Abstract
This paper provides evidence on how the adoption of Statement of Financial Accounting Standards (SFAS) No. 146, Accounting for Costs Associated with Exit or Disposal Activities (SFAS 146), has changed (1) restructuring-related earnings management, (2) the effect of restructurings on future operating performance, and (3) stock price responses to restructuring announcements. I find that restructuring charges reported after SFAS 146 have a weaker association with smoothing behavior. More importantly, I document that an improvement in future earnings following a normal restructuring charge is attenuated under SFAS 146, consistent with the standard’s adoption resulting in smaller and more frequent (i.e., more persistent) restructuring charges. I further document that the market places a higher valuation multiple on restructuring charges in the post-SFAS 146 regime, compared to the pre-SFAS 146 regime. Therefore the overall results suggest that the market appears to understand the higher persistence of restructuring charges reported under SFAS 146 and values the charges accordingly.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
For brevity, throughout this paper I often use the term “accuracy” when referring to the concept of representational faithfulness, which is discussed in Statement of Financial Accounting Concepts (SFAC) No. 2.
In this paper, however, I do not contrast the profitability characteristics of investments before and after the adoption of SFAS 146 and leave it as an avenue for future research.
Concerning the accuracy aspect, to the extent that SFAS 146 has been effective in reducing big bath or smoothing behavior, the accuracy of the abnormal charge, rather than that of the normal charge, is expected to improve under the standard. Concerning the persistence and timing aspect, suppose that management initiates a restructuring plan to close plants and terminate employees over the next 2 years, which is expected to cost the company $10 million. Prior to SFAS 146, management would estimate the cost and accrue the charge at the time of commitment to the plan. Under SFAS 146, management must identify the specific plants that will be closed and the specific employee group that will be laid off and can only accrue a charge where these specific identifications have been made. Assuming further that $5 million occurs in the first year, and $5 million in the second, the $10 million charge that would have been recorded in a year prior to the standard will now get split into two equal charges across the years. This change in the accounting rule is more likely to affect the persistence and timing of the normal charge than that of the abnormal charge.
Note that I do not draw a conclusion on restructuring-related earnings management based on a comparison of the magnitudes of abnormal restructuring costs before and after SFAS 146. A smaller magnitude of abnormal costs may result from a decrease in managers’ unintentional measurement errors and thus does not necessarily imply a decline in earnings management.
See Sect. 5.5 for details.
Restructuring-related liabilities and expenses include costs incurred for the following (Moehrle 2002): (1) employee benefits such as severance and termination benefits, (2) elimination and reduction of product lines, (3) consolidation or relocation of plant facilities, (4) new systems development or acquisition, and (5) retraining employees to use newly developed systems.
For example, SFAS 146 does not apply to the termination benefits provided to current employees who are involuntarily terminated under the terms of an ongoing benefit arrangement. Accordingly, in this case, firms may need to look to SFAS No. 112, Employers’ Accounting for Postemployment Benefits—an amendment of FASB Statements No. 5 and 43 (SFAS 112), which applies when there is a plan in place that addresses the payout employees would be entitled to at the time of an involuntary termination. Contrary to SFAS 146, however, SFAS 112 requires that postemployment benefits be recognized when it is probable that a liability has been incurred and the amount can be reasonably estimated (FASB 1992).
To elaborate on this, assuming that a given level of normal and abnormal charges (say, $1 each) is expected to increase future earnings by the same amount (say, by $2), the $2 increase in future earnings following the normal charge is likely to have a wider confidence interval (or a higher standard error) than the same increase in future earnings following the abnormal charge. This is because a normal charge is assumed to increase future earnings only if there is a fundamental improvement in the firm’s underlying performance, which may be difficult to be realized. In contrast, an abnormal charge is very likely to reverse and, as a result, mechanically increase future earnings.
A number of studies suggest that the market reacts positively to restructuring announcements (e.g., Brickley and Van Drunen 1990; Francis et al. 1996; Denis and Kruse 2000; Chalos and Chen 2002). On the other hand, Bartov et al. (1998) point out that the market reaction to restructurings is quite small even in studies where the results are statistically significant. Moreover, Bens (2002) reports negative stock price reactions to restructuring announcements, and Elayan et al. (1998) document negative price reactions to employee downsizing announcements.
Following Cready et al. (2010), I use after-tax restructuring costs throughout this paper. I also repeat all of my analyses using pre-tax restructuring costs and obtain similar results. While this is somewhat inconsistent with Cready et al.’s (2011) finding that before-tax implications of (quarterly) negative special items for future before-tax earnings differ substantially from after-tax implications for after-tax earnings, using annual numbers and focusing on a particular special item (i.e., restructuring costs) as in my paper might have reduced such differences in earnings (and stock) implications between before-tax and after-tax special items.
Since my sample includes all fiscal year-ends, I calculate for each fiscal year the annual percentage change in real U.S. GDP using monthly real GDP data. U.S. GDP data is obtained from the Federal Reserve Bank of St. Louis website: http://research.stlouisfed.org/fred2/.
Untabulated results indicate that INV_TOt−1, PPE_TOt−1, and SALE_EMPt−1 all exhibit autocorrelations of around 0.9. AR_TOt−1 and PMt−1 have autocorrelations of approximately 0.7.
Following Bens and Johnston (2009), I broaden the definition of industry to analyze big bath and smoothing behaviors. However, I assume that firms with changes in pre-restructuring earnings (scaled by the prior year’s sales) that are less than the first quartile of negative values (not all values as in Bens and Johnston) have an incentive to take a big bath. Similarly, firms with values for this variable that are higher than the third quartile of positive values are assumed to have a smoothing incentive. This definition is used in Riedl (2004) and strictly ensures that firms with sufficiently poor (good) accounting performance have a big bath (smoothing) incentive.
Unlike Bens and Johnston (2009), however, I exclude the proportion of the restructuring charge that is subsequently reversed and recorded as a gain over the three years following the charge. My reasoning is that, for example, a reversal in 2005 may come from a charge reported in 2002 (pre-SFAS 146 period) or a charge reported in 2004 (post-SFAS 146 period), resulting in an intermingling of regimes within a single observation. In addition, I do not include the percentage of insiders as identified by the Corporate Library database because doing so reduces my sample size by about 48 %, but the coefficient on this variable is insignificant.
I use changes (instead of levels) in earnings as the dependent variable because the changes specification correctly identifies transitory income components and is less likely to be affected by survivor biases, as discussed by Ball and Shivakumar (2005).
Note that the primary objective of this paper is to contrast the consequences of restructuring costs reported before and after SFAS 146, i.e., the relative associations across these two regimes. This differs from an analysis of absolute associations within each regime. Thus it is less crucial to establish an ex ante relationship between restructuring costs and future performance (and, by extension, market response).
Compustat began providing special items by type in 2000, but I focus on the 2001–2002 and 2004–2006 periods for the following reasons: first, observations in fiscal year 2000 are deleted because there are far fewer observations than in the following years. Second, I exclude observations in fiscal year 2003, which can be viewed as a transition period. Since SFAS 146 is effective for restructuring activities that are initiated after December 31, 2002, if firms with non-December fiscal year-ends initiated restructuring activities before the effective date during fiscal year 2003, then it is unclear whether restructuring charges incurred during that period were subject to SFAS 146. Third, observations in fiscal years 2007 and after are also deleted because 3-year-ahead earnings are generally unavailable for these observations at the time this paper was being written.
In December 2007, the FASB issued SFAS 141(R), Business Combinations, which requires the restructuring costs that the acquirer expected but was not obligated to incur to be expensed after the merger.
An interesting fact about this sample composition is that the annual frequency of reported restructuring costs has not increased after the adoption of SFAS 146. However, delaying recognition of part of a contemplated charge until it is incurred may result in a decline in the average normal restructuring costs as well as an increase in the frequency of reported restructurings. While the next section provides some evidence consistent with the former, I do not find evidence supporting the latter. This is probably because some firms had already recognized excessive restructuring costs in 2001–2002 in anticipation of the standard. Another possibility is that the overall economic conditions were worse prior to SFAS 146. For example, according to the National Bureau of Economic Research (NBER) data, the period from March 2001 to November 2001 is defined as a recession, whereas the 2004–2006 (post-SFAS 146) period is not. See http://www.nber.org/cycles/cyclesmain.html.
I require restructuring firms to have data for three years after the year with the restructuring charge to be included in my analyses. A 3-year window is commonly used in the literature, which documents that the benefits or detriments of a restructuring materialize, on average, after 3 years. In addition, several independent variables used in the test of earnings management require three-year-ahead data. Therefore, despite a potential survivorship bias, I impose this data requirement.
Throughout the paper, all continuous variables are winsorized at the top and bottom 1%. Following Bens and Johnston (2009), however, I winsorize AR_TOt−1 and INV_TOt−1 at 24. In the estimation of normal restructuring costs, I also winsorize these variables at the 99th percentile and obtain similar results.
I also use the condition index to check for multicollinearity and find that the largest one is less than 10. Thus my tobit model does not suffer from a collinearity problem based on a common rule of thumb that a condition index over 15 indicates a possible collinearity problem.
Dhrymes R2 is the squared correlation between the predicted and actual values for observations with nonzero values of RESTt and is commonly reported in the literature (e.g., Bens and Johnston 2009). Due to the variation in sample size across industries, I provide in the table a weighted-average Dhrymes R2 with the weight being determined by the sample size for the industry.
This methodology allows for cross-sectional correlation within a given year and industry. This adjustment is necessary because the overall magnitude of restructuring costs is likely to be affected by these two dimensions. However, I do not adjust t-statistics for serial correlation because a firm has at most two or three observations in each regime.
From a regression of seasonally differenced earnings in quarter t + 4 on negative special items in quarter t and the remaining component of earnings seasonally differenced with quarter t − 4, Burgstahler et al. (2002) report that the coefficient on negative special items is -1.277, which would be 1.277 if special items were multiplied by −1. While I use annual numbers and focus on restructuring costs, the corresponding coefficient for the pre-SFAS 146 period in my regression (reported in Table 6, Panel A) is 3.126, which appears to be substantially higher than 1.277. Relatedly, the coefficients on ∆Et and ∆Et*D∆Et in the regression are 0.368 and −1.328, respectively, suggesting that both positive and negative changes in earnings reported in the pre-SFAS 146 period are followed by an increase in future earnings. These somewhat abnormal results are probably due to the fact that firms generally reported higher earnings following the end of the recession of the early 2000s. As discussed earlier, however, this would bias against finding the results I expect.
References
Atiase, R., Platt, D., & Tse, S. (2004). Operational restructuring charges and post-restructuring performance. Contemporary Accounting Research, 21(3), 493–522.
Ball, R., & Shivakumar, L. (2005). Earnings quality in U.K. private firms: Comparative loss recognition timeliness. Journal of Accounting and Economics, 39(1), 83–128.
Balsam, S., Reitenga, A., & Sanchez, J. (2007). Corporate restructuring quality and CEO compensation. Working paper, Temple University and University of Arkansas.
Bartov, E., Lindahl, F., & Ricks, W. (1998). Stock price behavior around announcements of write-offs. Review of Accounting Studies, 3(4), 327–346.
Bens, D. (2002). The determinants of the amount of information disclosed about corporate restructurings. Journal of Accounting Research, 40(1), 1–20.
Bens, D., & Johnston, R. (2009). Accounting discretion: Use or abuse? An analysis of restructuring charges surrounding regulator action. Contemporary Accounting Research, 26(3), 673–699.
Bhojraj, S., Sengupta, P., & Zhang, S. (2009). Restructuring charges, regulatory changes and the accruals anomaly. Working paper, Cornell University and George Mason University.
Brickley, J., & Van Drunen, L. (1990). Internal corporate restructuring: An empirical analysis. Journal of Accounting and Economics, 12(1–3), 251–280.
Burgstahler, D., Jiambalvo, J., & Shevlin, T. (2002). Do stock prices fully reflect the implications of special items for future earnings? Journal of Accounting Research, 40(3), 585–612.
Chalos, P., & Chen, C. (2002). Employee downsizing strategies: Market reaction and post announcement financial performance. Journal of Business Finance & Accounting, 29(5/6), 847–870.
Cohen, D., Dey, A., & Lys, T. (2008). Real and accrual-based earnings management in the pre- and post-Sarbanes-Oxley periods. The Accounting Review, 83(3), 757–787.
Cready, W., Lopez, T., & Sisneros, C. (2010). The persistence and market valuation of recurring nonrecurring items. The Accounting Review, 85(5), 1577–1615.
Cready, W., Lopez, T., & Sisneros, C. (2011). Negative special items and future earnings: Expense transfer or real improvements? Working paper, University of Texas at Dallas.
Dechow, P., & Ge, W. (2006). The persistence of earnings and cash flows and the role of special items: Implications for the accrual anomaly. Review of Accounting Studies, 11(2/3), 253–296.
Denis, D., & Kruse, T. (2000). Managerial discipline and corporate restructuring following performance declines. Journal of Financial Economics, 55(3), 391–424.
Elayan, F., Swales, G., Maris, B., & Scott, J. (1998). Market reactions, characteristics and the effectiveness of corporate layoffs. Journal of Business Finance & Accounting, 25(3/4), 329–351.
Elliott, J., & Hanna, J. (1996). Repeated accounting write-offs and the information content of earnings. Journal of Accounting Research, 34(Supplement), 135–155.
Financial Accounting Standards Board (FASB). (1992). Employers’ accounting for postemployment benefits—An amendment of FASB Statements No. 5 and 43. Statement of Financial Accounting Standards No. 112. Norwalk, CT: FASB.
Financial Accounting Standards Board (FASB). (1995a). Accounting for the impairment of long-lived assets and for long-lived assets to be disposed of. Statement of Financial Accounting Standards No. 121. Norwalk, CT: FASB.
Financial Accounting Standards Board (FASB). (1995b). Emerging Issues Task Force minutes issues summaries. January. Norwalk, CT: FASB.
Financial Accounting Standards Board (FASB). (2001). Accounting for the impairment or disposal of long-lived assets. Statement of Financial Accounting Standards No. 144. Norwalk, CT: FASB.
Financial Accounting Standards Board (FASB). (2002). Accounting for costs associated with exit or disposal activities. Statement of Financial Accounting Standards No. 146. Norwalk, CT: FASB.
Francis, J., Hanna, J., & Vincent, L. (1996). Causes and effects of discretionary asset write-offs. Journal of Accounting Research, 34(Supplement), 117–134.
Kolev, K., Marquardt, C., & McVay, S. (2008). SEC scrutiny and the evolution of non-GAAP reporting. The Accounting Review, 83(1), 157–184.
Moehrle, S. (2002). Do firms use restructuring charge reversals to meet earnings targets? The Accounting Review, 77(2), 397–413.
Penman, S. (2007). Financial statement analysis and security valuation (3rd ed.). New York: The McGraw-Hill Companies.
Riedl, E. (2004). An examination of long-lived asset impairments. The Accounting Review, 79(3), 823–852.
Skinner, D., & Sloan, R. (2002). Earnings surprises, growth expectations, and stock returns or don’t let an earnings torpedo sink your portfolio. Review of Accounting Studies, 7(2/3), 289–312.
Strong, N., & Walker, M. (1993). The explanatory power of earnings for stock returns. The Accounting Review, 68(2), 385–399.
Watts, R., & Zimmerman, J. (1986). Positive accounting theory. Englewood Cliffs, NJ: Prentice Hall.
Acknowledgments
I thank my committee members for their guidance: Bjorn Jorgensen (sponsor), Partha Mohanram, Stephen Penman (chair), Mary Billings, and Amit Khandelwal. Helpful comments and suggestions have also been provided by Divya Anantharaman, Sid Balachandran, Masako Darrough, Michael Kirschenheiter, Steve Rock, Gil Sadka, Kenton Yee, Julian Yeo, Yuan Zhang, and seminar participants at Baruch College of the City University of New York, Chinese University of Hong Kong, George Mason University, Hong Kong Polytechnic University, Hong Kong University of Science and Technology, Korea University, National University of Singapore, Singapore Management University, Temple University, University of Alberta, University of Illinois at Chicago, University of Toronto, University of Waterloo, the 2008 AAA annual meeting, and the 2009 AAA FARS meeting. I thank Ethan Kinory for research assistance. This paper was supported by Faculty Research Fund, Sungkyunkwan University, 2013.
Author information
Authors and Affiliations
Corresponding author
Additional information
This paper is based on my dissertation completed at Columbia University.
Rights and permissions
About this article
Cite this article
Lee, Y.G. An examination of restructuring charges surrounding the implementation of SFAS 146. Rev Account Stud 19, 539–572 (2014). https://doi.org/10.1007/s11142-013-9260-9
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11142-013-9260-9