October 2006 pp. 59-95
Determinants and Market Reaction of
Assets Impairment in Taiwan
∗
Wan-Ting Hsieh
Kun Shan University of Technology
TsingZai C. Wu
†National Cheng Kung University
ABSTRACT: On July 1, 2004, Taiwan’s Financial Accounting Standards Committee of the Accounting Research and Development Foundation issued SFAS No. 35, Accounting for the Impairment of Assets. This accounting standard came effective in the financial year ending after December 31, 2005, with early application encouraged. Taiwan’s SFAS No. 35 provides for accounting for assets impairment conservatively by recognizing impairment loss to reflect the market value of long-lived assets but not the unrealized gains of assets. SFAS No. 35 will affect accounting earnings, carrying amounts of long-lived assets, and, supposedly, the stock prices for listed firms in Taiwan. This paper examines determinants of assets impairment and its market reaction. First, we examine the determinants of early adoption of Taiwan’s SFAS No. 35 for firms listed in the
Taiwan Stock Exchange and the GreTai Securities Market. We investigate the
motivation of adoption of assets impairment accounting in reporting purposes and operational purposes per literature. Our empirical results show that determinants for early adopters of Taiwan’s SFAS No. 35 are the taking a “big bath” purpose (the reporting motivation) and factors reflecting the accrual-based and cashflow-based recoverability of long-lived assets, such as changes in sales and changes in operational cash flows (operational motivations). Secondly, we examine factors affecting the “amounts” of assets impairment in both reporting and operational perspectives. Our empirical results show the following: (1) For early adopters, the amounts of assets impairments are associated with only reporting motivations (the taking a big-bath purpose, the income smoothing purpose, and the changes in top management). (2) For non-early adopters, the amounts of assets impairment are associated with, not only the reporting perspective (the income smoothing purpose, the changes in top management), but also the firms’ operational perspective (such as stock returns and sales
∗ We appreciate comments from two anonymous reviewers and participants of 2005 TAA annual meeting.
Any remaining errors are our sole responsibilities.
growth). When analyzing the “amounts” of assets impairment losses for five different types of assets (long-term investments, fixed assets, identifiable intangible assets, goodwill, and other assets), we find that impairment loss for goodwill is less likely to be associated with reporting motivation, probably because restoration of previously recognized impairment loss for goodwill is prohibited under SAFS No. 35. Finally, we examine market reactions to announcements of assets impairment in Taiwan. Our empirical tests of the shareholder wealth effects of impairment announcements reveal that, (1) the stock market reacts significantly and negatively to fourth-quarter impairment losses, but not to fourth-quarter unexpected earnings for early adopters of Taiwan’s SFAS No. 35; and (2) the stock market does not react significantly to first-quarter impairment losses, but does react thusly to first-quarter unexpected earnings for all listed firms. The latter is consistent with Francis et al. (1996). In addition to Francis’s explanations, we further interpret this result in terms of the same reporting dates for annual report and first-quarter report in Taiwan. Unlike the U.S. Securities Exchange Commission, which requires public firms to submit their 10-K on May 31 and their first-quarter 10-Q on April 15, Taiwan’s Securities and Future Bureau (SFB) requires that annual report and first-quarter report be submitted on the same day (April 30) by public firms in Taiwan. Simultaneous announcements of annual report and first-quarter report may further explain our empirical results: the unexpected fourth-quarter earnings reflected in the previous year’s annual report has no the information content at the end of the first quarter since information may have already been reflected in the beginning of the year. However, as expected, unexpected first-quarter earnings do have information content when they are announced. For assets impairment, on the contrary, investors react to the CPA-audited impairment losses in the annual report while no statistical evidence shows that investors react to CPA-reviewed first-quarter impairment losses.
Keywords: Assets Impairment, Taiwan’s SFAS No. 35, Determinants,
Market Reaction
I. INTRODUCTION
I
n the history of Taiwan’s Statements of Financial Accounting Standards (hereinafter SFAS), no standard like Taiwan’s SFAS No. 35 (2004), Accounting for theImpairment of Assets, caused such strong concern in the market and financial press,
except SFAS No.18, Accounting for Pension. As early as December 1995, the U.S. Financial Accounting Standards Board (hereinafter FASB) issued the U.S. SFAS 121,
Accounting for the Impairment of Long-Lived Assets and for Long-Lived Assets to be Disposed of. This statement requires that, if an asset is impaired on the financial report
date, an enterprise must recognize the impairment loss in order to avoid overstating the value of the asset. In addition to recording historical costs when acquisition and periodically depreciating or amortizing the original costs, long-lived assets are required to be carried at the lower of their carrying values or recoverable values. In January of 2001,
the International Accounting Standards Board (hereinafter IASB) also released International Accounting Statement No. 36, Impairment of Assets, for the accounting of impairment losses of long-lived assets. In addition, the U. S. FASB released U. S. SFAS 142, Goodwill and Other Intangible Assets, to regulate the accounting of goodwill and other intangible assets when acquired and after acquisitions, including their impairments, in July 2001.
For a long time, Taiwan’s Generally Accepted Accounting Principles required the recording of long-lived assets by their historical costs less accumulated depreciations (or amortizations). However, when increases in the consumer price index accumulate to 25% compared to the recorded costs, long-lived assets can be recognized by the current value to reflect assets appreciation. Nevertheless, before SFAS No. 35, Taiwan’s accounting standards for long-lived assets had never considered assets impairment. Long-lived assets could only be written-up, but never written-down, causing an imbalance phenomenon in the accounting records for long-lived assets. This limited accounting for long-lived assets seriously reduced the relevance of the financial reports. And, the carrying amounts of long-lived assets could not fully reflect value changes caused by progress in technology and effect of the business cycle. This accounting treatment had long been criticized by accounting and finance circles in Taiwan. In order to match the International Accounting Standards and to improve the transparency of financial statements, Taiwan’s Financial Accounting Standards Committee of the Accounting Research and Development Foundation of the Republic of China issued SFAS No. 35 Accounting for the Impairment
of Assets on July 1, 2004. This new accounting standard was effective for financial
statements ending on and after December 31, 2005, with early adoption encouraged. Taiwan’s SFAS No. 35 is based on the International Accounting Standard No. 36 and partly refers to related U. S. accounting standards, especially U.S. SFAF No. 121 and SFAS No. 142. The new accounting requires that an enterprise conduct an “impairment test” for the values of long-lived assets (including long-term equity investments, fixed assets, recognizable intangible assets, goodwill, and other assets) on the balance sheet date. If the carrying amount of a long-lived asset is higher than its “recoverable amount”, then an impairment loss must be recognized on the income statement instantly. Later on, if there were any indications that previously recognized impairment losses did not exist, the restoration of that loss would be permitted on that balance sheet date. (However, restoration of previously recognized impairment losses for goodwill is prohibited under Taiwan’s SFAS No. 35 because subsequent increases in the recoverable amounts of goodwill may be partly from goodwill generated internally in the enterprise, instead of the disappearance of the cause of asset impairments that existed previously. In accounting theory, a firm should not record internally generated goodwill. Therefore, the restoration of impairment loss for goodwill is not allowed.) Through recognition and restoration of impairment loss for ling-lived assets, Taiwan’s SFAS No. 35 conservatively reflects the value of long-lived assets since assets appreciation is not discussed in Taiwan’s SFAS No. 35.
On August 26, 2004, Wistron,1 a member of the Taiwan’s Acer group, announced in
a press conference that she would recognize an impairment loss of NT$5.5 billion for her
long-term investment in Aopen accounted by the equity method, and an impairment loss of NT$10.51 by revaluating the long-lived assets in Wistron Philippines (equivalently NT$1.88 impairment losses per share in total) in her 2004 semiannual report, according to the spirit of SFAS No. 35. Wistron’s annual earnings forecast thus decreased from a profit of NT$8.5 billion to a loss of NT$8 billion. (Equivalently, her EPS would be reduced from NT$1.89 to -NT$1.05). This information caused panic in the market, and Wistron’s stock price slumped by 13% within two days. In the beginning of 2005, Taiwan’s Yageo officially declared she would adopt SFAS No. 35 early in her 2004 annual report in a press conference on February 23, 2005. Yageo “cleared the deck” by recognizing impairment losses of NT$120 billion (equivalent to NT$5.23) for long-lived assets under SFAS No. 35 at the end of fourth quarter in 2004. Most impairment losses were from goodwill revaluation. These impairment losses corroded half of Yageo’s contributed capital and made Yageo’s EPS drop from NT$0.23 to -NT$52. On the next day, Yageo’s
stock price fell dramatically. Subsequently, Taiwan’s Chinalife announced the early adoption of SFAS No. 35 in the fourth quarter of 2004 on March 28, 2005. Chinalife claimed she would recognize an impairment loss of NT$37.93 billion (NT$4.78 per share) from her long-term investments. The Chinalife’s annual profit was turned to a loss. (Equivalently, her EPS was dropped from NT$0.66 to -NT$4.12). This news resulted in Chinalife’s stock price slump and lock to be traded due to a decrease in price by more than 7% on the next day. This event negatively affected all stock prices of listed financial institutions in Taiwan, and selling pressure on the stock market was intense on that day. The Taiwan Stock Exchange Index declined 87 points to 5961 points on March 29, 2005. On March 2005, the average U. S. Dow Jones stock index was up. The downward adjustment in Taiwan Stock Exchange Index around the end of March 2005 seemed to be reflecting the negative impact of SFAS No. 35.
From the beginning of 2005 on, Taiwan’s stock market was under the haze of SFAS No. 35. The “January effect” of Taiwan’s stock market didn’t show up in 2005. Worrying that implementing SFAS No. 35 would reduce firms’ net worth and negatively impact stock price, many firms explicitly hoped Taiwan government would rescind this accounting standard. Such a hope was not realized, however. On April 1, 2005, Taiwan’s Executive Yuan claimed that SFAS No. 35 (effective January 1, 2005) and the new Statute
of Labor Pension (effective July 1, 2005) were the two most influential regulations
affecting Taiwan’s economic performance in 2005. And, she directed the Ministries of Finance and Economic affairs to take special notice and respond prudently3. On April 4,
2005, Taiwan’s Premier held a Yuan meeting with his cabinet members to discuss Taiwan firms’ complaints about SFAS No. 35 in regard to its negative impact on firms’ earnings and stock price. The meeting concluded that although SFAS No. 35 might negatively affect stock market in the short term, its positive effect would show up in the long term. The reform of financial regulations in Taiwan could not halt. Since SFAS No. 35 came effective in January 2005, its impact on 2004 annual report (to be disclosed on April 30, 2005) would not be substantial. Ten days later, the Financial Supervisory Commission of Executive Yuan (thereafter FSC) made a special report to the Finance Committee of the
2 Economic Daily, February 23, 2005; Wealth Magazine, Vol. 278, May 2005.
Legislative Yuan on the “stock market reaction to the new SFAS No. 35” on April 13, 2005. The FSC argued that the overall effects of SFAS No. 35 were not substantial. However, some legislators did not agree. It was estimated that government-owned firms, such as Taiwan Sugar Co., Chinese Petroleum Co., Taiwan Power Co., etc., would recognize approximately NT$500 billion in assets impairment in total, which would greatly affect earnings return to the National Treasury. All non-operational funds under respective ministries, such as the Water Resource Fund, etc., would also recognize about NT$300 billion assets impairment. Furthermore, allowing firms themselves to determine the adoption timing and the assets being revaluated by appraisers, SFAS No. 35 left a lot of loopholes for insider trading4. Facing tremendous opposition from industrial and
commercial circles, the Business Services of the Ministry of Economic Affairs finally passed a resolution on April 23, 2005 allowing firms with contributed capital under NT 30 million a 2-year grace period to delay applying SFAS No. 35 until the end of 2007.
Undoubtedly, the accounting treatment of assets impairment under SFAS No. 35 will affect a firm’s net worth and its current earnings. However, estimating the amount of impairment loss will become a challenge to the credibility of a firm’s management and financial statements. SFAS No. 35 defines the impairment loss of an asset as the difference between its carrying amount and the recoverable value if the former is larger than the latter. The recoverable value of an asset is its “value in use” or its “net fair value”, whichever is larger where the “value in use” of an asset is determined by discounting future cash flows from the asset. Therefore, management has to estimate a discount rate and forecast future cash flows from a long-lived asset in order to determine the “value in use”. In addition, the fair value of an asset depends on the liquidity of that asset. If fair market value is unavailable, estimating the fair value of a long-lived asset shall be subject to the judgment of managers or appraisers. Hence, determining the impairment loss of an asset under SFAS No. 35 cannot escape the manager’s discretion.
Taiwan’s SFAS No. 35 was issued on July 1, 2004 and came effective for financial statements ending on and after December 31, 2005, with early adoption encouraged. Thereafter, firms in Taiwan could voluntarily adopt the new SFAS No. 35 early in the third quarter or the fourth quarter of 2004, or face mandatory adoption of this standard in the first quarter of 2005. Managers have full discretion on the adoption time of SFAS No. 35. Our first research issue in this study is to investigate the determinants of early adoption of SFAS No. 35 for listed firms in Taiwan. Previous studies on the adoption timing of U. S. SFAS No. 121 revealed that the determinants can be classified into two categories: management’s “reporting motivations” and a firm’s “operational factors”. Management’s reporting motivations reflect several scenarios. If a firm has unexpectedly high earnings performance, its management has incentive to recognize a large amount of impairment loss to smooth out earnings (the “income-smoothing” purpose); if a firm has unexpectedly poor earnings performance, management has incentive to clear the dirty deck of impaired assets to improve investors’ perceptions of the future financial performance of the firm (so-called “taking a big bath” purpose); when a firm changes its management, new managers may exercise greater scrutiny over existing assets and want to clean up assets impairments. Furthermore, some firms may determine the timing of
adopting SFAS No. 35 due to their respective “operational factors”.
In judging the impairment of long-lived assets, SFAS No. 35 unequivocally requires management to examine external and internal information (such as technology, market, economic and legal environment or the market of the assets) to see whether significant adverse changes affecting the fair value of long-lived assets occur. Therefore, based on management’s “reporting motivation” and a firm’s “operational factors”, managers may exploit the discretion allowed under SFAS No. 35 in determining the amount of impairment losses for long-lived assets. Our second research issue is to explore the determinants of the size of assets impairment.
Finally, we examine market reactions to the announcement of assets impairment losses. Previous studies suggest that impairment announcements may have any of three different information contents: The first one is that impairment information reflects a decrease in economic values of company assets. Under this viewpoint, a negative market reaction to this impairment announcement is expected. The second one is that impairment information is a signal of changes in strategy to improve future performance. A positive market reaction to this impairment announcement is expected. The third one is that impairment information may reflect a firm’s willingness and ability to manage earnings opportunistically. It is not clear whether investors react positively or negatively to this information.
This study investigates determinants of the timing and the amount of assets impairment decisions for listed firms in Taiwan. We also examine market reaction to impairment announcements. Our paper is organized as follows: Section two summarizes prior studies on assets impairment. Section three describes our hypotheses and research design. Section four presents sample selection, descriptive statistics and empirical results. Section five includes sensitivity tests. Section six presents our conclusions and suggestions.
II. LITERATURE REVIEW
Before the U.S. SFAS No. 121 was issued in December 1995, there was no mandatory accounting requirement for firms to reduce the carrying value of a long-lived asset that had become impaired. During this period of time, the only related FASB standard was the SFAS No. 5, Accounting for Contingencies (1975), which provided some general guidelines by requiring firms to recognize impairment loss only if an asset might be impaired and the amount of loss could be reasonably estimated. However, the discretionary flexibility of SFAS No.5 resulted in various accounting treatments and reportings of assets impairments. Before the issuance of SFAS No. 121, the U.S. GAAP, in fact, allowed managers a lot of discretion in deciding when to recognize a writedown and how much to write down the assets. And consequently, assets impairment decisions became a matter of managers’ discretion.
Most studies done in the U.S. on assets writedowns under SFAS No. 5 were information-content research. However, the empirical results were inconsistent as to whether assets writedowns constituted good or bad news. Some studies argued that firms’ recognizing assets impairment was good news, since it conveyed a signal of possible improvement in future performances. For example, Strong and Meyer (1987) examined the effect of announcements of asset writedowns under SFAS No. 5 on security returns.
They found negative average CARs around the impairment announcements date. However, the negative returns were reversed in sixty days after the impairment announcement date. Therefore, they endorsed the argument that recognizing assets impairment conveyed signals of improving future performance. Rees, the Gill and Gore (1996) held the same position. Examining the relationship between assets writedowns under SFAS No. 5 and abnormal accruals, Rees et al. found that firms recognized significantly more negative abnormal accruals in the years of assets writedowns, and these accruals were not reversed in subsequent years. They argued that the writedowns and concurrent discretionary accruals were appropriate responses of managers to a changing economic environment, rather than managers’ opportunistic earnings management. Managers could use their discretion to signal value-relevance information to investors so that assets writedowns helped in valuing a firm.
However, some argued that recognizing assets impairment reflected real damage to the firm’s earning power. Elliott and Shaw (1988) found that U.S. firms usually recognized assets writedowns under SFAS No. 5 in audited annual reports, instead of in interim reports. They showed that assets writedowns would cause long-term decreases in stock returns. Their empirical results did not support the argument that recognizing assets impairment conveyed a signal of improving future performance. Francis, Hanna, and Vincent (1996) obtained similar results. Studying firms’ characteristics for assets writedowns decisions and market reaction to writedowns under SFAS No. 5, Francis et al. discovered that assets writedowns were positively related to changes in CEOs. They also found negative market reaction to assets writedown announcements. Bartov, Lindah, and Ricks (1998) found significantly negative stock returns in years after a firm recognized assets writedowns under SFAS No. 5. They argued that the market might not fully understand the economic consequences of assets impairment.
Nevertheless, some studies documented no significant market reaction to assets writedown under SFAS No. 5. Zucca and Campbell (1992) argued that recognizing assets impairment was, for a firm, managing earnings (e.g., taking “big baths” or smoothing income). They found no significant market reaction to assets writedown announcements under SFAS No. 5, since investors could see through manipulations. Hogan and Jeter (1998) also found no significant relationship between CAR and assets writedowns under SFAS No. 5. They argued that asset writedowns might reflect changes in CEOs, rather than substantial impairments in long-lived assets.
With respect to motivations to write down long-lived assets, Strong and Meyer (1987), Elliott and Shaw (1988), Zucca and Campbell (1992), Francis et al. (1996), Rees et al.(1996), Hogan and Jeter (1998) all found that firms’ writedown decisions were generally associated with managers reporting motivations. Examining the motivation of revaluating long-lived assets for 72 Australian firms during the years 1981 through 1990, Easton, Eddy and Harris (1993) interviewed CFOs of these sample firms and found that assets revaluation showed very weak association with stock performance. Revaluation of long-lived assets was not done in a timely manner. Further, Easton and Eddey (1997) extended the Easton et al. (1993) study by including a recent period (1990~1993) when Australian firms most likely suffered assets impairment. Barth and Clinch (1998) also extended the Easton et al. (1993) study to examine the explanatory power of assets revaluations regarding stock price and two-years-ahead earnings forecasts. Both studies
concluded that assets revaluation for a firm generally was not done in a timely manner in general. Using U.S. data, Heflin and Warfield (1997) obtained similar results. They documented that American firms delayed recognizing assets writedowns for three years on average. They indicated managers recognizing assets writedowns aimed at managing earnings, instead of recording the real assets impairment. Managers tended to recognize assets writedowns in the years when earnings were really poor to “clear the deck” or in the years when earnings were extremely good to smooth reported earnings. Deng and Lev (1998) also documented that, after recognizing assets impairment, a firm’s returns on assets and returns on equity tend to increase in the next year.
After U.S. SFAS No. 121 was issued in December 1995, few studies examined the impact of SFAS No. 121. Comprix (2000) documented that firms recognizing assets impairment under SFAS No.121 would make the book values of their long-lived assets closer to market values. Also, investors would value a firm lower if it recognized assets impairment. In addition, the records of earnings performance and the history of assets impairment could influence investors’ valuation of the firm. Riedl (2004) found the assets writedowns were highly related with economic factors before SFAS No.121. However, after SFAS No. 121, firms recognizing assets impairment aimed at taking a “big bath” to manage earnings, reflecting managers’ opportunistic reporting behavior rather than providing private information to reduce information asymmetry. Consequently, Riedl argued that the quality of financial reporting after SFAS No.121 decreased.
In Taiwan, there was no accounting standard for assets impairment before Taiwan’s SFAS No. 35 was issued in 2004, nor was there any academic study relating to assets writedowns under Taiwan’s SFAS No. 9, Accounting for Contingencies and Subsequent
Events (1986), to the best of our knowledge. Due to expected severe impacts of Taiwan’s
SFAS No. 35 on financial statements and stock prices, we investigate the determinants of the timing and amount of assets impairment for listed firms in Taiwan. In addition, we examine market reaction to impairments announcements in Taiwan in this study.
III. HYPOTHESES DEVELOPMENT
We first investigate the determinants of adoption timing of SFAS No. 35 for listed firms in Taiwan. Taiwan’s SFAS No. 35 was promulgated on July 1, 2004 and was effective on January 1, 2005. However, early adoption was allowed. Therefore, managers could use reporting flexibility in determining whether to adopt SFAS No. 35 on time in the first quarter of 2005 or earlier in 1994. The adoption time of SFAS No. 35 is at the managers’ discretion.
Zucca and Campbell (1992) argued that recognizing assets writedowns under SFAS No. 5 was one of the means of earnings management. Some managers managed earnings for taking “big baths,” others for income smoothing. Riedl (2004) also found that, after SFAS No.121 was issued, the main motivation for recognizing assets impairment losses was for taking “big baths” for improving the future profitability of the firm. Francis et al. (1996) documented that in the years of changing CEOs, recognizing assets writedowns was very common. This implies the new CEO intends to clean up losses that occurred in the tenures of preceding CEOs to make a fresh start. This is literature evidence that recognizing assets impairment losses is associated with managers’ reporting motivation.
Furthermore, SFAS No. 35 requires that, if there is any indication that an asset may be impaired, managers should estimate the recoverable amount of impaired long-lived assets in order to make sure the long-lived assets are impaired. Such “indications” include simultaneously internal and external information. Therefore, firms adopting SFAS No. 35 early are driven, not only by managers’ reporting motivation, but also by the firms’ operational factors. Thus, our first hypothesis is proposed as follows：
Hypothesis 1: The adoption timing of SFAS No. 35 is associated with both management’s “reporting motivations” and a firm’s “operation factors.”
The proxies and predicated signs of “reporting motivations” and “operational factors” are described in the next section.
Secondly, we explore the amount of assets impairment for listed firms in Taiwan. Although Taiwan’s SFAS No. 35 provides some guidance on the evaluation, measurement, recognition of impairment loss, this statement could not totally eliminate management’s discretion on the timing and amount of assets impairment. In addition, before assets impairment loss is determined, management is required by SFAS No. 35 to consider not only firm’s internal information, but also external information outside the firm, such as significant unfavorable current and future changes in industry technology, market, economic, legal environments, and the market to which the asset belongs. Therefore, the amount of assets impairment may be affected by both management’s “reporting motivations” and a firm’s “operational factors.” Accordingly, our second hypothesis is proposed as follows:
Hypothesis 2: The amount of assets impairment decision is associated with management’s “reporting motivations” and a firm’s “operational factors.”
The proxies and expected signs of “reporting motivations” and “operational factors” are also described in the next section.
Finally, we examine the information content of assets impairment for Taiwan’s listed firms. Francis et al. (1996) indicated that announcement of assets writedowns might be interpreted by investors in three different ways. First, it might signal a real decrease in the economic value of assets so that negative market reaction to the impairment announcement would occur. Second, it might signal changes in management strategy and improvement in future profitability so that positive market reaction to the impairment announcement would happen. Third, it might signal management’s manipulation of earnings so that market would not react to the assets impairment announcement. Previous studies of market reaction to assets impairment were inconsistent. Strong & Meyers (1987), Rees et al. (1996) found significantly positive market reaction to the assets writedowns announcements. On the contrary, Elliot and Shaw (1988), Francis et al. (1996) documented significantly negative market reaction to the assets writedowns. Nevertheless, Zucca and Campbell (1992), Hogan and Jeter (1998) found insignificant market reaction to assets writedowns.
Individual investors accounted for 77.57%5 of total trading volume in the Taiwan Stock
Exchange in 2004. Institutional investors that have ability to hire accounting and finance professionals to examine financial statements only accounted for 22.43% of total trading volume. Therefore, we do not expect that the current Taiwan stock market as a whole could fully understand financial reports. As indicated in Section I, Taiwan’s stock market reactions to early adopters of SFAS No. 35 were extremely negative, causing serious concerns from Taiwan’s Executive Yuan, Legislative Yuan, Financial Supervisory Commission and the financial press. Therefore, we expect that investors will react negatively to the assets impairment information for firms early adopting SFAS No. 35 and recognize assets impairment loss in 2004 annual reports (which are audited by external auditors). However, since assets impairment involves management judgment and estimation, unaudited quarterly financial statements receive less credibility. Accordingly, the negative market reaction to the assets impairment loss in the unaudited 2005 first-quarter financial report would be reduced. Thus, our third hypothesis is proposed as follows:
Hypothesis 3: Taiwan’s stock market would negatively respond to assets impairment losses disclosed in the 2004 annual report (which was audited by external auditors), and in the first quarter of 2005 (which was not audited by external auditors) to a lesser extent.
IV. RESEARCH DESIGN
Determinants of the Timing and Amount of Assets Impairment Decision
Taiwan’s SFAS No. 35 was issued on July 1, 2004, and effective for financial statements ending on and after December 31, 2005, with early adoption allowed. That is, managers could early adopt SFAS No. 35 in the second, third, or fourth quarter of 2004, or adopt it in the first quarter of 2005. Elliott and Shaw (1988) found that U.S. firms usually recognized assets impairment losses in the audited annual report. We also find that Taiwan’s listed firms chose early adoption of SFAS No. 35 in the 2004 annual report (fourth quarter), rather than in the second or third quarter of 2004.
We investigate first the determinants of adoption timing of SFAS No. 35. Our experiment group consists of early adopters of SFAS No. 35 in the 2004 annual report for listed firms in Taiwan; the control group consists of firms matching our experiment group in terms of market value (size), industry, and recognizing impairment loss in the first quarter of 2005. Secondly, we explore the determinants of amount of assets impairment losses. Our experiment group thus consists of firms recognizing assets impairment losses in 2004 annual report or in the first quarter of 2005, while the control group consists of firms matching our experiment group in terms of market value (size), industry, and not recognizing impairment loss in the same periods. Finally, we examine information content of assets impairment for listed firms in Taiwan. The sample consists of listed
5 “Statistics of Investors Structure in terms of Trading Volume on TSEC Market,” by Securities and Futures
Bureau, Financial supervisory Commission, Executive Yuan, 2005. (http://www.sfb.gov.tw/statistics/point/9507/t16.xls)
firms in Taiwan with assets impairment disclosure.
We define assets impairment loss as the difference between the carrying amount of long-lived assets and its economic value (recoverable amount), including decrease in investment income resulting from impairment loss on long-term equity investments by the equity method. Impairment losses for five types of long-lived assets (that is, long-term equity investment, fixed assets, goodwill, identifiable intangible assets and other assets) are available from footnote of each firm’s financial statements.
Logistics regression is employed to examine the adoption timing of SFAS No. 35 for listed firms in Taiwan in Hypothesis 1. We estimate a logistics regression as follows:
ADOPT
i=
α
0+
α
1BATH
i+
α
2SMOOTH
i+
α
3∆
MGT
i+
α
4FIN
i+
α
5∆
INDROA
ii i i
RET
SALE
ELEC
+
+
∆
+
α
6α
7α
8+
α
9∆
OCF
i+
α
10MTB
i+
ε
i (1) where, iADOPT
: indicator variable, which equals 1 if firm i recognizes impairment loss in2004 annual report, and 0 otherwise.
Hypothesis 2 explores the determinants of the amount of assets impairment. If a firm does not recognize assets impairment loss, the value of dependent variable(amount of impairment) is zero. Once assets impairment loss is recognized, the dependent variable will be a ratio scale. The dependent variable, therefore, is a truncated datum. Thus, we employ a tobit regression to explore the amount of assets impairment (Hypothesis 2). We estimate a tobit regression as follows:
it it
it it
it
it
BATH
SMOOTH
MGT
FIN
INDROA
WOTA
=
α
0+
α
1+
α
2+
α
3∆
+
α
4+
α
5∆
it it it it itit
RET
SALE
OCF
MTB
ELEC
α
α
α
α
ε
α
+
+
∆
+
∆
+
+
+
6 7 8 9 10 (2)where, it
WOTA
: firm i’s impairment loss6 at quarter t, divided by firm i’s total assets
6 The assets impairment losses under SFAS No. 35 may not be deductible from undistributed earnings that
are subject to 10% surtax. The Tax Agency of the Ministry of Finance, R.O.C. announced that:
"SFAS No.35 Accounting for the impairment of assets requires that long-lived assets, such as fixed assets, identifiable intangible assets, goodwill, and long-term equity investment by equity method, should valuated in the lesser of carrying value and its recoverable amount. The difference between the recoverable amount and carrying value shall be recognized as impairment loss. If such impairment loss (except for goodwill) does not exist or is reduced in subsequent period, the firm should restore previously recognized impairment loss as income in that period.”According to Article 66-9 of the income tax law, the 10% surtax on undistributed earnings is based on taxable income. In calculating the undistributed earnings, gain or loss derived from the time difference between financial accounting and tax regulation would not be deductible or addable. Hence, unrealized impairment loss or restoring previously recognized impairment loss under SFAS No.35 belongs to the timing differences between financial accounting and tax regulation. Accordingly, impairment loss and its restored amount would not be a deduction or an addition in calculating the 10% surtax of undistributed earnings. Impairment loss under SFAS No.35
at the end of quarter t-1. it
BATH
: proxy for taking big baths. If∆
E
it7 < median of the “unexpectednegative earnings”8, then
BATH
it=∆
E
it; otherwise,BATH
it= 0.9 itSMOOTH
: proxy for income-smoothing. If∆
E
it> median of the “unexpectednegative earnings” 10 , then
SMOOTH
it =∆
E
it ; otherwise, itSMOOTH
= 0. 11it
MGT
∆
: indicator variable, which equals 1 if firm i changes its top management (defined as CEO, chairman of the boardor CFO) from year t-1 to t; and 0 otherwise.it
FIN
: extent of demand for capitals ＝ (total amounts of firm i’s issuance of equity capital and corporate bonds in quarter t, divided by firm i’s total assets at the end of quarter t-1.)it
INDROA
∆
: growth rate of return on assets for the industry ＝ (the median of growth rate of return on assets from quarter t-1 to t in firm i’s industry).it
ELEC
: indicator variable, which equals 1 if firm i belongs to electronics industry at the end of quarter t; and 0 otherwise.it
RET
: firm i’s quarterly stock returns ＝ (fim i’s stock returns from quartert-1 to t).
it
SALE
∆
: firm i’s growth in quarterly sales ＝ (firm i’s sales in quarter t - firm i’s sales in quarter t-1 ) / firm i’s total assets at the end of quarter t-1.it
OCF
∆
: firm i’s growth in cash flows from operations＝ (firm i’s cash flows from operations in quarter t - firm i’s cash flows from operations in quarter t-1) / firm i’s total assets at the end of quarter t-1.it
MTB
: indicator variable, which equals 1 if firm i's market-to-book ratios is below 1 at the end of quarter t; and 0 otherwise.
belongs to unrealized loss. Tax regulation allows firms to recognize losses in assets only when assets are damaged, discarded, disposed of, or disinvestment or liquidation of investee companies.” (Taxation Agency, 2005/2/25) Apparently, firms recognizing assets impairment loss under SFAS No. 35 in Taiwan cannot save taxes. Therefore, we do not consider the tax effect of impairment loss in this paper.
7 Unexpected earnings (ΔE
it) = (firm i’s pre-impairment earnings in quarter t － firm i’s earnings in
quarter t-4) / firm i’s total assets at the end of quarter t-1.
8 Unexpected negative earnings is defined as unexpected earnings being less than zero. We then rank all
unexpected negative earnings for all firms in the same period and find its median.
9 As suggested by Reidl (2004).
10 Unexpected positive earnings is defined as unexpected earnings being more than zero. We then rank all
unexpected positive earnings for all firms in the same period and find its median.
Reporting motivations
Management has considerable discretion on the timing of adopting SFAS No. 35 and on the amount of assets impairment. Zucca and Campbell (1992) indicated that if managers’ objective in recognizing assets impairment is to manipulate earnings, they might take either the “income-smoothing” tactic or the “taking-big-baths” tactic. Murthy (1985), Antle & Smith (1986), Lambert et al. (1987) argued that management’s incentive plans generally tie in with reported income. Managers have motivations to maximize their bonuses through manipulating accounting earnings. Income smoothing literature argues that if management incentive plans are based on the income-smoothing purpose, managers will do so for their personal interest. Thus, management will selectively recognize assets impairment losses in periods with high earnings to attain the goal of income smoothing.Taking big baths literature argues that when earnings are abnormally low, management has incentive to “clear the deck” by recognizing assets impairment losses to signal that “the worst period has already passed and the future will be bright.”
Following Riedl (2004), we use independent variables BATH and SMOOTH to represent the taking-big-baths tactic and the income-smoothing tactic, respectively. BATH is defined as the unexpected negative earnings if they are lower than the median for all firms in the same quarter, and zero otherwise. It represents that the firm has unexpected poor earnings performance. On the contrary, SMOOTH is defined as the unexpected positive earnings if it is higher than the median for all firms in the same quarter, and zero otherwise. It represents that the firm has unexpected good earnings performance. When management employs the taking- big-baths tactic to manipulate earnings, it will adopt SFAS No. 35 in the period of exceptionally poor earnings to “clear the deck.” We expect a negative relation between BATH and early adoption of SFAS No. 35 or the amount of impairment losses. On the other hand, if management uses the income-smoothing tactic to manipulate earnings, it will adopt SFAS No. 35 in the period of exceptionally good earnings performance to recognize a large amount of impairment losses. We expect a positive relation between SMOOTH and the early adoption of SFAS No. 35 or the amount of assets impairment losses.
Other reporting motivations include change in top management. New management has incentive to “clear the deck” by recognizing assets impairment losses to improve its future financial performance. Following Francis et al. (1996), ΔMGT is defined as changes in CEO, chairman of the board, or CFO. We expect a positive relation between ΔMGT and the early adoption of SFAS No. 35 or the amount of assets impairment losses.
When a firm needs long-term capital, it may issue capital stock or corporate bond. However, the ability to raise capital and the cost of the new capital depend on a firm’s reported earnings. In order to raise additional capital at lower cost, management would be reluctant to recognize assets impairment losses. We use FIN (the amount of seasonal equity and debt issuances deflated by total assets) to capture the need for long-term capital. We expect a negative relation between FIN and the early adoption of SFAS No. 35 or the amount of assets impairment losses.
Operational factors
SFAS No. 35 requires that, when there are indications that an asset may be impaired on the financial reporting date, management should estimate the recoverable amount of the impaired long-lived assets to make sure the long-lived assets are impaired and to adjust the value of assets according to current position. Such “indications” include internal information and external information. Examples of internal information for assets impairment are: evidence of physical damage or obsolescence of an asset, or significant adverse changes in the scope or in the manner in which an asset is used or is expected to be used have occurred or will take place in the near future. Examples of external information for assets impairment are: the decline of an asset’s market value during the period being significantly larger than the expected decline as a result of the passage of time or normal use; or significant adverse changes in industrial technology, market, economy, legal environment, or in the market to which the asset belongs having occurred during the period or will take place in the near future. Since an assets impairment decision could be driven by industry condition, a firm’s operational condition and assets usage condition, we further consider the operational factors in order to capture the cross-sectional variation of impairment losses.
Assets impairment may be related with changes in industrial technology and business environment. Firms in a declining industry are more likely to recognize larger amounts of impairment losses; and firms in a growing industry are less likely to have assets impairment. Therefore, the industry performance can influence the timing of adopting SFAS No. 35 and the amount of assets impairment losses. We use the change in industry return on assets (∆ INDROA) to capture the industry effect on the timing and amount of assets impairment decisions, and we predict a negative association between them, as suggested in Riedl (2004). In addition, due to the fact that the electronics industry has a higher reinvestment ratio and shorter life cycle, firms in this industry are more likely to recognize assets impairment losses. We expect a positive relationship between the electronics industry indicator (ELEC) and the early adoption of SFAS No. 35 or the amount of assets impairment losses.
Furthermore, assets impairment losses may be affected by a firm’s past operational and assets-usage conditions. Francis et al. (1996) suggested that the worse a firm’s past stock price performance has been, the more likely management will clear the deck of impaired assets. Since a firm’s stock price reflects investors’ expectation on the firm’s future performance, we employ stock returns before assets impairment (RET) to capture investors’ expectation on the firm’s future performance, and expect a negative relationship between RET and the early adoption of SFAS No. 35 or the amount of assets impairment losses.
Poor usage and idle capacity of an asset will reduce its value. Riedl (2004) argued that recognizing assets impairment losses implies lowering the recovery of specific assets. He used sales growth (∆SALE) to capture assets recovery in an accrual basis and used operational cashflows growth (∆OCF) to capture assets recovery in cash basis. We expect a negative relationship between ∆ SALE (or ∆ OCF) and the amount of assets impairment losses. However, when determining the adoption timing of SFAS No. 35, management will consider whether a firm can bear the impact of recognizing impairment losses. A firm will early adopt SFAS No. 35 only when its sales growth and operational cashflows
long-term market reaction to the recognized impairment losses.
REFERENCES
Antle, R., and A. Smith. 1986. An empirical investigation into relative performance evaluation of corporate executives. Journal of Accounting Research 24, 1-39.
Barth, M., and G. Clinch. 1998. Revalued financial, tangible, and intangible assets: Associations with shares and non-market-based value estimates. Journal of
Accounting Research 36, 199-233
Bartov, E., F. Lindah, and W. Ricks. 1998. Stock price behavior around announcements of write-offs. Review of Accounting Studies 3, 327-346.
Bunsis, H. 1997. A description and market analysis of write-off announcements. Journal
of Business Finance and Accounting 24, 1385-1400.
Comprix, J. 2000. Write-offs and restructuring charges: Evidence from SFAS No. 121 and EITF 94-3 mandatory disclosures. Unpublished dissertation, University of Illinois at Urbana-Champaign.
Deng, Z., and B. Lev. 1998. The valuation of acquired R&D. Working paper, New York University.
Easton, P., P. Eddey, and T. Harris. 1993. An investigation of revaluations of tangilble long-lived assets. Journal of Accounting Research 31, 1-38.
Easton, P., and P. Eddey. 1997. The relevance of asset revaluations over an economic cycle. Australian Accounting Review: 22-30
Elliott, J., and W. Shaw. 1988. Write-offs as accounting procedures to manage perceptions.
Journal of Accounting Research 26, 91-119.
FASB. 1975. Statement of Financial Accounting Standards No. 5: Accounting for
Contingencies. Norwalk, CT: Financial Accounting Standards Board.
————. 1995. Statement of Financial Accounting Standards No. 121: Accounting for
the impairment of long-lived assets and for long-lived assets to be disposed of.
Norwalk, CT: Financial Accounting Standards Board.
————. 2001. Statement of Financial Accounting Standards No. 142: Accounting for
the goodwill and other intangible assets. Norwalk, CT: Financial Accounting
Standards Board.
Francis, J., J. Hanna, and L. Vincent. 1996. Causes and effects of discretionary asset write-offs. Journal of Accounting Research 34, 117-134.
Heflin, F., and T. Warfield. 1997. Managerial discretion in accounting for asset write-offs. Working paper, University of Wisconsin-Madison.
Hogan, C., and D. Jeter. 1998. The information content of restructuring charges: A contextual analysis. Working paper, Vanderbilt University.
IASC. 2001. IFRS No. 36: Impairment of Assets. NY: International Accounting Standards Committee.
Lambert, R., D. Larcker. And G. Baker. 1987. An analysis of the use of accounting and market measures of performance in executive compensation contracts. Journal of
Murphy, K. 1985. Corporate performance and managerial remuneration: An empirical analysis. Journal of Accounting and Economics 7, 11-42.
Rees, L., S. Gill, and R. Gore. 1996. An investigation of asset write-downs and concurrent abnormal accruals. Journal of Accounting Research 34, 157-169.
Riedl, J. 2004. An examination of long-lived asset impairment. Accounting Review 79, 823-852
Strong, J. and J. Meyer. 1987. Asset writedowns: Managerial incentives and security returns. Journal of Finance 42, 643-661.
Taiwan’s Financial Accounting Standard Committee. 2005. SFAS No. 35, Accounting for
the Impairment of Assets, Taipei: Accounting Research and Development
Foundation, Taiwan.
Zucca, L. and D. Campbell. 1992. A closer look at discretionary writedowns of impaired assets . Accounting Horizons 6, 30-41.
