M&A Accounting and Firm Valuation Paul Healy Harvard Business School. For many companies, mergers and acquisitions are important investments

Transcription

1 M&A Accounting and Firm Valuation Paul Healy Harvard Business School Introduction For many companies, mergers and acquisitions are important investments that are sources of growth and transformation. In 2014 alone, their global value amounted to $3.2 trillion. Many of these deals are bet the farm investments for acquirers with potential to transform their firms and even their industries. But prior research indicates many fail to add value for acquiring shareholders (see Healy, Palepu and Ruback, 1997). Given their economic significance, it is important that M&A reporting standards provide timely information on which acquisitions have added value and which have not. Pooling of interests accounting failed to meet this objective. It valued an acquisition at its book value reported on the target firm's books prior to the acquisition. This made it challenging for investors to compare the post-acquisition earnings generated by the combined firm with the actual cost of the net assets acquired. In contrast, by recognizing the cost of assets acquired at their fair values at the date of acquisition, purchase accounting enables investors to judge whether the net assets acquired have generated an attractive book return. Yet purchase accounting raises a number of important questions for standard setters. How should the difference between the purchase price of the business and the fair value of separable tangible net assets be reported? Should the acquirer be required to separately value identifiable intangibles, such as brands and in-process R&D for the target? Should goodwill (the residual) be allocated across divisions? 1

2 Should intangible assets be amortized after an acquisition, or should they be retained at cost and subject to an impairment test? FAS 141 eliminated the pooling option and required that all combinations be accounted for under the purchase (or acquisition) method, with identifiable intangibles acquired reported at their fair values at date of purchase. Identifiable intangibles with finite lives, such as patents and copyrights, are amortized over their useful lives unless they are deemed impaired, in which event they are written down to their estimated fair value. In contrast, identifiable intangibles with indefinite lives are reported at fair value at date of purchase and subject to an annual impairment test. Any residual value, which reflects the value of intangibles that cannot be separately identified, is recorded as goodwill and allocated across firms business units. FAS 142 required that, since it has an indefinite life, goodwill is not amortized but subject to regular impairment tests. Advocates for current M&A rules argue that they provide investors with information needed to judge whether an acquisition has been successful on a timely basis. Opponents counter that these rules provide managers with too much latitude in reporting for difficult to value intangibles. They note that given the evidence that many acquisitions fail to add value for acquiring shareholders, much of the goodwill reported is bad will and that managers have incentives to fail to recognize these losses on a timely basis (see Holthausen and Watts, 2001; Watts, 2003). Considerable evidence has been compiled in the academic literature on the impact of merger accounting standards. The findings suggest that impairment charges under FAS 142 provide new information to investors (Chambers, 2007; 2

3 Chen, Kohlbeck and Warfield, 2008; Bens, Heltzer and Segal, 2011; and Li, Shroff, Venkataraman and Zhang, 2011). However, there is also evidence that managers take advantage of the judgment provided them under the latest accounting standards to: (a) overstate goodwill, which since it is unamortized increases future earnings (see Shalev, Zhang and Zhang, 2013; and Paugam, Astolfi and Ramond, 2015), and (b) delay recognizing impairments on intangibles, leading stocks to be temporarily be mispriced (Ramanna and Watts, 2012; Li and Sloan, 2015). In this paper I revisit questions on M&A accounting using data for one of the largest acquisitions in history, the Time Warner acquisition of AOL. The acquisition was widely viewed as one of the worst in history, providing ample opportunity to better understand at a micro level how the acquisition was initially reported, when impairment tests were reported, and how the market interpreted the reporting. In the spirit of the conference, given this field evidence, I conclude with some reflections on accounting research and M&A reporting standards. M&A Accounting and AOL s Acquisition of Time Warner On January , AOL announced the acquisition of Time Warner for $162 billion. Despite revenues of only $4.8 billion, dwarfed by Time Warner s $27.3 billion, AOL s high stock valuation enabled its shareholders to end up owning 55% of the combined company and Time Warner (TW) shareholders 45%. TW had itself been forged out of the mergers of Warner Bros. movie studio (which included HBO and Cinemax), Time magazine and its publishing business, and Ted Turner s TBS cable company (which included CNN). AOL, an internet service 3

4 provider, was itself no stranger to acquisitions having acquired Netscape and its largest competitor, Compuserve. The acquisition was intended to allow AOL to transition from being a pure internet service provider into a media and entertainment giant. In anticipation of the growth of broadband access to online content, AOL had sought access to the new technology via deals with telecommunications, cable and satellite companies so that it could provide its subscribers with faster access to the internet than phone lines. The acquisition of TW was seen as a way to accelerate this process. In addition, AOL s 22 million subscribers could be used to leverage TW s television, movie, and publishing content. The deal was also attractive to TW since it had unsuccessfully sought to enter the online market to leverage its content. In reflecting on the combination fifteen years later, McGrath (2015) explained that at the time, a lot of people thought that the merger was a brilliant move and worried that their own companies would be left behind. The whole thing was transformative. Had [the] initial assumptions been borne out, we might be talking today about what a visionary deal it was. Of course, with the benefit of hindsight, the deal is now seen as one of the more spectacular merger failures in history. McGrath explained the problems. Merging the cultures of the combined companies was problematic from the get go. Certainly the lawyers and professionals involved with the merger did the conventional due diligence on the numbers. What also needed to happen, and evidently didn t, was due diligence on the culture. The aggressive and, many said, arrogant AOL people horrified the more staid and corporate Time Warner side. Cooperation and promised synergies failed to materialize as mutual disrespect came to color their relationships. A few scant months after the deal closed, the dot com bubble burst and the economy went into recession. Advertising dollars evaporated, and AOL was 4

5 forced to take a goodwill write-off of nearly $99 billion in 2002, an astonishing sum that shook even the business-hardened writers of the Wall Street Journal. AOL was also losing subscribers and subscription revenue. The total value of AOL stock subsequently went from $226 billion to about $20 billion. The acquisition is also noteworthy because it was one of the first to adopt FAS 141 and 142. Given the visibility and intensive scrutiny of the acquisition, it makes a potentially useful field study to investigate how the standards were implemented and whether they provided timely information on the merged firm s performance. Valuation of Intangibles As reported in Table 1, AOL Time Warner recognized $7.3 billion on intangible assets subject to amortization (primarily for TW s music and film rights) and $37.7 billion for intangibles not subject to amortization (notably TW s cable and sports franchises, and brands and trademarks). These assets reflected the standalone value of TW s identifiable intangibles. On its December 31, 1999 balance sheet, TW valued its music and film right intangibles at $0.8 billion, and the cable television and sports franchise intangibles at $8.5 billion, indicating under acquisitions accounting AOL Time Warner was able to revalue these assets significantly. In addition, AOL Time Warner reported goodwill for $127.4 billion. This reflected (a) the pre-acquisition value of goodwill for TW as a stand-alone business (since TW alone was worth than the sum of its parts); (b) the incremental value of goodwill arising as a result of the merger, presumably reflecting expected merger synergies; and (c) any overpayment by AOL made to acquire TW. Prior to the 5

6 acquisition, on December 31, 1999, TW had reported goodwill of $15.5 billion. However, the standalone value of TW s goodwill would have been substantially higher than this amount. How much, though, is unclear. After the acquisition was completed, as required by FAS 141, AOL Time Warner allocated goodwill across its business segments. The allocation, reported in Table 1, apportions the majority of the goodwill to the AOL, cable and networks segments. What is less clear, however, is how one would make such an allocation. After all, goodwill represents the value of assets that cannot be separately identified; it depends on the interactions between the segments that generate value for the firm. As noted above, some of this goodwill arises from the pre-acquisition value of TW s goodwill, which reflects any synergies between its business units prior to combining with AOL. For example, there may be synergies between the cable and film businesses, which were also brought together under a merger. Yet it is not straightforward to allocate any such goodwill between these units. I suspect that it would have been useful to investors to understand how much of the $127 billion of goodwill reported by AOL Time Warner represented its valuation of goodwill for TW as a stand-alone business. The second goodwill component arises from any projected synergies for the combination of AOL and TW. At the time of the acquisition, industry experts argued that the merged firm could generate synergies by using AOL s network of subscribers to leverage Time Warner s cable and entertainment businesses, and by providing AOL s subscribers with faster broadband access to the internet. But if 6

7 such is the case, it is unclear whether the related goodwill should be attributed to the AOL business segment, to the Cable, or to the Entertainment segments. If these business units were strictly independent, it would be straightforward to separately estimate the value of goodwill for each. But given the businesses were combined precisely because management expected that they would create synergies, leading the whole to exceed the sum of the parts, any such allocation seems arbitrary. Intangible Impairment FAS 142 requires that goodwill be carried on the books at its fair value at purchase, and not subject to periodic amortization. Instead, each year its current assessed fair value is compared to its book value to decide whether an impairment charge should be recorded. In contrast, under the former standard, FAS 121, impairments were recognized when the undiscounted estimated future cash flows attributable to goodwill fell below book value, a less stringent standard that the FASB recognized might happen rarely for an indefinite lived asset. AOL Time Warner s goodwill was materially impacted by the new standard for impairment. In early 2002, the company reported that as a result of applying the fair value benchmark for write-downs under FAS 142, rather than the undiscounted cash flow benchmark, it would have to write down goodwill by $54.2 billion. As reported in Table 2, this charge was taken against the goodwill allocations to TW business units. Consistent with the findings of Beatty and Weber (2006), AOL Time Warner opted to report the impact of the adoption of FAS 142 by showing the impairment charge as the effect of an accounting change, reported below the line. 7

8 Table 1 Summary of 2001 segment data for AOL Time Warner Panel A: Goodwill and operating data by segment Segment Goodwill, Dec 31 Revenues EBITDA Operating profit Operating assets AOL $27,729 22% $8,615 22% $2,914 30% $2, % $6,616 3% Cable 33,259 26% 6,028 15% 2,628 27% (748) -66% 72,087 37% Filmed Entertainment 9,110 7% 8,759 22% 1,017 11% % 18,623 9% Networks 33,562 26% 7,050 18% 1,797 19% (328) -29% 51,696 26% Music 5,477 4% 4,036 10% 419 4% (498) -44% 18,341 9% Publishing 18,283 14% 4,689 12% 909 9% (96) -8% 29,065 15% $127, % $39, % $9, % $1, % $196, % Panel B: Other Intangible assets Intangible assets subject to amortization Music catalogues and copyrights $2,927 40% Film library 3,363 46% Customer lists and other intangible assets % $7, % Intangible assets not subject to amortization Cable television franchises $26,574 70% Sports franchises 480 1% Brands, trademarks and other intangible assets 10,654 28% $37, % 8

9 Table 2 Changes in AOL Time Warner goodwill in 2002 Impact of accounting change Fourth quarter impairment Dec. 31, 2001 Additions Dec. 31, 2002 AOL $27,729 $8,536 $0 $(33,489) $2,776 Cable 33, (22,976) (10,550) 0 Filmed Entertainment 9,110 (71) (4,091) 0 4,948 Networks 33,562 (4) (13,077) 0 20,481 Music 5,477 (35) (4,796) (646) 0 Publishing 18,283 (243) (9,259) 0 8,781 $127,420 $8,803 $(54,199) $(44,685) $36,986 In addition, at the end of the fourth quarter, AOL Time Warner announced that it would have to take a further $44.7 billion charge for goodwill impairment, primarily against the AOL unit s allocation. These impairments ($98.9 billion) resulted in the complete write-down of goodwill in the cable and music businesses, and reduced goodwill for the AOL unit by 93%. Given the challenges in valuing the goodwill for each business unit, one has to wonder how these impairment allocations were generated. Why did the $54.2 billion impairment affect only the allocations to TW s segments and not those of AOL, particularly given concerns about the decline in the online subscription business with the collapse of the dot com market? Consistent with this concern, the company s 2002 performance, reported in table 3, showed a 38% decline in EBITDA for the AOL business unit, whereas the cable unit grew by 5% and Networks by 13%. 9

10 Table 3 Segment revenues and EBITDA for AOL Time Warner in 2002 Revenues EBITDA Percent change Percent change AOL $8,615 $ 9,094 6% $2,914 $1,798-38% Cable 6,028 7,035 17% 2,628 2,766 5% Filmed Entertainment 8,759 10,040 15% 1,017 1,232 21% Networks 7,050 7,655 9% 1,797 2,032 13% Music 4,036 4,205 4% % Publishing 4,689 5,422 16% 909 1,155 27% $39,177 $43,451 $9,684 $9,465 Timeliness of Impairments How timely were AOL Time Warner s impairments? To help answer this question, I conducted three analyses. The first examined stock price changes around the public announcements by the company of the impairments. The second examined the performance of AOL Time Warner subsequent to the impairments to determine whether its reported book returns exceeded its cost of capital. And the third examined the timing of the impairments relative to changes in the value of the company s stock. AOL Time Warner s public announcements of the impairment charges were made on January 7, 2002 (when it disclosed that the new accounting standard would require an impairment of between $40 billion and $60 billion), April 24, 2002 (when it reported the actual impairment under the accounting change of $54.2 billion), and January 29, 2003 (when it reported the additional $44.7 billion impairment). 10

11 The market reactions to the first two announcements were 0.2% and 2.0% respectively (versus -0.2% and -1.0% for the Nasdaq index), suggesting that the market had either anticipated the impact of the accounting change/impairment or did not view it as economically significant. Only the announcement of the $44.7 additional impairment elicited much reaction, with the company s stock dropping 12.1% (versus -1.5% for the Nasdaq), reflecting negative news conveyed by the impairment charge and AOL Time Warner s revenue and operating earnings news. Relative to the fourth quarter of 2001, AOL Time Warner s Q revenues declined by 25% and its operating earnings by 52%. If the impairment charges taken by AOL Time Warner represented an unbiased assessment of the decline in fair value of aggregate goodwill, one might anticipate that in subsequent years the company would be able to earn a book return on assets that at least equaled its cost of capital. Table 4 reports book returns on beginning net operating assets from 2002 to 2009, the year the AOL business was spun off. Despite the large goodwill impairment charge in 2002, AOL Time Warner reported ROAs ranging from -15.4% to 6.2%, with an average of 0.4%. After excluding 2008, when the company took another large impairment, the average ROA was only 3%, consistently lower than any plausible cost of capital. AOL Time Warner, therefore, continued to underperform following the large impairment in 2002, indicating that its assets (and intangibles in particular) continued to be overvalued. In 2008, AOL Time Warner formally recognized that its intangible assets were overvalued, and took a $2.2 billion impairment on AOL goodwill, a $7.1 billion 11

12 charge for the impairment of publishing goodwill and identifiable intangible assets, and a $14.8 billion impairment of cable franchise rights. The impairment of the cable franchise is noteworthy because its value reflected the revaluation at the time of the AOL deal. Table 4 Operating ROA for AOL Time Warner ROA Intangible assets Total net operating assets % 169, , % 81,192 83, % 83,639 85, % 83,215 83, % 83,940 84, % 92,903 98, % 91,692 98, % 68,278 79, % 38,935 44,927 ROA is defined as net operating profit after tax divided by beginning of year net operating assets. Net operating profit after tax is net income plus the after-tax cost of net interest. Net operating assets are working capital plus long-term assets less non-interest bearing long term debt. To further examine the timeliness of the 2008 intangible impairment I examine the market valuation of AOL Time Warner. Table 5 reports quarterly multiples of the company s enterprise value (EV) to book capital, 1 intangible assets, and intangible assets to book capital from 2001/Q1 (when the merger was first reported) to 2009/Q4. Throughout this period, intangibles ranged from 80-95% of total capital, signifying their economic significance. The EV-book capital multiple fell below one in the fourth quarter of 2001, suggesting that the market perceived that the company s intangibles were impaired. AOL Time Warner implicitly 1 Enterprise value is the market capitalization of equity plus the book value of net debt. Book capital is the book value of equity plus net debt. 12

13 acknowledged this one year later (2002/Q4) when it took the $44.6 billion write-off. The EV-book capital multiple remained above one until the fourth quarter of 2007, when it fell to 0.99, and declined steadily throughout This appeared to prompt the third significant impairment of $25.1 billion. However, even this appears to have been insufficient, since the multiple continued to be less than one until the first quarter of Overall, AOL Time Warner s recognition and impairment of intangibles maps into many of the findings in the academic literature. The company initially took a significant $55 billion impairment as a result of the change in accounting standards, reporting the change below the line. This announcement met with little market response. In contrast, the fourth quarter 2002 impairment of $44.6 billion, presumably in response to the decline in AOL s subscriber base, was accompanied by a material 12% drop in stock price, signifying that either the impairment and/or declines in quarterly revenues and operating earnings were surprises to investors. Finally, there is also evidence consistent with the company delaying further write-downs despite reporting ROAs well below the cost of capital for five consecutive years. In 2008, after one year of EC-book capital multiples below one, it opted to take another write-down for $25.1 billion, but even this appears to be too little to late, since the multiple continued to be well below one. 13

14 Table 4 History of AOL Time Warner valuation and intangibles Yr/Qtr EV/ Book capital multiple Intangibles Intangibles as % book capital 2001/ , % 2001/ , % 2001/ , % 2001/ , % 2002/ , % 2002/ , % 2002/ , % 2002/ , % 2003/ , % 2003/ , % 2003/ , % 2003/ , % 2004/ , % 2004/ , % 2004/ , % 2004/ , % 2005/ , % 2005/ , % 2005/ , % 2005/ , % 2006/ , % 2006/ , % 2006/ , % 2006/ , % 2007/ , % 2007/ , % 2007/ , % 2007/ , % 2008/ , % 2008/ , % 2008/ , % 2008/ , % 2009/ , % 2009/ , % 2009/ , % 2009/ , % 14

15 Discussion and Summary So what can we learn from the accounting for the AOL Time Warner merger? First, it reinforces the view that acquisitions are high-risk undertakings. Large acquisitions can be particularly challenging to implement successfully. Planned synergies can fail to materialize for a variety of reasons - conflicts in corporate cultures, changing technologies that threaten industry leaders, responses of competitors, etc. As a result, goodwill and the other intangible assets reported as a result of an acquisition are highly risky and likely to trigger subsequent impairment tests. Second, the practice of decomposing goodwill into business unit allocations, required by FAS 141, seems arbitrary given goodwill reflects the synergies from combining the firm s identifiable assets, and is not by itself a separable asset. For AOL Time Warner, synergies could arise from combining AOL s subscriber business with TW s cable and entertainment business, but it is difficult to identify which particular unit generated the goodwill. In addition, stand-alone goodwill for TW reflects synergies between its various business units. But again, the allocation of goodwill from these sources to individual business units seems arbitrary. Instead, it might be more useful for merging companies to report the share of goodwill applicable to the target as a stand-alone entity, and the portion attributable to projected synergies in the two businesses. The relative magnitude of these amounts would allow financial statement users to better understand the risks from an acquisition either in realizing the projected synergies or in the assessing whether the target was fairly valued as a stand-alone business. Given the difficulty 15

16 in separating these components after the merger has occurred, I suspect that it makes little sense to track these allocations over time and to assign impairments to specific mergers and these two goodwill components. Although, some companies do track actual merger synergies over time to understand successes and failures of merger investments. Third, goodwill impairments are unlikely to be timely events. Stock markets can respond quickly to new information on a firm s prospects, whereas accountants typically revalue intangibles once a year. As a result, investors are likely to anticipate that intangible assets are impaired prior to their being reported by management. Impairments are likely to confirm that reported poor performance is due to failures to achieve anticipated merger synergies or from overpayment for the target, rather than deteriorations in the acquiring firm s business. As such they help financial statements to reflect differences between successful and failed acquisitions. But they may not provide much new information to financial markets relative to more timely quarterly revenues and earnings that are used as inputs to judge whether impairments are called for. Fourth, the failure to impair intangibles after sustained poor performance, as was the case for AOL Time Warner, is one indicator that managers are reluctant to recognize the acquisition s failure. Concern about such delays lead some to advocate for systematically amortizing goodwill (also with impairment tests). This would have forced AOL Time Warner to write-down its goodwill and identifiable intangibles with indefinite lives between 2002 and Such a policy may be more conservative, but it is difficult to understand how a mechanical write-down of 16

17 intangibles provides information that helps financial statements to reflect differences between successful and failed acquisitions. Proponents of amortization argue that, like fixed assets, intangibles get used up over time and it is important to recognize their use as an expense. However, there is an important distinction between fixed assets and intangibles. New fixed asset outlays are reported as an asset, whereas outlays to preserve or enhance intangibles, such as advertising or R&D, are typically expensed. As a result, amortizing intangibles and simultaneously expensing outlays that preserve or grow their value effectively double counts their expense. Unless we decide to permit the capitalization of outlays that preserve and grow goodwill, which I would not advocate, the current system will inevitably lead to an inconsistency in the reporting of outlays for fixed assets and goodwill. Fifth, although research findings demonstrate that managers exercise judgment to delay impairments under FAS 142, it is worth remembering that management exercised considerable judgment under the prior standards. They could groom an acquisition to meet criteria for pooling of interests, which eliminated the risk of having to recognize goodwill impairments from merger failures. And managers of firms that used purchase accounting were subject to a less stringent impairment test and could exercise considerable reporting judgment to reduce the risk of impairments even when acquisitions had failed. Would AOL Time Warner s financial statements have reflected the merger s failure on a more timely basis had pooling been used? Or had purchase accounting been used but a less stringent impairment test and management judgment were available to defer 17

18 recognizing any loss? Despite concerns about the delays in impairment discussed above, I suspect the answer to both these questions would be no. In conclusion, additional analysis seems to be essential before we can draw strong conclusions about the effectiveness of FAS 141 and 142 in holding management more accountable for acquisition outcomes. The following are some questions that might be worth pursuing: Are there fewer under-performing acquisitions post-fas 141/142? Are managers of under-performing acquisitions disciplined more promptly for poor decisions (through early retirement or termination, investor activism or takeovers) under FAS 141/142? Do divestitures of failed acquisition targets take place more frequently or more quickly after FAS 142? Can field research allow us to better understand how managers justify deferring impairing intangibles for failed mergers? Given limitations on the timeliness of impairments, how do investors judge whether an acquisition has succeeded or failed on a timely basis? 18

19 References Beatty, A., and J. Weber Accounting Discretion in Fair Value Estimates: An Examination of SFAS 142 Goodwill Impairment. Journal of Accounting Research 44, Bens, D., W. Heltzer, and B. Segal The Information Content of Goodwill Impairments and SFAS 142. Journal of Accounting, Auditing and Finance 26, Chambers, D Has Goodwill Accounting Under SFAS 142 Improved Financial Reporting? Unpublished paper, University of Kentucky. Chen, C., M. Kohlbeck, and T. Warfield Timeliness of Impairment Recognition: Evidence from the Initial Adoption of SFAS 142. Advances in Accounting 24, Financial Accounting Standards Board (FASB) 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: FASB. Financial Accounting Standards Board (FASB) Statement of Financial Accounting Standards No. 141: Business Combinations. Norwalk, CT: FASB. Financial Accounting Standards Board (FASB) Statement of Financial Accounting Standards No. 142: Goodwill and Other Intangible Assets. Norwalk, CT: FASB. Hayn, C., and P. Hughes Leading Indicators of Goodwill Impairment. Journal of Accounting, Auditing and Finance 21, Healy, P., K. Palepu and R. Ruback Which Mergers are Profitable: Strategic or Financial. Solan Management Review 38, Holthausen, R., and R. Watts The Relevance of Value-Relevance Literature for Financial Accounting Standard Setting. Journal of Accounting and Economics 31, Li, Z., P. Shroff, R. Venkataraman, and X. Zhang Causes and Consequences of Goodwill Impairment Losses. Review of Accounting Studies 16, Li, K. and R. Sloan Has Goodwill Accounting Gone Bad? Unpublished paper, University of California Berkeley. Paugam, I., P. Astolfi and O. Ramond Accounting for Business Combinations: Do Purchase Allocations Matter? Journal of Accounting and Public Policy 34,

20 391. Ramanna, K The Implications of Unverifiable Fair-Value Accounting: Evidence from the Political Economy of Goodwill Accounting. Journal of Accounting and Economics 45, Ramanna, K., and R. Watts Evidence on the Use of Unverifiable Estimates in Required Goodwill Impairment. Review of Accounting Studies 17, Shalev, R., I. Zhang and Y. Zhang CEO Compensation and Fair Value Accounting: Evidence from Purchase Price Allocation. Journal of Accounting Research 51, Watts, R Conservatism in Accounting Part I: Explanation and Implications. Accounting Horizons 17,