = α0 + α1Integration Issue Measure + αiControli + Industry & Year Fixed Effects + ε
(1)
14
The dependent variable, GUIDANCE_ERROR, is the absolute difference between management
EPS guidance and actual EPS (measured in the first full fiscal year following the deal), scaled by
share price one month prior to fiscal year-end (Rogers and Stocken 2005; Feng, Li, and McVay
2009). We also specify the dependent variable, GUIDANCE_ERROR, as an indicator variable
coded 1 for high guidance error; coded 0 otherwise.3 Integration Issue Measure represents our
three test variables, which are used alternately in model 1. First, ABN_AFEE is the acquirer’s
abnormal audit fees (i.e., regression residual audit fee), measured in the fiscal year the deal is
completed using the audit fee model from Ashbaugh, LaFond, and Mayhew (2003).4 Results
from estimating the audit fee model are included in Appendix B.5 Second, ABN_AUDRPT_LAG
(abnormal audit report lag) is the audit report lag measured for the fiscal year the deal is
completed less the average audit report lag in the three years preceding the deal, where audit
report lag is defined as the number days between fiscal year-end and the audit opinion signature
date (Bamber Bamber, and Schoderbek 1993; Knechel and Payne 2001; Krishnan and Yang
2009).6 Third, INTEGRATION_ISSUES, aggregates ABN_AFEE and ABN_AUDRPT_LAG by
transforming each into decile ranks (0 to 9) and dividing by 9. The two transformed values (each
ranging in value from 0 to 1) are then added, resulting in an aggregate measure of integration
3 High and low amounts of guidance error are determined using sample median guidance error. When we use the indicator variable, HIGH_GUIDANCE_ERROR, we estimate model 1 using a logistic regression model. 4 In addition to including all variables from the Ashbaugh et al. model, we include lagged audit fees as a control for firm-specific factors that could affect current year audit fees but that are unrelated to the acquisition in the current year. 5 We chose the Ashbaugh et al. (2003) audit fee model because, relative to other audit fee models, it requires only a limited set of Audit Analytics and Compustat based variables. Yet, the model has high explanatory power; results from our model estimation, presented in Appendix B, yielded an adjusted-R2 of 0.883. Further, the Asbaugh et al. model includes a control variable for M&A activity, which is necessary in our setting. 6 Results, presented later in the paper, are robust to measuring audit report lags using both raw audit report lag days (i.e., not abnormal audit report lag days) and the natural log of audit report lag days.
15
issues (ranging in value from 0 to 2). We expect positive coefficients for all three test variables
indicating that greater accounting-related integration issues is positively associated with
management guidance accuracy.
Control variables are based on prior acquisition research and include pre-acquisition
measures of acquirer size (Moeller et al. 2004; ACQ_SIZE), profitability (Dechow et al. 2011;
ACQ_ROA), leverage (Maloney et al. 1993; ACQ_LEV), book-to-market ratio (Dong et al., 2006;
Tuch and O’Sullivan, 2007; ACQ_BTM), and return volatility (Dechow et al. 2011;
ACQ_STD_RET). We also include controls capturing acquirer pre-acquisition internal control
(Darrough et al. 2018; ACQ_ICW) and financial reporting quality (Kothari et al. 2005; Krishnan
et al. 2011; ABS_PADACC), as well as the acquirer’s use of a Big 4 audit firm (BIG4). Finally,
we include deal specific controls for the target company’s public/private status (PUBLIC_MA),
deal announcement period returns (MM_RET3), relative deal size (DL_REL_SIZE), form of
payment (Ghosh 2001; Gu and Lev, 2011; DL_STOCK), type of deal (i.e., diversifying vs. non-
diversifying, Morck et al. 1990; DL_DIVER), and length of time to deal completion (Wangerin
2012; DL_DUE_DIL).
Abnormal Audit Fees and Audit Report Lags and Long-Term Post-M&A Performance
To test the association between accounting-related integration issues we use two
measures of long-term, post-acquisition performance established in prior literature: changes in
acquirer cash flows between the pre- and post-acquisition periods and post-acquisition goodwill
impairments. We first estimate the following model using OLS to test the association between
long-term changes in acquirer cash flows and abnormal audit fees and abnormal audit report
lags:
CHG_CFO = γ0 + γ1Integration Issue Measure + γiControli + Industry & Year Fixed Effects + θ
(2)
16
The dependent variable in model 2, CHG_CFO, is the difference between the average cash flow
from operations (CFO) for the post-acquisition periods, t+1 through t+3, and the pre-acquisition
periods, t-3 through t-1, where period t is the year in which the acquisition was completed. If
CFO data are unavailable for periods t+3 or t-3, average CFO is calculated using two (instead of
three) years of data in the pre- and post-deal periods. Integration Issue Measure represents our
three test variables, ABN_AFEE, ABN_AUDRPT_LAG, and INTEGRATION_ISSUES, which are
used alternately in model 2. All of these variables were defined above. Our second hypothesis
predicts that high (low) post-acquisition accounting integration quality will translate into strong
(poor) long-term, post-acquisition financial performance. As such we expect to find a negative
association between our measures of accounting-related integration issues, ABN_AUDFEE and
ABN_AUDRPT_LAG, and CHG_CFO. The control variables used are consistent with those used
in model 1 and are defined, in detail, in the appendix.
We next estimate the following model using logistic regression to test the association
between long-term changes in acquirer cash flows and abnormal audit fees and abnormal audit
report lags:
Pr(GW_IMPAIRMENT = 1) = F{δ0 + δ1Integration Issue Measure + δiControli + Industry & Year Fixed Effects}
(3)
The dependent variable in model 3, GW_IMPAIRMENT, is an indicator variable coded 1 if the
acquirer recorded a goodwill impairment in the year of the deal (period t) or in the post-deal
period (periods t+1 through t+3); coded zero otherwise (Chen et al. 2018). Goodwill captures
expected synergies and future economic benefits generated in a business combination that do not
meet the criteria for asset recognition (Johnson and Petrone 1998). Therefore, goodwill
impairments in the years following acquisitions provide evidence that synergies were not
17
achieved or future economic benefits were not realized. Integration Issue Measure represents our
three test variables, ABN_AFEE, ABN_AUDRPT_LAG, and INTEGRATION_ISSUES, which are
used alternately in model 3. All of these variables were defined above. Our second hypothesis
predicts that strong (poor) post-M&A accounting integration will be associated with good (bad)
financial outcomes. Therefore, we predict a positive association between our inverse measures of
post-M&A accounting quality, ABN_AFEE, ABN_AUDRPT_LAG, and INTEGRATION_ISSUES,
and post-acquisition goodwill impairments. All control variables are defined in detail in the
Appendix and are consistent with those discussed in connection with model 1 above.
IV. SAMPLE SELECTION AND RESULTS
Sample and Data
Table 1 details our sample selection procedures. We begin with a sample of acquisitions
obtained from the Thomson One SDC database. We require acquisitions to have been completed
and to have involved a US public acquirer and a US target (either public or private).7 While our
analyses cover firm years between 2002 and 2016, we restrict our sample to M&A deals
completed before 2013 so that we have sufficient data to measure our dependent variables,
GUIDANCE_ERROR, CHG_CFO, and GW_IMPAIRMENT in the post-deal period. We also
focus on the post-SOX period (i.e., years 2002 and later), require relative deal sizes (the ratio of
transaction value to acquirer market value) of ≥ 5 percent, and non-missing acquirer identifiers.8
Our initial sample obtained from the SDC database included 5,725 acquisition observations.
Finally, after requiring data necessary for the construction of our dependent, control, and test
7 In our multivariate analyses, we control for the target company’s public/private status. 8 One of our control variables, ACQ_ICW, requires internal control data which is available only in periods 2002 and later (Darrough, Huang, and Zur 2018)
18
variables, our final sample for the management guidance error, change in cash flows, and
goodwill impairment analyses includes 1,314, 2,681, and 3,976 observations, respectively.9
Management Guidance Error
Next, we begin our analyses of post-acquisition accounting integration’s effect on
management guidance error. As discussed above, post-acquisition management guidance error
proxies for the quality of accounting information produced from the newly integrated financial
reporting and internal control systems of the acquirer and target companies. Our expectation is
that, if post-merger integration goes well (poorly), the quality of accounting information
produced from the financial reporting and internal controls systems will be high (low).
Management guidance error is an inverse measure of accounting information quality. Similarly,
abnormal fees and abnormal audit report lags are inverse measures of accounting integration
quality. Therefore, we expect to find a positive association between our measures of accounting
integration quality, ABN_AFEE and ABN_AUDRPT_LAG, and GUIDANCE_ERROR.
Table 2, Panels A and B, present descriptive statistics for our dependent variable
(GUIDANCE_ERROR) and our three test variables (ABN_AFEE, ABN_AUDRPT_LAG, and
INTEGRATION_ISSUES), as well as t-tests for our two test variables and control variables.
When we partition our sample into high/low guidance error groups (Panel B) we find that mean
abnormal audit fees and abnormal audit report lags for the high guidance error group
(HIGH_GUIDANCE_ERROR = 1) are significantly larger than those for the low guidance error
group (HIGH_GUIDANCE_ERROR = 0). The same is true for our aggregate integration issues
9 These final observation counts relate to analyses that use ABN_AFEE as a test variable. When ABN_AUDRPT_LAG is used as our test variable our sample sizes for the management guidance error, change in cash flows, and goodwill impairment analyses are further reduced to 1,277, 2,659, and 3,942 observations respectively. When INTEGRATION_ISSUES is used as our test variable our sample sizes for the management guidance error, change in cash flows, and goodwill impairment analyses are further reduced to 1,227, 2,538, and 3,760 observations respectively.
19
measure.10 Next, among our control variables (the bottom portion of Panel B), we find firms with
high guidance error have significantly higher book-to-market ratios (ACQ_BTM), return
volatility (ACQ_STD_RET), and relative deal sizes (DL_REL_SIZE). We also find that firms
with high guidance error purchased public targets (PUBLIC_MA) and used Big 4 auditors (BIG4)
less often. These same firms are smaller and less profitable (ACQ_SIZE and ACQ_ROA).
In Table 2, Panel C, we provide a Pearson correlation matrix. Consistent with our
univariate results presented in Panel B, we find positive correlations between
GUIDANCE_ERROR and ABN_AFEE, ABN_AUDRPT_LAG, and INTEGRATION_ISSUES,
although the correlation for ABN_AFEE is not significant (p=0.21). We also note that our two
proxies for post-acquisition integration quality, ABN_AFEE and ABN_AUDRPT_LAG, have a
positive correlation of 0.140 (p≤0.01), which provides some comfort that these proxies are not
entirely overlapping measures of our theoretical construct. Finally, as expected, both ABN_AFEE
and ABN_AUDRPT_LAG are highly correlated with the aggregate INTEGRATION_ISSUES
measure (Pearson correlations of 0.692 and 0.567, respectively; p≤0.01 for both).
Next, Table 3 provides results from our multivariate tests of H1. Again, our expectation
is that when post-acquisition accounting integration goes well (poorly) the quality of accounting
information produced from the financial reporting and internal control systems will be positively
(negatively) affected. We, therefore, expect to find positive associations between
GUIDANCE_ERROR and ABN_AUDFEE, ABN_AUDRPT_LAG, and INTEGRATION_ISSUES.
Consistent with this prediction, in columns 1-3 we find a positive associations for all three test
variables. In addition, when we specify our dependent variable as a binary variable measuring
high and low amounts of guidance error in columns 4-6, we find positive associations for all
10 The high/low guidance error sample partitions in Table 2, Panel B are defined using sample medians.
20
three test variables. Overall, the results in columns 1-6 are consistent with post-acquisition
accounting integration quality affecting the quality of information produced from the acquirer’s
post-merger accounting and internal control systems.
Finally, among our control variables we find evidence that GUIDANCE_ERROR is
negatively associated with the use of a Big 4 auditor (BIG4, columns 1-3), M&A announcement
returns (MM_RET3, columns 1-3), and acquirer size (ACQ_SIZE, columns 1-6) and profitability
(ACQ_ROA, column 5). Conversely, GUIDANCE_ERROR is positively associated with
discretionary accruals (ABS_PADACC, column 5), book-to-market ratio (ACQ_BTM, columns 4-
Available at: https://asc.fasb.org/topic&trid=2303972. Angwin, D. 2004. Speed in M&A Integration: the First 100 Days. European Management
Journal 22 (4): 418-430. Ashbaugh, H., R. LaFond, and B.W. Mayhew. 2003. Do Nonaudit Services Compromise Auditor
Independence? Further Evidence. The Accounting Review 78 (3): 611-639.
Paugam, L., P. Astolfi , and O. Ramond. 2015. Accounting for Business Combinations: Do Purchase Price Allocations Matter? Journal of Accounting and Public Policy 34(4): 362-391.
Bamber, E.M., L.S. Bamber, and M.P. Schoderbek. 1993. Audit Structure and Other Determinants of Audit Report Lag: An Empirical Analysis. Auditing: A Journal of Practice & Theory 12 (1): 1-23.
Biddle, G., and G. Hilary G. 2006. Accounting Quality and Firm-Level Capital Investment. The Accounting Review 81 (5): 963–982.
Biddle, G., G. Hilary, and R. Verdi. 2009. How Does Financial Reporting Quality Relate to Investment Efficiency? Journal of Accounting and Economics 48 (2-3): 112-131.
Bens, D.A., T.H. Goodman, and M. Neamtiu. 2012. Does Investment-Related Pressure Lead to Misreporting? An Analysis of Reporting Following M&A Transactions. The Accounting Review 87 (3): 839-865.
Chen, C-W., D.W. Collins, T.D. Kravet, and R.D. Mergenthaler. 2018. Financial Statement Comparability and the Efficiency of Acquisition Decisions. Contemporary Accounting Research 35 (1): 164-202.
Darrough, M., R. Huang, and E. Zur. 2018. Acquirer Internal Control Weaknesses in the Market for Corporate Control. Contemporary Accounting Research 35 (1): 211-244.
Dechow, P., and I. Dichev. 2002. The Quality of Accruals and Earnings: the Role of Accrual Estimation Errors. The Accounting Review 77 (Supplement): 55-79.
Dechow, P.M., W. Ge, C.R. Larson, and R.G. Sloan. 2011. Predicting Material Accounting Misstatements. Contemporary Accounting Research 28 (1): 17-82.
Deloitte. 2018. On the Board’s Agenda: Post-merger Integration. Available at:
Dickinson, V., D.D. Wangerin, and J.J. Wild. 2016. Accounting Rules and Post-Acquisition Profitability in Business Combinations. Accounting Horizons 30 (4): 427-447.
Dong, M., D. Hirshleifer, S. Richardson, and S.H. Teoh. 2006. Does Investor Misvaluation Drive the Takeover Market? Journal of Finance 61 (2): 725–762.
30
Duchin, R. and B. Schmidt. 2013. Riding the Merger Wave: Uncertainty, Reduced Monitoring, and Bad Acquisitions. Journal of Financial Economics 107 (1): 69-88.
Feldman, M.L. and M.F. Spratt. 1998. Five Frogs on a Log: A CEO's Field Guide to Accelerating the Transition in Mergers, Acquisitions and Gut Wrenching Change. Harper Collins.
Feng, M., C. Li, and S. McVay. 2009. Internal Control and Management Guidance. Journal of Accounting and Economics 48 (2-3): 190-209.
Francis, J.R., and X. Martin. 2010. Acquisition Profitability and Timely Loss Recognition. Journal of Accounting and Economics 49 (1-2): 161-178.
Financial Accounting Standards Board (FASB). Business Combinations. “Statement of Financial Accounting Standards No. 141.” Norwalk, CT: FASB, 2001a.
Financial Accounting Standards Board (FASB). Goodwill and Other Intangible Assets. “Statement of Financial Accounting Standards No. 142.” Norwalk, CT: FASB, 2001b.
Financial Accounting Standards Board (FASB). Business Combinations. “Statement of Financial Accounting Standards No. 141-R.” Norwalk, CT: FASB, 2007.
Goodman, T.H., M. Neamtiu, N. Shroff, and H.D. White. 2014. Management Forecast Quality and Capital Investment Decisions. The Accounting Review 89 (1): 331-365.
Gu, F. and B. Lev. 2011. Overpriced Shares, Ill-Advised Acquisitions, and Goodwill Impairment. The Accounting Review 86 (6): 1995-2022.
Ghosh, A., 2001. Does Operating Performance Really Improve Following Corporate Acquisitions? Journal of Corporate Finance 7 (2): 151-178.
Hackenbrack, K.E., N.T. Jenkins, and M. Pevzner. 2014. Relevant but Delayed Information in Negotiated Audit Fees. Auditing: A Journal of Practice & Theory 33 (4): 95–117.
Harris, S.B. 2017. Statement on Adoption of an Auditing Standard on the Auditor’s Report. Public Company Accounting Oversight Board (PCAOB) Open Board Meeting, Washington. Available at: https://pcaobus.org/News/Speech/Pages/Harris-statement-auditors-report-standard-adoption-6-1-17.aspx
Haspeslagh, P.C. and D.B. Jemison. 1991. Managing Acquisitions: Creating Value through Corporate Renewal (Vol. 416). New York: Free Press.
Harp, N.L., B.G. Barnes. 2017. Internal Control Weaknesses and Acquisition Performance. The Accounting Review 93 (1): 235-258.
Hayes, A.F. 2018. Introduction to Mediation, Moderation, and Conditional Process Analysis: A Regression-Based Approach (2nd Edition). New York, NY: The Guilford Press.
Hayn, C. and P.J. Hughes. 2006. Leading Indicators of Goodwill Impairment. Journal of
Accounting, Auditing & Finance 21 (3): 223-265.
31
Hirst, D.E., L. Koonce, and S. Venkataraman. 2008. Management Earnings Forecasts: A Review and Framework. Accounting Horizons 22 (3): 315-338.
Homburg, C. and M. Bucerius. 2006. Is Speed of Integration Really a Success Factor of Mergers
and Acquisitions? An Analysis of the Role of Internal and External Relatedness. Strategic Management Journal 27 (4): 347-367.
Hribar, P., T.D. Kravet, R. Wilson. 2014. A New Measure of Accounting Quality. Review of
Accounting Studies 19 (1): 506–538. Johnson, T.L. and K.R. Petrone. 1998. Is Goodwill an Asset? Accounting Horizons 12 (3): 293-
303.
Jones, J.J. 1991. Earnings Management During Import Relief Investigations. Journal of Accounting Research 29 (2): 193–228.
Kanodia, C., D. Lee. 1998. Investment and Disclosure: The Disciplinary Role of Periodic Performance Reports. Journal of Accounting Research 36 (1): 33-55.
Knechel, W.R., J.L. Payne. 2001. Additional Evidence on Audit Report Lag. Auditing: A Journal of Practice & Theory 20 (1): 137-146.
Kothari, S.P., A.J. Leone, and C. Wasley. 2005. Performance Matched Discretionary Accrual Measures. Journal of Accounting and Economics 39 (1): 163-197.
Kravet, T.D. 2014. Accounting Conservatism and Managerial Risk-Taking: Corporate Acquisitions. Journal of Accounting and Economics 57 (2-3): 218-240.
Krishnan, J. and J.S. Yang. 2009. Recent Trends in Audit Report and Earnings Announcement Lags. Accounting Horizons 23 (3): 265-288.
Krishnan, J., Y. Wen, and W. Zhao. Legal Expertise on Corporate Audit Committees and Financial Reporting. The Accounting Review 86 (6): 2099-2130.
Laux, C. and C. Leuz. 2009. The Crisis of Fair-Value Accounting: Making Sense of the Recent Debate. Accounting, Organizations and Society 34 (6-7): 826-834.
Lajoux, A., C. Elson. 2000. The Art of M&A Due Diligence. New York: McGraw-Hill.
Maloney, M.T., R.E. McCormick, and M.L. Mitchell. 1993. Managerial Decision Making and Capital Structure. Journal of Business 66 (2): 189-217.
Marquardt, C. and E. Zur. 2015. The Role of Accounting Quality in the M&A Market. Management Science 61 (3): 604-623.
McNichols, M. and S. Stubben. 2008. Does Earnings Management Affect Firms’ Investment Decisions? The Accounting Review 83 (6): 1571-1603.
McNichols, M. and S. Stubben. 2015. The Effect of Target-Firm Accounting Quality on Valuation in Acquisitions. Review of Accounting Studies 20 (1): 110-140.
32
Moeller, S.B., F.P. Schlingemann, and R.M. Stulz. 2004. Firm Size and the Gains from Acquisitions. Journal of Financial Economics 73 (2): 201–228.
Morck, R., A. Shleifer, and R.W. Vishny. 1990. Do Managerial Objectives Drive Bad Acquisitions? Journal of Finance 45 (1): 31–48.
Nikolaev, V. 2014. Estimating Accounting Quality. Working Paper, University of Chicago.
Raman, K., L. Shivakumar, and A. Tamayo. 2013. Targets’ Earnings Quality and Bidders’ Takeover Decisions. Review of Accounting Studies 18 (4): 1050-1087.
Rogers, J.L. and P.C. Stocken. Credibility of Management Forecasts. The Accounting Review 80 (4): 1233-1260.
Scholz, S. 2008. The Changing Nature and Consequences of Public Company Financial Restatements: 1997-2006. The Department of Treasury. Available at: http://www.lexissecuritiesmosaic.com/uploaded/ResourceCenter/FinancialRestatements_1997_2006.pdf.
Scholz, S. 2014. Financial Restatement: Trends in the United States (2003-2012). Center for
Audit Quality. Available at: https://www.thecaq.org/sites/default/files/financial-restatement-trends-in-the-united-states-2003-2012.pdf.Shalev, R. 2007. Managerial Discretion in Business Combinations Reporting Under SFAS 141. Working Paper, Columbia University.
SEC. 2004. Management’s Report on Internal Control Over Financial Reporting and Certification of Disclosure in Exchange Act Periodic Reports: Frequently Asked Questions (revised October 6, 2004). https://www.sec.gov/info/accountants/controlfaq1004.htm.
Shalev, R., I.X. Zhang, Y. Zhang. 2013. CEO Compensation and Fair Value Accounting: Evidence from Purchase Price Allocation. Journal of Accounting Research 51 (4): 819-854.
Skaife, H.A., D.D. Wangerin. 2013. Target Financial Reporting Quality and M&A Deals that Go Bust. Contemporary Accounting Research 30 (2): 719-749.
Tuch, C. and N. O'Sullivan. 2007. The Impact of Acquisitions on Firm Performance: A Review of the Evidence. International Journal of Management Reviews 9 (2): 141-170.
Valeri, L. and T.J. VanderWeele. 2013. Mediation Analysis Allowing for Exposure-Mediator Interactions and Causal Interpretation: Theoretical Assumptions and Implementation with SAS and SPSS Macros. Psychological Methods 18 (2): 137-150.
Wangerin, D.D. 2012. M&A Due Diligence and Its Consequences for Post-Acquisition Financial Statements. Working paper, Michigan State University.
33
Appendix A Variable Definitions
Variable Definition GUIDANCE_ERROR The absolute difference between management EPS guidance and actual EPS
(measured in the first full fiscal year following the deal), scaled by share price one month prior to fiscal year-end (Rogers and Stocken 2005; Feng, Li, and McVay 2009).
HIGH_GUIDANCE_ERROR An indicator variable coded 1 if GUIDANCE_ERROR is ≥ sample median GUIDANCE_ERROR; coded 0 otherwise.
INTEGRATION_ISSUES This variable aggregates ABN_AFEE and ABN_AUDRPT_LAG by transforming each into decile ranks (0 to 9) and dividing by 9. The two transformed values for ABN_AFEE and ABN_AUDRPT_LAG (each ranging in value from 0 to 1) are then added, resulting in an aggregate measure of integration issues (ranging in value from 0 to 2).
CHG_CFO Change in industry-adjusted acquirer cash flows (CFO) between pre- and post-deal periods, where CFO is calculated as cash flow from operations divided by average total assets. CHG_CFO is the difference between the average CFO for the post-deal periods, t+1 through t+3, and the pre-deal periods, t-3 through t-1, where period t is a year in which the acquisition was completed. If CFO data are unavailable for periods t+3 or t-3, average CFO is calculated using two (instead of three) years of data in the pre- and post-deal periods.
GW_IMPAIRMENT GW_IMPAIRMENT is an indicator variable coded 1 if the acquirer recorded a goodwill impairment in the year of the deal (period t) or in the post-deal period (periods t+1 through t+3); coded zero otherwise.
ABN_AFEE ABN_AFEE is the acquirer’s abnormal audit feess, measured in the year of the deal using the audit fee model from Ashbaugh, LaFond, and Mayhew (2003). Results from this estimation are presented in Appendix B.
ABN_AUDRPT_LAG Audit report lag measured for the year of the deal less average audit report lag in the three years preceding the deal, where audit report lag is defined as the number days between fiscal year-end and the audit opinion signature date (Bamber Bamber, and Schoderbek 1993; Knechel and Payne 2001; Krishnan and Yang 2009).
ACQ_ICW Indicator variable coded 1 if the acquirer reported ICWs in the fiscal year prior to the deal announcement; coded 0 otherwise (Darrough, Huang, and Zur 2018).
ABS_PADACC Absolute performance-adjusted discretionary accruals (Kothari, Leone, and Wasley 2005; Krishnan, Wen, and Zhao 2011).
PUBLIC_MA An indicator variable coded 1 if the target company was a publicly traded company prior to the M&A; coded 0 otherwise.
BIG4 An indicator variable coded 1 if the acquirer used a Big 4 auditor (PwC, KPMG, Deloitte, or E&Y); coded 0 otherwise.
MM_RET3 Announcement returns measured as three-day market model adjusted returns. ACQ_SIZE Acquirer size. Measured as the natural logarithm of acquirer's market value
fifty trading days prior to an acquisition announcement. ACQ_ROA Acquirer's pre-acquisition ROA. Measured as operating income after
depreciation scaled by average total assets (oiadp / ((at +lag_at)/2)) at the fiscal year end prior to acquisition announcement. Industry-adjusted.
ACQ_LEV Acquirer's pre-acquisition leverage. Measured as the sum of short-term debt and long-term debt scaled by total assets ((dlc+dltt)/at) at the fiscal year end prior to acquisition announcement.
Appendix A continues on next page.
34
Appendix A (continued) Variable Definitions
Variable Definition ACQ_BTM Acquirer's pre-acquisition book-to-market ratio. Measured as book value of
equity divided by market value of equity (ceq/(prcc_f *csho)) at the fiscal year end prior to acquisition announcement.
ACQ_STD_RET Acquirer's standard deviation of daily returns computed over the one-year period ending one month before an acquisition announcement (One-year period is deemed to have 250 trading days).
DL_REL_SIZE Relative deal size. Measured as the ratio of the transaction value to the market value of the acquirer.
DL_STOCK Indicator variable equal to one if at least 50 percent of the consideration paid for the target consists of stock, and zero otherwise.
DL_DIVER Indicator variable equal to one for a diversifying acquisition (i.e., an acquisition where the acquirer and target operate in different two-digit SIC codes), and zero otherwise.
DL_DUE_DIL Proxy for managerial effort in M&A. The number of weekdays between the signing of the acquisition agreement and the effective date of the deal.
Audit Fee Model (Appendix B)
lnAUDIT_FEE The natural log of the audit fee in millions of dollars.
BIG4 An indicator variable coded 1 if the firm used a Big 4 auditor (PwC, KPMG, Deloitte, or E&Y); coded 0 otherwise.
lnMVE The natural log of the firm’s market value of equity defined as the firm’s price per share at fiscal year-end multiplied by the number of shares outstanding measured in millions of dollars.
MERGER An indicator variable coded 1 if the firm engaged in an M&A (identified by Compustat variable AQC); coded 0 otherwise.
FINANCING An indicator variable coded 1 if the number of shares outstanding increased by at least 10 percent or long-term debt increased by at least 20 percent; coded 0 otherwise.
MB The firm’s market to book ratio defined as its market value of equity divided by book value.
LEVERAGE The firm’s total liabilities divided by total assets.
ROA The firm’s return-on-asset ratio calculated as income before extraordinary items divided by average total assets.
AR_IN The sum of the firm’s receivables and inventory divided by its total assets.
NEGATIVE_ROA An indicator variable coded 1 if the firm’s ROA was negative; coded 0 otherwise.
SPECIAL_ITEM An indicator variable coded 1 if the firm reports special items (Compustat variable SPI).
Lagged lnAUDIT_FEE The natural log of prior year’s audit fee in millions of dollars.
This appendix presents the results from estimating the audit fee model from Ashbaugh, LaFond, and Mayhew (2003) for the period 2002-2013 for the full population of firms with data available in the Audit Analytics and Compustat databases. In addition to including all variables from the Ashbaugh et al. model, we include lagged audit fees as a control for firm-specific factors that could affect current year audit fees. The residual values obtained from estimating the model represent our abnormal audit fee measure, ABN_AFEE. The dependent variable, lnAUDIT_FEE, is the natural logarithm of audit fees. All variables are defined in detail in Appendix A. All continuous variables have been winsorized at the 1st and 99th percentiles.
36
TABLE 1 Sample Selection
N for Dependent Variable:
GUIDANCE_ERROR CHG_CFO GW_IMPAIRMENT (1) (2) (3) M&A data from SDC: completed deals involving US public acquirers and
US targets (both public and private), between 2002 and 2013a, with non-missing acquirer identifiers (necessary for merger with Compustat and Audit Analytics databases), and where relative deal size exceeds 5%.
5,725 5,725 5,725
Less:
M&A transactions with missing dependent variable data (3,989) (2,426) 0 M&A transactions with missing control variable data (346) (423) (1,462)
Subtotal 1,390 2,876 4,263
Observations with non-missing ABN_AFEE test variable 1,314 2,681 3,976 Observations with non-missing ABN_AUDRPT_LAG test variable 1,277 2,659 3,942 Observations with non-missing INTEGRATION_ISSUES test variable 1,227 2,538 3,760
This table summarizes our sample selection process. One column is presented for each dependent variable – GUIDANCE_ERROR, CHG_CFO, and GW_IMPAIRMENT – used in our analyses. GUIDANCE_ERROR is the absolute difference between management EPS guidance and actual EPS (measured in the first full fiscal year following the deal), scaled by share price one month prior to fiscal year-end (Rogers and Stocken 2005; Feng, Li, and McVay 2009). CHG_CFO measures change in industry-adjusted acquirer cash flows (CFO) between pre- and post-deal periods, where CFO is calculated as cash flow from operations divided by average total assets. CHG_CFO is the difference between the average CFO for the post-deal periods, t+1 through t+3, and the pre-deal periods, t-3 through t-1, where period t is a year in which the acquisition was completed. If CFO data are unavailable for periods t+3 or t-3, average CFO is calculated using two (instead of three) years of data in the pre- and post-deal periods. GW_IMPAIRMENT is an indicator variable coded 1 if the acquirer recorded a goodwill impairment in the year of the deal (period t) or in the post-deal period (periods t+1 through t+3); coded zero otherwise. Finally, we use three test variables in our analyses: ABN_AFEE and ABN_AUDRPT_LAG. ABN_AFEE is the acquirer’s abnormal audit fees, measured in the year of the deal using the audit fee model from Ashbaugh, LaFond, and Mayhew (2003). Second, ABN_AUDRPT_LAG (abnormal audit report lag) is audit report lag measured for the year of the deal less average audit report lag in the three years preceding the deal, where audit report lag is defined as the number days between fiscal year-end and the audit opinion signature date (Bamber Bamber, and Schoderbek 1993; Knechel and Payne 2001; Krishnan and Yang 2009). Third, INTEGRATION_ISSUES, aggregates ABN_AFEE and ABN_AUDRPT_LAG by transforming each into decile ranks (0 to 9) and dividing by 9. The two transformed values for ABN_AFEE and ABN_AUDRPT_LAG (each ranging in value from 0 to 1) are then added, resulting in an aggregate measure of integration issues (ranging in value from 0 to 2). All variables, including control variables, are defined in detail in Appendix A. a Our analyses cover the years 2002 to 2016. However, we restrict our sample to M&A deals completed between 2002 and 2013 so that we have sufficient data to measure our dependent variables in the post-deal period (i.e., after 2013).
***, **, * indicate significance at the 0.01, 0.05, 0.10 levels, respectively (two-tail). In Panel A, we present descriptive statistics for our dependent and test variables. GUIDANCE_ERROR is the absolute difference between management EPS guidance and actual EPS (measured in the first full fiscal year following the deal), scaled by share price one month prior to fiscal year-end (Rogers and Stocken 2005; Feng, Li, and McVay 2009). Alternatively, we use the variable HIGH_GUIDANCE_ERROR, which is an indicator variable coded 1 if GUIDANCE_ERROR is ≥ sample median GUIDANCE_ERROR and coded 0 otherwise. We use three test variables in our analyses. First, ABN_AFEE is the acquirer’s abnormal audit fees, measured in the year of the deal using the audit fee model from Ashbaugh, LaFond, and Mayhew (2003). Second, ABN_AUDRPT_LAG (abnormal audit report lag) is audit report lag measured for the year of the deal less average audit report lag in the three years preceding the deal, where audit report lag is defined as the number days between fiscal year-end and the audit opinion signature date (Bamber Bamber, and Schoderbek 1993; Knechel and Payne 2001; Krishnan and Yang 2009). Third, INTEGRATION_ISSUES, aggregates ABN_AFEE and ABN_AUDRPT_LAG by transforming each into decile ranks (0 to 9) and dividing by 9. The two transformed values for ABN_AFEE and ABN_AUDRPT_LAG (each ranging in value from 0 to 1) are then added, resulting in an aggregate measure of integration issues (ranging in value from 0 to 2). All variables, including control variables, are defined in detail in Appendix A. All continuous variables have been winsorized at the 1st and 99th percentiles. In Panel B, we present t-tests for differences in mean values between HIGH_GUIDANCE_ERROR = 0 and HIGH_GUIDANCE_ERROR = 0 subsamples. Finally, in Panel C we present a Pearson correlation matrix. Correlations where p-value ≤ 0.10 are listed in bold font. a N=1,227; p-value≤0.01.
39
TABLE 3 The Impact of Accounting Integration Issues on Management Guidance Error
(1.17) (1.28) (1.32) (1.19) (0.90) (1.23) Industry Fixed Effects Included Included Included Included Included Included Year Fixed Effects Included Included Included Included Included Included Adjusted/Pseudo-R2 0.113 0.130 0.131 0.229 0.263 0.265 N 1,314 1,277 1,227 1,314 1,277 1,227
Table continues on next page.
40
TABLE 3 (continued) The Impact of Integration Issues on Management Guidance Error
This table presents the results from estimating the following models:
GUIDANCE_ERROR = α0 + α1* Integration Issue Measure + αi*Controli + Industry & Year Fixed Effects + ε
(1a)
Pr(HIGH_GUIDANCE_ERROR = 1) = F{β0 + β1* Integration Issue Measure + βi*Controli + Industry & Year Fixed Effects}
(1b)
Results presented in columns 1 and 2 were obtained, using model 1a, from OLS regressions (heteroscedasticity robust standard errors). Results presented in columns 3 and 4 were obtained, using model 1b, from logistic regressions. The samples used are summarized in Table 1. The dependent variable GUIDANCE_ERROR is the absolute difference between management EPS guidance and actual EPS (measured in the first full fiscal year following the deal), scaled by share price one month prior to fiscal year-end (Rogers and Stocken 2005; Feng, Li, and McVay 2009). Alternatively, in columns 3 and 4 we use the variable HIGH_GUIDANCE_ERROR, which is an indicator variable coded 1 if GUIDANCE_ERROR is ≥ sample median GUIDANCE_ERROR and coded 0 otherwise. We include three, alternative Integration Issue Measures. First, ABN_AFEE is the acquirer’s abnormal audit fees, measured in the year of the deal using the audit fee model from Ashbaugh, LaFond, and Mayhew (2003). Second, ABN_AUDRPT_LAG (abnormal audit report lag) is audit report lag measured for the year of the deal less average audit report lag in the three years preceding the deal, where audit report lag is defined as the number days between fiscal year-end and the audit opinion signature date (Bamber Bamber, and Schoderbek 1993; Knechel and Payne 2001; Krishnan and Yang 2009). Third, INTEGRATION_ISSUES, aggregates ABN_AFEE and ABN_AUDRPT_LAG by transforming each into decile ranks (0 to 9) and dividing by 9. The two transformed values for ABN_AFEE and ABN_AUDRPT_LAG (each ranging in value from 0 to 1) are then added, resulting in an aggregate measure of integration issues (ranging in value from 0 to 2). All variables, including control variables, are defined in detail in Appendix A. All continuous variables have been winsorized at the 1st and 99th percentiles. ***, **, and * indicate significance at the 0.01, 0.05, and 0.10 levels, respectively (two-tail).
41
TABLE 4 Post-M&A Outcomes: Descriptive Statistics and Univariate Comparisons
Panel A: Descriptive Statistics for Dependent and Control Variables N Mean STD P25 Median P75 CHG_CFO 2,681 -0.015 0.083 -0.051 -0.008 0.021 ALM_ABN_AFEE 2,681 0.041 0.390 -0.168 0.012 0.237 ABN_AUDRPT_LAG 2,659 4.498 25.724 -4.333 1.333 10.000 INTEGRATION_ISSUES 2,538 1.000 0.470 0.667 1.000 1.333
TABLE 4 (continued) Post-M&A Outcomes: Descriptive Statistics and Univariate Comparisons
Panel A provides descriptive statistics for the dependent and independent test variables used in our post-M&A outcomes analyses. Panel B provides univariate comparisons of dependent variables, CHG_CFO and GW_IMPAIRMENT, across quintiles of our independent test variables, ABN_AFEE, ABN_AUDRPT_LAG, and INTEGRATION_ISSUES. Finally, Panel C provides descriptive statistics for our control variables. The dependent variable CHG_CFO measures change in industry-adjusted acquirer cash flows (CFO) between pre- and post-deal periods, where CFO is calculated as cash flow from operations divided by average total assets. CHG_CFO is the difference between the average CFO for the post-deal periods, t+1 through t+3, and the pre-deal periods, t-3 through t-1, where period t is a year in which the acquisition was completed. If CFO data are unavailable for periods t+3 or t-3, average CFO is calculated using two (instead of three) years of data in the pre- and post-deal periods. GW_IMPAIRMENT is an indicator variable coded 1 if the acquirer recorded a goodwill impairment in the year of the deal (period t) or in the post-deal period (periods t+1 through t+3); coded zero otherwise. We use three test variables in our analyses. ABN_AFEE is the acquirer’s abnormal audit fees, measured in the year of the deal using the audit fee model from Ashbaugh, LaFond, and Mayhew (2003). Second, ABN_AUDRPT_LAG (abnormal audit report lag) is audit report lag measured for the year of the deal less average audit report lag in the three years preceding the deal, where audit report lag is defined as the number days between fiscal year-end and the audit opinion signature date (Bamber Bamber, and Schoderbek 1993; Knechel and Payne 2001; Krishnan and Yang 2009). Third, INTEGRATION_ ISSUES, aggregates ABN_AFEE and ABN_AUDRPT_LAG by transforming each into decile ranks (0 to 9) and dividing by 9. The two transformed values for ABN_AFEE and ABN_AUDRPT_LAG (each ranging in value from 0 to 1) are then added, resulting in an aggregate measure of integration issues (ranging in value from 0 to 2). All variables, including control variables, are defined in detail in Appendix A. All continuous variables have been winsorized at the 1st and 99th percentiles.
44
TABLE 5 The Impact of Integration Issues on Post-M&A Outcomes
(-1.89) (-1.24) (-1.64) (-0.74) (-0.90) (-0.94) Industry Fixed Effects Included Included Included Included Included Included Year Fixed Effects Included Included Included Included Included Included Adjusted/Pseudo-R2 0.152 0.131 0.142 0.093 0.095 0.100 N 2,681 2,659 2,538 3,976 3,942 3,760 N (GW Impairment) 1,379 1,374 1,318
Table continues on next page.
45
TABLE 5 (continued)
The Impact of Integration Issues on Post-M&A Outcomes This table presents the results from estimating the following models:
CHG_CFO = γ0 + γ1* Integration Issue Measure + γi*Controli + Industry & Year Fixed Effects + θ (2)
Pr(GW_IMPAIRMENT = 1) = F{δ0 + δ1* Integration Issue Measure + δi*Controli + Industry & Year Fixed Effects}
(3)
Results presented in columns 1 and 2 were obtained, using model 1a, from OLS regressions (heteroscedasticity robust standard errors). Results presented in columns 3 and 4 were obtained, using model 1b, from logistic regressions. The samples used are summarized in Table 1. The dependent variable CHG_CFO measures change in industry-adjusted acquirer cash flows (CFO) between pre- and post-deal periods, where CFO is calculated as cash flow from operations divided by average total assets. CHG_CFO is the difference between the average CFO for the post-deal periods, t+1 through t+3, and the pre-deal periods, t-3 through t-1, where period t is a year in which the acquisition was completed. If CFO data are unavailable for periods t+3 or t-3, average CFO is calculated using two (instead of three) years of data in the pre- and post-deal periods. GW_IMPAIRMENT is an indicator variable coded 1 if the acquirer recorded a goodwill impairment in the year of the deal (period t) or in the post-deal period (periods t+1 through t+3); coded zero otherwise. We include three, alternative Integration Issue Measures. First, ABN_AFEE is the acquirer’s abnormal audit fees, measured in the year of the deal using the audit fee model from Ashbaugh, LaFond, and Mayhew (2003). Second, ABN_AUDRPT_LAG (abnormal audit report lag) is audit report lag measured for the year of the deal less average audit report lag in the three years preceding the deal, where audit report lag is defined as the number days between fiscal year-end and the audit opinion signature date (Bamber Bamber, and Schoderbek 1993; Knechel and Payne 2001; Krishnan and Yang 2009). Third, INTEGRATION_ISSUES, aggregates ABN_AFEE and ABN_AUDRPT_LAG by transforming each into decile ranks (0 to 9) and dividing by 9. The two transformed values for ABN_AFEE and ABN_AUDRPT_LAG (each ranging in value from 0 to 1) are then added, resulting in an aggregate measure of integration issues (ranging in value from 0 to 2). All variables, including control variables, are defined in detail in Appendix A. All continuous variables have been winsorized at the 1st and 99th percentiles. ***, **, and * indicate significance at the 0.01, 0.05, and 0.10 levels, respectively (two-tail).
(1.54) (1.31) (0.83) (0.74) Controls Included Included Included Included Industry Fixed Effects Included Included Included Included Year Fixed Effects Included Included Included Included Adjusted/Pseudo-R2 0.117 0.131 0.231 0.264 N 1,314 1,277 1,314 1,277
Panel B: the Impact of Integration Issues on Post-M&A Outcomes Dependent Variable = CHG_CFO GW_IMPAIRMENT (1) (2) (3) (4) Coefficient (t-statistic) Intercept -0.034** -0.028* -1.459*** -1.327***
Controls Included Included Included Included Industry Fixed Effects Included Included Included Included Year Fixed Effects Included Included Included Included Adjusted/Pseudo-R2 0.1520 0.130 0.101 0.103 N 2,681 2,659 3,976 3,942 N (Goodwill Impairment) 1,379 1,374
This table tests whether an acquirer’s (a) prior integration experience, or (b) recent acquisition intensity are associated with management guidance error (Panel A) and post-M&A outcomes (Panel B). In Panel A, models (1a) and (1b) were modified to incorporate measures of prior integration experience and recent acquisition intensity:
Results presented in columns 1 and 2 were obtained, using model 1a, from OLS regressions (heteroscedasticity robust standard errors). Results presented in columns 3 and 4 were obtained, using model 1b, from logistic regressions. The samples used are summarized in Table 1. In Panel B, models (2) and (3) were modified to incorporate measures of prior integration experience and recent acquisition intensity:
In both Panels A and B, we include two Prior Integration Issue Measures. First, Prior Deals ABN_AFEE is the average of the acquirer’s abnormal audit fees for all deals executed in the three-year period preceding the current acquisition. Second, Prior Deals ABN_AUDRPT_LAG is the average of the acquirers abnormal audit report lags for all deals executed in the three-year period preceding the current acquisition. All control variables are included in the estimation of models (1a) and (1b) but are not presented here in the interest of brevity. All variables are defined in Appendix A and in previous tables. All continuous variables have been winsorized at the 1st and 99th percentiles. ***, **, and * indicate significance at the 0.01, 0.05, and 0.10 levels, respectively (two-tail).
48
TABLE 7 Additional Analyses: Do Announcement Period Returns Predict Integration Issues?
This table tests whether announcement period returns (MM_RET3) can predict integration issues as measured by ABN_AFEE and ABN_AUDRPT_LAG. Results in all columns were obtained by estimating OLS models with heteroscedasticity robust standard errors. Consistent with our main sample, samples here were restricted to observations (a) falling in the years 2002 and later, (b) with relative deal sizes ≥ 5%, and (c) non-missing dependent and independent variables. The results above suggest that announcement period returns have little explanatory power in predicting post-merger integration issues. All variables are defined, in detail, in previous tables and in the Appendix. All continuous variables have been winsorized at the 1st and 99th percentiles. ***, **, and * indicate significance at the 0.01, 0.05, and 0.10 levels, respectively (two-tail).
49
TABLE 8 Falsification Tests: Test Variables Measured in the Pre-M&A Period
(0.51) (0.74) Controls Included Included Included Included Industry Fixed Effects Included Included Included Included Year Fixed Effects Included Included Included Included Adjusted/Pseudo-R2 0.118 0.118 0.254 0.258 N 1,216 1,128 1,216 1,128
(-0.52) (-0.51) Controls Included Included Included Included Industry Fixed Effects Included Included Included Included Year Fixed Effects Included Included Included Included Adjusted/Pseudo-R2 0.125 0.117 0.093 0.093 N 2,506 2,370 3,723 3,518 N (GW_IMPAIRMENT) 1,311 1,246
This table presents the results of falsification tests in which the independent test variables, ABN_AFEE and ABN_AUDRPT_LAG, were measured one year prior to the M&A deal (i.e., period t-1, where period t is the fiscal year during which the M&A deal was completed). Panel A presents the results from estimating models 1a and 1b (our management guidance error models) using pre-M&A test variables. Panel B presents the results from estimating models 2 and 3 (our change in cash flow and goodwill impairment models). All variables are defined, in detail, in previous tables and in the Appendix. All continuous variables have been winsorized at the 1st and 99th percentiles. ***, **, and * indicate significance at the 0.01, 0.05, and 0.10 levels, respectively (two-tail).
50
Figure 1 Path Analysis: HIGH_GUIDANCE_ERROR as Mediator Variable
Panel A: Changes in Cash Flows as Outcome Variable (N=872)
Figure continues on next page.
INTEGRATION_ ISSUES
HIGH_ GUIDANCE_ERROR
CHG_CFO
+0.530 (p-value≤0.01)
-0.008 (p-value=0.07)
Direct Effect: -0.065
(Confidence Interval: -0.120 to +0.007)
Indirect Effect through HIGH_GUIDANCE_ERROR: -0.008
(Confidence Interval: -0.020 to -0.001)
51
Figure 1 (continued) Path Analysis: HIGH_GUIDANCE_ERROR as Mediator Variable
Panel B: Goodwill Impairments as Outcome Variable (N=1,227) Figure 1 presents the results of two path analyses. Panel A uses CHG_CFO as the outcome variable, while Panel B uses GW_IMPAIRMENT. In these analyses we are interested in testing whether post-merger accounting integration issues affect post-merger outcomes through the quality of information produced from the newly merged firm’s accounting system. In other words, we are testing whether HIGH_GUIDANCE_ERROR is a mediator variable in the relationship between INTEGRATION_ISSUES and our two post-merger outcome variables, CHG_CFO and GW_IMPAIRMENT. Using Stata (package “binary_mediator”), we ran path analyses in which the following models were estimated:
ACQ_ROA, ACQ_LEV, ACQ_BTM, ACQ_STD_RET, DL_REL_SIZE, DL_STOCK, DL_DIVER, DL_DUE_DIL, Industry & Year Fixed Effects)
CHG_CFO or GW_IMPAIRMENT = f(HIGH_GUIDANCE_ERROR, INTEGRATION_ISSUES, ACQ_ICW, ABS_PADACC, PUBLIC_MA, BIG4,
MM_RET3, ACQ_SIZE, ACQ_ROA, ACQ_LEV, ACQ_BTM, ACQ_STD_RET, DL_REL_SIZE, DL_STOCK, DL_DIVER, DL_DUE_DIL, Industry & Year Fixed Effects)
Within the larger path model, when either HIGH_GUIDANCE_ERROR or GW_IMPAIRMENT are used as the dependent variable, a logit model is estimated. When CHG_CFO is used as the dependent variable, an OLS model is estimated. The path model estimates both the direct and indirect effects of post-merger accounting integration issues on post-merger outcomes and tests each for significance (bootstrapped standard errors based on 1,000 repetitions are used to calculate bias-corrected confidence intervals). In the bottom portion of each path diagram above, the estimated direct and indirect effects and their associated bootstrapped bias-corrected confidence intervals (90 percent) are reported. Similar results are obtained when we use Stata’s “sem” command. However, we do not report results based on this command because we use a binary mediator variable (which requires the estimation of a logit model as part of the path model) and the “sem” command treats all mediator variables as linear (output is based upon OLS estimates). Finally, our results remain significant when we allow for interaction effects between our test variable (INTEGRATION_ISSUES) and our mediator variable (HIGH_GUIDANCE_ERROR) by implementing the Stata package “paramed” (based on Valeri and VanderWeele 2013). All significant path coefficients are reported in bold font. All variables are defined in Appendix A and in prior tables.