Journal of accounting, auditing finance tập 28, số 01, 2013 1

Journal of Accounting, Auditing & Finance http://jaf.sagepub.com/ Dedication Journal of Accounting, Auditing & Finance 2013 28: DOI: 10.1177/0148558X12472129 The online version of this article can be found at: http://jaf.sagepub.com/content/28/1/3.citation Published by: http://www.sagepublications.com On behalf of: Sponsored by The Vincent C Ross Institute of Accounting Research, The Leonard N Stern School of Business Additional services and information for Journal of Accounting, Auditing & Finance can be found at: Email Alerts: http://jaf.sagepub.com/cgi/alerts Subscriptions: http://jaf.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Permissions: http://www.sagepub.com/journalsPermissions.nav >> Version of Record - Dec 14, 2012 What is This? Downloaded from jaf.sagepub.com at Taylor's University on December 22, 2012 Dedication Journal of Accounting, Auditing & Finance 28(1) The Author(s) 2013 Reprints and permission: sagepub.com/journalsPermissions.nav DOI: 10.1177/0148558X12472129 http://jaaf.sagepub.com This issue is dedicated to Professor Lee J Yao, PhD who passed away on November 14, 2012 due to complications from cancer at the age of 54 He was an expert in interdisciplinary research on forensic accounting, finance and information systems He published widely on issues in these diverse topics, resulting in three books and two book chapters, and more than forty articles in leading refereed journals His papers have appeared in several leading journals on Accounting, Finance, Electronic Markets, Accounting Information Systems, Accounting Education, International Accounting, Information Management, Computer Information Systems, and Quantitative Finance Downloaded from jaf.sagepub.com at Taylor's University on December 22, 2012 Not All That Glitters Is Gold: The Effect of Attention and Blogs on Investors’ Investing Behaviors Journal of Accounting, Auditing & Finance 28(1) 4–19 ÓThe Author(s) 2013 Reprints and permission: sagepub.com/journalsPermissions.nav DOI: 10.1177/0148558X12459606 http://jaaf.sagepub.com Nan Hu1, Yi Dong2, Ling Liu1, and Lee J Yao3 Abstract This article investigates the relationship between a firm’s visibility in blogspaces, termed blog exposure, and the cross-sectional stock returns We show that blog exposure is fundamentally different from the traditional media coverage, and securities with low blog exposure earn higher returns than stocks with high blog exposure We further illustrate that such an effect is more prominent for stocks with low institutional ownership Contrary to traditional media coverage, the return premium associated with blog exposure cannot be explained by either the illiquidity hypothesis or the investor recognition hypothesis based on the rational-agent framework Instead, our results suggest that blog effect can be attributed to the limited attention theory and cannot be arbitraged due to investors’ selfattribution and short-sale constraints Our research points out the importance of blogs in information dissemination, especially for the stocks with limited attention Keywords blog, expected return, limited attention, self-attribution, short-sale constraints, word of mouth communication Introduction In this article, we investigate whether a firm’s visibility within blogspaces, termed blog exposure, can affect security pricing even if the blog itself does not supply any authentic news When controlling for environmental factors such as traditional media coverage and analyst coverage, and various risk factors such as size, book-to-market (BM) ratio, beta, and momentum, our results indicate that stocks with high blog exposure earn lower returns than stocks with low blog exposure These results are more pronounced among stocks with lower institutional ownership We also show that a portfolio composed of stocks with low institutional ownership produces sustained future abnormal returns Such returns peak approximately 10 months after formation of the portfolio, and reverse sharply after that, University of Wisconsin, Eau Claire, USA University of International Business and Economics, Beijing, China Loyola University, New Orleans, USA Corresponding Author: Yi Dong, No 10, Huixin Dongjie, Chaoyang District, Beijing, China Email: docljy@gmail.com Downloaded from jaf.sagepub.com at Taylor's University on December 22, 2012 Hu et al displaying a long-term return reverse pattern In addition, the abnormal returns of a portfolio composed of stocks with high institutional ownership meander around zero over the next 12 months after formation of the portfolio We examine three plausible explanations for this return premium, two from the rational-agent framework (liquidity and investor recognition) and one from the behavior finance framework (limited attention with short-sale constraints) Surprisingly, contrary to media coverage, for which the return premium can be explained by liquidity and investor recognition based on the rational-agent framework (Merton, 1987), the return premium of blog exposure cannot be explained by that framework Instead, our findings show that a blog exposure premium originates from the joint forces of biased disclosure of bloggers and limited attention of consumers Such overvaluation becomes larger over time due to consumers’ biased self-attribution, while the shortsale constraints prevent such a premium from being arbitraged Our results are most consistent with the notion in Hirshleifer, Lim, and Teoh (2009) In addition, our results show that the blog exposure effect is not subsumed by traditional media exposure, as documented by Fang and Peress (2009) However, without mainstream media (e.g., news) planting a seed of discussion, blog exposure actually has no impact on security returns Understanding the relation between blog exposure and stock returns is very important for the marketing community and the finance community because it demonstrates how a firm’s marketing strategy within blogspaces can influence elements of its finance strategy, such as cost of capital As a form of word of mouth (WOM)1, blogs represent the fastest growing medium of personal publishing and the newest method of individual expression and opinion on the Internet.2 In 2004, blogs were a fairly new phenomenon with only million bloggers worldwide (Wright, 2006) However, at the time of writing this article, according to www.BlogPulse.com, there were more than 126 million blogs on the World Wide Web We believe blogs are playing a role that is as important as that of newspapers because (a) information in a blog is not a simple reflection of what is covered by traditional news In fact, many blogs address topics that are not covered by the mainstream media at all Blogs might either lead or follow traditional news and (b) blogs disseminate information to a much broader audience faster and with in-depth analysis In fact, compared with other online media, blogs are viewed as more credible In addition, compared with traditional sources, more than three quarters of respondents view blogs as moderate to very credible This article adds to the growing body of studies of the valuation of online WOM literature (Antweiler & Frank, 2004; Tumarkin & Whitelaw, 2001; Wysocki, 1999) Our article is also closely related to but distinct from Fang and Peress (2009)’s media coverage study that shows that, by helping to reach a broad population of investors, mass media can alleviate information frictions and affect stock price even if it does not contain authentic news However, Fang and Peress focus on studying the impact of traditional media coverage, measured by the number of newspaper articles about a firm on its stock returns We concentrate on examining the influence of nontraditional media coverage—blog exposure—on security pricing, while controlling for the traditional media coverage Furthermore, our results are distinct from Fang and Peress’ conclusions The return premium of news coverage can be explained by the rational-agent framework (Fang & Peress, 2009) However, such a return premium of blog exposure can only be explained by the joint forces of investors’ behavior and short-sale constraints Another branch of literature related to our study is the behavior finance literature that recognizes that attention influences investors’ selling and purchasing behavior, and causes asset pricing deviation from its fundamental value The underlying reason is that investors Downloaded from jaf.sagepub.com at Taylor's University on December 22, 2012 Journal of Accounting, Auditing & Finance face a formidable search problem when buying a stock Investors address such an issue by limiting their choice sets (Barber & Odean, 2008) to those stocks that have recently caught their attention (Odean, 1999) Because investors are overconfident (Daniel, Hirshleifer, & Subrahmanyam, 1998; Odean, 1998) and biased toward self-attribution (Daniel et al., 1998), stocks over bought heavily by individual investors will enjoy short-term positive contemporaneous returns These results emphasize the attention driven by traditional media, such as securities mentioned in the news or securities that have gone through large volume or price changes Our study, however, focuses on the attention driven by nontraditional media, blogs One practical implication of our results is that a firm’s visibility within blogspaces, regardless of whether the blog discussion is positive or negative, influences investors’ purchase decision As a nontraditional media, blog discussions serve the role of disseminating information that was traditionally covered by conventional information channels such as analysts’ forecasts or newspapers Such a role is especially important for stocks with limited attention In fact, as we documented, because blog discussions not affect the expected stock returns when there is a lack of echo from the mainstream media, companies should plant the seed of a discussion to foster the conversation within blogspaces Marketing managers of a firm can use blogs not only to communicate more efficiently with its customers, partners, suppliers, and other stakeholders but also to work closely with finance managers to lower the cost of capital by delivering information to a broader audience in a faster manner The rest of this article is organized as follows In the next section, we discuss our data collection processes, elaborate our variable definitions, and present our empirical results In the section ‘‘Explaining the Blog Visibility Effect,’’ we discuss three possible causes of the blog exposure effect, and in the section ‘‘Conclusion,’’ we present our concluding remarks Data Collection and Empirical Results Data Collection In this article, blog visibility/exposure is defined as the extent to which a company’s products or services are discussed in blogspaces We collect such a measurement from www.BlogPulse.com using its conversation track tool Our data collection is composed of two steps In Step 1, we identify the brand names of products or services associated with a company by searching a company’s web site, reading its financial reports, or using researchers’ domain knowledge In Step 2, using the names identified in Step as keywords, for each company, we retrieve its blog visibility over time using the conversation tracker tool provided by www.BlogPulse.com Of the total daily blogging activities traced by www.BlogPulse.com, this measurement represents the percentage of the total blog conversation related to a particular firm, its products, or its services It measures how (and how much) a firm’s current customers, potential customers, competitors, industry peers, and so on are talking about the products or services of the firm Hence, it represents the visibility of a firm within the overall blogspaces Figure shows one example of the blog trend for Advance Auto Parts Inc.3 The data on a firm’s media exposure were collected from Factiva Factiva is a database of the Dow Jones and Reuters companies It provides timely, domestic, and international information, such as articles from the Dow Jones and Reuter’s newswires and The Wall Street Journal This information covers market data, firm and industry news, financial Downloaded from jaf.sagepub.com at Taylor's University on December 22, 2012 Hu et al Figure Blog coverage data collection Note: In this figure, we use Advance Auto Parts Inc as an example to show how we collect the exposure of a firm within blogspaces using www.blogpulse.com quotes, and newspaper articles We collected and counted all the nonredundant news items each day for the fiscal year of each company Our analysts’ forecast data were collected from the Institutional Brokers Estimates System The company accounting data were obtained from CompuStat, and the stock return data were from the Center for Research in Security Prices (CRSP) For detailed variable definitions, please refer to the appendix Following Amihud (2002), we delete one firm with extreme illiquidity value The final sample contains 404 Standard & Poor (S&P) 500 firms with daily blog visibility and traditional media coverage information from March 1, 2006, to August 22, 2006 We give detailed variable definitions in the appendix Following Fang and Peress (2009), to minimize the noise of daily data, for each firm, we aggregate its daily blog visibility and Factiva news to a monthly level to represent its blog visibility and media coverage, respectively Table shows the summary statistics of our key variables, including blog coverage and media coverage As we can see, only 1% of the stocks in our sample not have media coverage, but more than 25% of the stocks in our sample not have blog coverage It seems that for the S&P 500 firms, and for this sample period, traditional media has broader coverage in terms of the number of stocks discussed than blog conversations However, this does not necessarily mean that, in general, traditional news media has broader coverage than blogs Table shows the Pearson correlation among blog coverage, media coverage, and other firm characteristics We find blog coverage and media coverage are positively correlated, and analysts, news, and blogs have the tendency to feature the same set of stocks This is reasonable because analysts, news, and blogs pay attention to large firms and well-known firms Furthermore, it seems that traditional media cares more about firms with a lower Downloaded from jaf.sagepub.com at Taylor's University on December 22, 2012 Journal of Accounting, Auditing & Finance Table Summary Statistics of Blog Exposure and Media (Factiva) Coverage Blog M SD Skewness Quintile 100% maximum 99% 95% 90% 75% Q3 50% median 25% Q1 10% 5% 1% 0% minimum Factiva 0.40 1.22 5.66 11.89 7.38 1.90 0.80 0.22 0.04 0.00 0.00 0.00 0.00 0.00 M SD Skewness Quintile 100% maximum 99% 95% 90% 75% Q3 50% median 25% Q1 10% 5% 1% 0% minimum 236.42 415.45 3.50 2,791.00 2,267.00 1,085.00 535.00 224.00 95.00 41.00 19.00 7.00 0.00 0.00 Note: See the appendix for detailed variable definitions institutional ownership, whereas bloggers not differentiate between firms with high institutional ownership and those with low institutional ownership Blog Coverage and Short-Term Cross-Section of Stock Return In this section, we study whether a firm’s exposure within blogspaces affects its security pricing by regressing stock returns on blog coverage, with media coverage, beta, size, BM ratio, and other risk factors controlled (Table 3, Model 1) For a robustness check, we take an incremental approach by adding momentum, media coverage, percentage of institutional ownership, and illiquidity one by one, and present the results in Models 2, 3, and of Table 3, respectively To control for the potential confounding effect caused by the difference across months and industries, we also include month fixed effect and industry fixed effect (two-digital standard industrial classification from CRSP) in all our models For Model 1, the coefficient of blog exposure is 20.0026 with t-value at 21.93, which is negatively associated with the following month’s stock return Therefore, stocks with high blog exposure tend to have lower stock returns compared with the stocks with low blog coverage Such an impact is not subsumed when we add the traditional media exposure and other risk factors (Models and 4) Furthermore, using Factiva as proxy for media coverage, we found that media coverage has an insignificant coefficient regardless of whether we exclude the blog exposure The untabulated table shows that our results are qualitatively the same if we exclude those firm-month observations in which there are earning announcements for the firm in that month Therefore, we conclude that the blog effect is not driven by month effect, industry effect, or earnings announcements Our media coverage results are different from those of Fang and Peress (2009) However, in their article, they focus on four influential newspapers with large subscriptions Our media coverage includes all the newspapers included in the Factiva database Another reason for the different results might lie in the sample selection In their article, Downloaded from jaf.sagepub.com at Taylor's University on December 22, 2012 Hu et al Table Pearson Correlation Blog 2353 Factiva 0.1550 \0.0001 2,353 Size 0.1334 \0.0001 2,346 BM 20.1214 \0.0001 2,261 Momentum 20.0740 0.0003 2,352 Dispersion 0.0125 0.5513 2,266 IO 0.0088 0.6695 2,353 Coverage 0.1293 \0.0001 2,281 Illiquidity 20.0212 0.3043 2,353 Idiorisk 20.0038 0.8558 Factiva Size BM Momentum Dispersion IO Coverage Illiquidity Idiorisk Blog 2,353 0.5198 \0.0001 2,346 20.0383 0.0685 2261 20.0242 0.24 2,352 0.0050 0.8131 2,266 20.1053 \0.0001 2,353 0.3358 \0.0001 2,281 20.0369 0.0732 2,353 20.0725 0.0004 2,346 20.2583 \0.0001 2,259 0.0886 \0.0001 2,345 0.0076 0.7169 2,263 0.0614 0.0029 2,346 0.6064 \0.0001 2,278 20.2677 \0.0001 2,346 20.3569 \0.0001 2,261 20.1616 \0.0001 2,261 20.0384 0.0722 2,191 20.0905 \0.0001 2,261 20.1358 \0.0001 2,201 0.0348 0.0983 2,261 20.0612 0.0036 2,352 0.0218 0.3008 2,266 0.2115 \0.0001 2,352 0.0195 0.3528 2,281 20.0602 0.0035 2,352 0.1148 \0.0001 2,266 20.0500 0.0173 2,266 0.0012 0.9549 2,266 20.0399 0.0578 2,266 0.0035 0.8681 2,353 0.0353 0.0916 2,281 20.1454 \0.0001 2,353 0.0578 0.005 2,281 20.1501 \0.0001 2,281 20.0635 0.0024 2,353 0.2501 \0.0001 Note: BM = book to market; IO = institutional ownership For detailed variable definitions, refer to the appendix they argue that the most significance of high media coverage, low return effect comes from the small and illiquid firms However, in our article, we use only S&P 500 firms, which are all big firms Blog Coverage and Long-Term Cross-Section of Stock Returns In the previous section, we documented the short-term return premium associated with low blog exposure firms In this section, we study the long-term return impact of blog exposure Figure represents the cumulative abnormal returns of stocks with different blog coverage, starting from Month after the portfolio formation Each month, we sort our sample into three groups according to their monthly blog exposure Then, based on capital asset pricing model, for each stock, we estimate its abnormal return in the subsequent 13 months Low blog curve (Figure 2) represents the average abnormal returns for the lowest blog coverage group, whereas high blog curve represents those of the highest blog coverage group Figure shows that the highest blog coverage portfolio consistently has insignificant abnormal returns starting from the 1st month after the formation of the portfolio to the next 13 Downloaded from jaf.sagepub.com at Taylor's University on December 22, 2012 10 Journal of Accounting, Auditing & Finance Table Blog Coverage, Media Coverage, and Stock Returns Model Intercept Blog Factiva Beta Size BM Momentum Coverage IO Illiquidity Adjust_R2 0.0116 (20.45) 20.0026 (21.93) 20.0026 (21.1) 0.0024 (1.72) 0.0042 (1.78) 1119 Model 0.0295 (21.03) 0.0000 (20.87) 20.0028 (21.78) 0.0012 (0.70) 0.0040 (1.72) 1107 Model 20.0043 20.0033 0.0000 0.0003 0.0021 0.0012 20.0323 (20.15) (22.45) (20.65) (20.13) (1.23) (0.51) (25.79) 1253 Model 0.0041 20.0033 0.0000 0.0002 0.0019 0.0013 20.0311 20.0001 20.0034 20.0203 1246 (20.12) (22.39) (20.65) (20.06) (0.87) (0.56) (25.36) (20.25) (20.68) (20.95) Note: BM = book to market; IO = institutional ownership; SIC = standard industrial classification In this table, the dependent variable is the following month’s stock return The independent variables include traditional risk factors and other firm characteristics that may affect expected stock returns Our samples include 404 distinct firms We run pooled regression, controlling month fixed effect and industry fixed effect (2-digital SIC) Nextmonthre5a bbeta beta bsize size bBM BM bmomentum momentum bIO IO bcoverage coverage billiquility illiquility bblog blog bFactiva Factiva: Figure Cumulative abnormal return Note: CAPM = capital asset pricing model; CAR = coverage abnormal return Figure presents the cumulative abnormal return of a high blog coverage sample and a low blog coverage sample starting from Month after the portfolio formation Each month we sort our sample into two groups according to its monthly blog coverage Then, based on CAPM model, for each stock, we estimate its cumulative abnormal return in the subsequent 13 months The blue line represents the average CAR for the low blog coverage group, whereas the red line plots that of the high blog coverage group months, and the low blog coverage portfolio enjoys a positive coverage abnormal return in the following 13 months In addition, there is a return reversal when the cumulative abnormal returns peak at the 10th month Downloaded from jaf.sagepub.com at Taylor's University on December 22, 2012 84 Journal of Accounting, Auditing & Finance apparent bias in estimating fair values Finally, Ryan (2008) and Chen, Liu, and Ryan (2008) address the roles played by certain accounting issues in the subprime crisis We add another dimension to these latter articles by showing the inherent conflicts behind the measurements of fair values In sum, our research suggests that caution be exercised in interpreting results in studies (future and existing studies) that use or defend fair values Although our sample is relatively small in number, it nonetheless represents virtually the entire industry of credit card securitizers, measured in terms of total assets For instance, our sample includes of the top 50 Bank Holding Companies, and of our sample firms (Bank of America, JP Morgan, and Citigroup) are the top in terms of total assets (US$6.2 trillion, combined).5 Thus, our study has implications for the existing studies mentioned above (all of which specifically address securitizations) even though we focus on credit card securitizers We believe that the severity of the inherent agency conflict in the particular industry that we study should not be interpreted to mean that the conflicts are less severe or confined to this particular industry Thus, we contend that the use of (or continued use of) specific statistical controls for the agency environment are critical for any future study that uses fair values Sales Accounting and Transition Choices Sales Accounting In sales accounting, the transferor typically transfers a pool of financial assets (receivables) to another entity, the transferee, who pays the transferor for the financial assets with capital obtained through issuances of debt or equity to third-party investors The financial assets held by the transferee provide collateral for the securities sold to the investors The transferor often retains a slice (tranche) of the receivables The retention of a tranche in the receivables often resulted in credit and liquidity risk remaining with the transferor This retention of risk apparently allowed the transferor to obtain more attractive financing terms In fact, this is a primary explanation for the transferor being able to sell the tranches at values greater than their assigned cost, that is, at a profit, suggesting that the parts are greater than the whole Also, the transferee was structured as a bankruptcy remote vehicle, which in combination with the transferor’s retention of risk, contributed to the transferor being able to extract a profit In essence, these two protections made third-party investors willing to pay a premium for the cash flows It is surprising that the gains on sales of receivables did not receive more descriptive disclosure about how they originated and were measured Firms are required under SFAS No 157 (FASB, 2006) to value these tranches for accounting purposes, preferably with reference to prices in the market place (even if the market was problematic in such pricing) Where market prices were absent or inappropriate, firms applied the methodologies either for Level or Level assets, both of which, however, gave managers considerable discretion and possibilities for opportunistic pricing In addition, credit rating agencies’ conflicted relationships with transferors arguably affected the estimates of credit ratings in a biased way Therefore, it is surprising, admittedly in hindsight, to observe the degree of trust placed in firms to measure fair values, of the transferred receivables and those retained, because of the degree of agency conflict Under certain circumstances, the transferee was officially known as a ‘‘Qualifying Special Purpose Entity’’ (QSPE) The transferee was deemed a QSPE when the outside creditors (investors) who are holders of the beneficial interests are allowed to pledge or Downloaded from jaf.sagepub.com at Taylor's University on December 22, 2012 Bryan and Lilien 85 transfer their interest Other criteria for QSPE status are that the transferee is a passive entity, and, as mentioned, the entity is bankruptcy remote, so that the transferred assets are separate from the transferor Until recently, QSPEs were stand-alone entities and were not consolidated with the transferors Meeting requirements for QSPE status was quite attractive, because the transferor no longer had to consider the need for consolidation and therefore could automatically use sales accounting and raise financing off-balance sheet.6 Even though creditors did not directly buy the receivables, they bought interests that were backed by the receivables and the interests flowed through the QSPE back to the transferor When the criteria for QSPE status (and for sales accounting) were not met, transferors would evaluate whether the transferee required consolidation If consolidation was deemed to be necessary, the transferor would consolidate the receivables and the debt issued by the transferee and use financing accounting During the financial crisis, regulators and standard setters came to believe that some firms were misapplying the QSPE criteria, resulting in many securitizations occurring in entities that were not passive.7 Therefore, the FASB issued standards which have essentially brought about the demise of the QSPE and the subsequent reconsolidation of the transferred receivables, as well as recognition of the related financing that had been on the QSPE’s books Entities that were previously automatically exempted from consideration for consolidation now had to be assessed SFAS No 167 (FASB, 2009b) contains a series of tests to see whether a transferor can bypass consolidation If these tests are met, then the firm would determine if sales accounting could still be used under SFAS No 166 (FASB, 2009a) Sales accounting for financial assets such as receivables can be thought of as bifurcation of financial assets into ‘‘component parts.’’ These parts consist of receivables that are retained versus those that are sold The difference between the values attributed to those component parts that are considered sold and their ultimate selling price is recognized as a gain on the sale Thus, sales accounting is often referred to as ‘‘component part’’ accounting or as ‘‘bifurcation’’ accounting Transition Choices For enterprises required to consolidate an entity as a result of SFAS No 167, standard setters gave firms three transition choices, described below.8 The consolidating firm (the transferor) would measure the assets, liabilities, and noncontrolling interests of the transferee at amounts equal to those as if financing had been used from the beginning The transferor would consolidate the transferee’s receivables and debt and would adjust Retained Earnings for all previously recognized income effects, including the gain on the sale of the receivables This choice is difficult to implement due to the difficulty of determining those historical values for securitizations that occurred many years ago However, in our sample, General Electric used this transition method The reconsolidated receivables may require a provision for bad debt The reporting enterprise could alternatively consolidate and present assets, liabilities, and noncontrolling interests at fair value In our sample, CompuCredit used this method If the activities of the transferee were primarily related to securitizations or other forms of asset-backed financings, and if the assets could be used only to settle Downloaded from jaf.sagepub.com at Taylor's University on December 22, 2012 86 Journal of Accounting, Auditing & Finance Table Transfer of Receivable to QSPE Assets 140.0 228.6 Liabilities Owners’ equity 111.4 Cash Receivables Gain Table Reclassification of Receivable to Retained Interest (Investment) Assets 121.4 221.4 Liabilities Owners’ equity Retained interest Receivable Table Fair Value Adjustment of Retained Interest Assets 18.6 Liabilities Retained interest Owners’ equity 18.6 Gain obligations of the reporting entity, then the assets and liabilities could be measured at their unpaid principal balances This reporting choice is used by the remaining firms in our sample It is also the method that many firms in the financial services sector lobbied for, as revealed in their comment letters Simplified Examples of Transition Choices The following simplified illustration shows the differences among the transition choices However, we begin with the securitization itself Assume a firm (the transferor) has receivables with a total book value of US$50 The fair value of the receivables if broken into tranches is deemed to be US$70 of which US$40 is transferred to a QSPE The assigned cost of the receivable sold (‘‘cost of goods sold’’) is US$28.6 (40 / 70 50) In Table 1, we give the entry (in intuitive form, rather than conventional ‘‘debit and credit’’ form) to transfer (sell) the receivables to the transferee The transferee buys the receivables for US$40 with cash raised through a debt offering (or other beneficial interest) We assume in our illustration that the transferee issued debt of US$40 The transferor reclassifies the retained tranche from a receivable to ‘‘Retained Interest,’’ which is an investment (such as Trading or Available-for-Sale) on its books The value of the Retained Interest is US$21.4 (30 / 70 50) (Our allocations of cost not consider the issues of ‘‘participating interests’’ under revisions to accounting standards; Table 2.) We assume in this example that the transferor marks to market through profit-and-loss its retained tranche to US$30 (The fair value of the Receivables was estimated at US$70, US$40 of which was sold to the transferee.) The fair value adjustment (see Table 3) is therefore 18.6 (30 – 21.4) Downloaded from jaf.sagepub.com at Taylor's University on December 22, 2012 Bryan and Lilien 87 Table Re-recognition of Receivable and Consolidation of Debt (Transition Choice 1) Assets 128.6 Liabilities 140.0 Receivable Owners’ equity Debt 211.4 Retained earnings Table Reclassification of Retained Interest to Receivable (Transition Choice 1) Assets 221.4 121.4 Liabilities Owners’ equity Retained interest Receivable Table Reversal of Fair Value Adjustment of Retained Interest (Transition Choice 1) Assets 28.6 Liabilities Retained interest Owners’ equity 28.6 Retained earningsa a If the firm had classified the Retained Interest as Available-for-Sale, the offset would have been to Accumulated Other Comprehensive Income, rather than Retained Earnings If the financial asset had been designated as held-tomaturity, the accounting would have been at amortized cost, unless there was an ‘‘other than temporary’’ impairment Transition Choice 1: Reversal of Securitization As If Financing Had Been Used For the first transition choice, the transferor will undo the above in the following way First, the transferor records the return of the receivable and also records the debt issued by the transferee Retained Earnings is reduced for the amount of the gain on the sale (Table 4) Next, the transferor also reclassifies the Retained Interest back to status as a Receivable (Table 5) Finally, the transferor will undo the mark-to-market on the Retained Interest (Table 6) Under this transition method, the firm simply reverses the prior entries The assets are returned, the debt is recognized, and the earnings effects are undone, albeit not through current income, but rather through Retained Earnings.9 A major concern arising from this method is the potential for double counting earnings When sales accounting was initially used, the firm booked a gain on the sale of US$11.4 This transition method reverses the gain against Retained Earnings Therefore, the undoing of the income effect is not shown in current earnings Subsequently, the assets that are reconsolidated on the transferor’s books will produce income again Namely, the pool of receivables will earn interest income that will be booked in current and future Income Statements In essence, therefore, income produced from the same pool of assets will be double counted One consequence of double counting earnings relates to compensatory arrangements and firm valuations To the extent that employees are compensated on the same earnings twice or that financial models use the same earnings twice, distortions in incentive effects and firm valuations will naturally result (Another problem, which we discuss in Transition Choice 3, is the provisioning for loan losses.) Downloaded from jaf.sagepub.com at Taylor's University on December 22, 2012 88 Journal of Accounting, Auditing & Finance Table Re-recognition of Receivable and Consolidation of Debt (Transition Choice 2) Assets 140.0 Liabilities Receivable 140.0 Owners’ equity Debt Table Reclassification of Retained Interest to Receivable (Transition Choice 2) Assets 230.0 130.0 Liabilities Owners’ equity Retained interest Receivable Transition Choice 2: Reversal of Securitization Using Fair Value Now assume that the transferor uses fair values in undoing the securitization We assume that the total fair value of the receivables continues to be US$70 The transferor will reconsolidate the Receivable back on its books at its fair value of US$40 (The transferee had purchased the receivables at fair value or US$40.) The transferor will also consolidate the debt issued by the transferee to buy the receivables (Table 7) The other tranche of the Receivable is on the books of the transferor as Retained Interest, valued at US$30 We reclassify the Retained Interest as a Receivable (Table 8) Arriving at fair values for these assets can be extremely difficult Due to the lack of an observed price in liquid markets, the firm would make assumptions and use internal models Transition Choice 3: Reversal of Securitization Using Current Principal Balances The final transition choice allows firms to use the current principal balances of the receivable and the debt held by the transferee This transition choice is on the surface the simplest for transferors to implement because it entails consolidating current balances, as opposed to initial balances (which could be several years in the past) or balances at fair values (which would require using a valuation technique or model) However, this choice is not simple because the receivables must be consolidated at values that are adjusted for loan loss provisions That is, an allowance for nonperforming assets (uncollectible accounts) must be measured (As we mentioned above, the first transition choice shares this attribute of having to estimate bad debt on the reconsolidated receivables.) As we will show in our sample, these allowances were of tremendous magnitude and heretofore unknown because of lack of reporting requirements for the QSPEs For this simplified example, assume that the transferee had collected US$15 of the receivable and had reduced the principal of the debt by an equal amount Thus, the principal balances of the Receivable and Debt are both US$25 (40 – 15) We ignore both interest earned and interest paid as it is not germane for our purposes The transferor would book the following to undo sales accounting and consolidate the transferee (Table 9) The transferor would also undo the gain from the sale of the receivable and reduce the receivable from its fair value to its cost The reduction of US$11.4 (40.0 – 28.6) is recognized below (Table 10) Downloaded from jaf.sagepub.com at Taylor's University on December 22, 2012 Bryan and Lilien 89 Table Re-recognition of Receivable and Consolidation of Debt (Transition Choice 3) Assets 125.0 Liabilities Receivable 125.0 Owners’ equity Debt Table 10 Reversal of Gain on Transfer of Receivable (Transition Choice 3) Assets 211.4 Liabilities Owners’ equity 211.4 Receivable Retained earnings Table 11 Reversal of Fair Value Adjustment and Reclassification of Retained Interest (Transition Choice 3) Assets 230.0 121.4 Liabilities Owners’ equity 28.6 Retained interest Receivable Retained earnings Table 12 Recognition of Bad Debt Allowance, Net of Tax Assets 210.0 13.0 Liabilities Allowance Deferred tax asset Owners’ equity 27.0 Retained earnings The transferor would undo the mark to market through Retained Earnings and reclassify the net amount of Retained Interest as Receivable (Table 11) The balance in the Receivable is US$35 Its book value began as US$50 It was written up to US$70 Of the US$70, US$11.4 adjustment was held on the books of the transferee and US$8.6 on the books of the transferor Upon adoption of the new standard and this particular transition choice, the transferor writes down the receivable by US$20 (the sum of US$11.4 and US$8.6) to take the value back to the book value of US$50 The transferee’s collection of US$15 of cash reduces the carrying value to US$35 However, now assume that the reconsolidated receivables need to be adjusted by a loss provision (Allowance) of US$10 (assumed) The Allowance represents a future deduction Therefore, a Deferred Tax Asset is also recorded, and we assume a 30% tax rate for ease of calculation (Table 12).10 The problem with this method (as well as the first) is that the provision for loan losses will not pass through the Income Statement of the transferor Rather, the charge for the provision avoids income recognition by going directly to Retained Earnings.11 (As we illustrate in the Discover case below, the provision for bad debt is the largest adjustment in the entry in the transition.) As is the case with any provisioning, if the provisions are overaccrued, future income will be enhanced as the overaccrued allowance is released The likelihood of overaccruing is arguably increased in this instance because earnings are not affected by the provisioning in the first place A very possible consequence is that income will be earned in Downloaded from jaf.sagepub.com at Taylor's University on December 22, 2012 90 Journal of Accounting, Auditing & Finance three different ways on the same assets: (a) gain on the initial sale, (b) accretion of interest on the reconsolidated receivable, and (c) reversals of provisions on receivables The problem is illustrated below by Chimera Investment Company (2009 10-K, filed with the Securities and Exchange Commission): The consolidation is expected to reverse approximately $98.1 million in previously recorded GAAP gains on sales of assets related to the re-securitizations undertaken in 2009 The reversal of this gain will be accreted into interest income over the remaining life of the re-securitized assets (p 61) Chimera alludes to the compensation issue It states in the same filing that the ‘‘management fee’’ is based on its stockholders’ equity, adjusted for one-time effects from changes in accounting principles Thus, any charge to retained earnings would appear to be removed Although Chimera does not explicitly state that the compensation base would be affected (increased) by accreted interest (as it is ongoing), one could arguably draw that precise conclusion If accreted interest is included in the compensation base, the management fee would be based twice upon earnings from the same asset pool This is problematic from an investor’s viewpoint and from a governance viewpoint In conclusion, all three transition methods come with issues The first and third transition choices discussed above have the double (triple)-counting problem, and the second choice has the problem of estimating fair values In addition, the standard setters have not addressed in a clear way how the valuations of component parts can total more than the pool of receivables as a single unit As these component parts are reconsolidated on the balance sheets of the transferors, this previously recognized ‘‘premium’’ has disappeared inside Retained Earnings without understanding its source in the first place The issue is important because conceivably other asset pools representing future cash flows could be ‘‘tranched,’’ then revalued with a similar type premium, and finally reconsolidated under a mixed attribute model allowing for different accounting treatments for the reconsolidated components A final fundamental issue is how the accounting standards allowed transferors to book loan loss provisions in a way that is less than fully transparent (by using Retained Earnings) Even though problems about the creditworthiness of the receivables began in earlier time periods (while at the QSPE and before reconsolidation), the transferor was the originator of the receivable as well as the creator of the QSPE Therefore, a reasonable question would be whether the transferor should not book in current income the charge for the provision, especially because in many cases, the transferor also retained servicing rights for the receivables and would therefore have some direct knowledge of the performance of the receivables Discover Financial and Its Peer Group Discover Financial We now turn the case of Discover Financial Services (Discover) In this section, we also show the findings for our sample of the other major firms in the sector Discover is a good example to understand the accounting treatment because it is ‘‘pure play’’ credit card company, and it is no longer a division of Morgan Stanley.12 On December 1, 2009, Discover consolidated its credit card securitizations, which were previously accounted for as sales They chose to use carrying amounts as of December 1, Downloaded from jaf.sagepub.com at Taylor's University on December 22, 2012 Bryan and Lilien 91 Table 13 Re-recognition of Discover’s Receivables Assets 127.0 22.1 20.1 22.2 22.3 11.2 10.8 Liabilities Receivables Allowance I/O strip AFS HTM Other assets Deferred tax asset 122.3 11.3 Debt Other Owners’ equity 21.3 Retained earnings Note: I/O = interest-only; AFS = available-for-sale securities; HTM = held-to-maturity securities 2009 Above, we recreate the entry to consolidate the transferee The amounts are taken from Discover’s 10-K (Table 13) (We highlight the loan loss provision and discuss further below.) The adoption resulted in recognizing credit card receivables of US$27 billion and debt of US$22.3 billion Retained interests that consisted of interest-only (I/O) strips, availablefor-sale securities (AFS) and held-to-maturity securities (HTM) were also reclassified as part of the receivables.13 The adoption resulted in a provision for loan losses, net of tax effect, of US$1.3 billion, as shown below and highlighted above: –2.1 Allowance 0.8 Deferred Tax Asset = –1.3 Retained Earnings Thus, the entire reduction in retained earnings equals the newly created loan loss provision, net of income taxes Discover’s underperforming receivables are revealed for the first time upon transition from sales accounting to financing accounting The transition charge of US$1.3 billion is against the ‘‘past’’ (Retained Earnings) and will not ever flow through the Income Statement Also, if the US$1.3 billion is attributable to third-party beneficial interests (that is, with recourse to the investors), then a note disclosure would appear to be appropriate, but none was made in this instance Therefore, we infer that the transferor bears the responsibility for the underperforming assets, which calls into question the relationship between the transferor and the QSPE all along.14 However, SFAS No 140 (FASB, 2000) was based on a controls model and not a risks-and-rewards model, so the transferor could retain credit risk and still use sales accounting Under Generally Accepted Accounting Principles (GAAP), AFS are accounted for at fair value with any unrealized gains and losses recognized in Accumulated Other Comprehensive Income (AOCI) An impairment loss is recognized if deemed other than temporary and the firm did not have the intent and ability to hold this instrument to recovery.15 Held-to-maturity investments are carried at amortized cost and tested for impairment based on estimates of probable incurred losses Finally, credit card portfolios are accounted for using a loan loss model under SFAS No We would expect the unrealized losses in AOCI and impairments on HTM to capture the additional loan loss provisions, when losses are borne by the transferor (i.e., holder of retained interests) The fair value of retained interests should reflect impairments in underlying securities If losses are attributed to the third-party transferee, the eventual gain on write down of notes payable owed to the transferee offsets the additional loan loss provision taken earlier However, this may result in an accounting mismatch, because loan losses are recognized at date of adoption through retained earnings The gains on writing-off notes payable flow through income in future periods Downloaded from jaf.sagepub.com at Taylor's University on December 22, 2012 92 Journal of Accounting, Auditing & Finance Table 14 Discover’s Investments at Fair Value At November 30, 2009 Available-for-sale investment securities Certificated retained interests in DCENT Credit card asset-backed securities of other issuers Held-to-maturity investment securities Certificated retained interests in DCENT and DCMT State and political subdivisions of states Less than 12 months More than 12 months Fair value Unrealized losses Fair value Unrealized losses 1,149,143 (115,857) 889,848 (10,152) 127,509 (34) 1,865,484 (430,655) 51,778 (6,162) Note: DCENT = Discover Card Master Trust; DCMT = Discover Card Execution Note Trust As shown in the exhibit from Discover (Table 14), total unrealized losses sum to US$562,860 (115,857 34 430,655 10,152 6,162) This amount is approximately 27.0% of the provision of US$2.1 billion Although the QSPEs are now largely recognized on the books of the transferor, better disclosure would have been helpful in understanding the economics of the transactions between the transferor and the transferee Looking ahead, enhanced disclosure will facilitate this important understanding for continuing and future structures that are similar to those now being terminated under the new accounting standards In particular, it would be insightful to know how the valuations of the retained interest compare with the valuations of those tranches transferred to the outside entity Although the values of retained interest were known and audited, the values of the tranches transferred were not, nor were the underlying net assets within the QSPE It could be reasonably assumed that the retained interests and transferred tranches would be valued similarly, given their underlying commonality in the original pools of assets (such as credit card receivables), but we learn only upon reconsolidation that the transferor had to make significant adjustments These adjustments were necessary for the retained interest and the reconsolidated tranches, as we show above in the consolidation entry and in the table of unrealized losses Why this is the case is a cause of concern, and standard setters should pursue the reasons Are the accounting models so dissimilar that the pools of assets are measured in ways that not capture the same attribute of ‘‘value,’’ be it market value or net realizable value? Alternatively, incentives exist to measure the pools of assets in opportunistic ways that distort the economic reality of deteriorating values? By increasing transparency between the tethered pools of assets, it creates the self-checking mechanism on determining values, if the valuation models in fact capture similar economic realities Thus, we believe the need exists to reconcile the disparity between the loan loss provision under the FASB contingency model and the fair value model (and impairment model for held-to-maturity investments) used in accounting for retained interests This would better enable an understanding of the underlying causes of the disparities and the economics of the transaction In its 10-K, Discover presents financial information on a so-called ‘‘managed basis.’’ This basis presents Discover’s financial results as if it had been using ‘‘modified’’ financing Downloaded from jaf.sagepub.com at Taylor's University on December 22, 2012 Bryan and Lilien 93 accounting Discover indicates this is no longer necessary, because they have now adopted financing accounting for GAAP The previous applications of GAAP (sales accounting) and the managed basis (financing accounting) present a contradiction If the managed basis is financing, then the question arises as to why sales accounting was even used Pro Forma Disclosures Under requirements from the Securities and Exchange Commission (SEC), firms must reconcile their non-GAAP (pro forma) disclosures (such as Discover’s ‘‘managed basis’’ disclosures) to GAAP Therefore, Discover must show the adjustments from its managed basis to GAAP On adoption of the new accounting standards, Discover provides ‘‘additional adjustments to adjusted’’ numbers of its managed basis However, if the managed basis was meant to provide a better assessment of risk and rewards, then Discover’s ‘‘additional adjustments to the adjusted’’ number could imply that this was not the case Excerpts from Discover’s discussion of its managed basis are below First, we made securitization adjustments to reverse the effect of loan securitization by recharacterizing securitization income to report interest income, income expense, provision for loan losses, discount and interchange revenue and loan fee income in the same line items as non-securitized loans These adjustments result in a ‘‘managed basis’’ presentation, which we have historically included in our quarterly and annual reports to reflect the way in which our senior management evaluated our business performance and allocated resources in the past (Form 10-K, p 48) Then, additional adjustments were made to reflect results as if the trusts used in our securitization activities had been fully consolidated in our historical results  Elimination of interest income and interest expense related to certificated retained interests classified as investment securities and associated intercompany debt;  An adjustment to the provision for loan losses for the change in securitized loan receivables;  Elimination of the revaluation gains or losses associated with the I/O strip receivable, which was derecognized upon adoption; and  An adjustment to reflect the income tax effects related to these adjustments (Form 10-K, p 48) Discover’s ‘‘modified’’ adjusted results not completely portray the financing model, but rather an interim step, which in turn has to be adjusted As seen in Table 15, the ‘‘As Reported’’ and ‘‘Managed’’ Net Income numbers are US$120,394, but after the ‘‘Additional Adjustments’’ to bring the numbers to ‘‘full financing,’’ the Net Income number falls to a loss of US$196,328 Discover continues with the presentation in its first quarter 2010 presentation, where the firm provides ‘‘As Adjusted Adjustments,’’ shown in Table 16 Sample of Major Credit Card Securitizers We extend our analysis of Discover by including a sample of transferors with credit card trusts We identified these entities through an EDGAR search of financial institutions with Downloaded from jaf.sagepub.com at Taylor's University on December 22, 2012 94 Journal of Accounting, Auditing & Finance Table 15 Discover’s As Reported and Managed Net Income For the months ended February 28, 2009 (dollars in thousands) As reported Interest income 815,793 Interest expense 312,720 Net interest income 503,073 Provision for loan losses 937,813 Net interest income after provision for loan losses 2434,740 Antitrust litigation settlement 474,841 Other income 715,115 Total other income 1,189,956 Total other expense 559,123 Income (loss) before income tax expense 196,093 Income tax expense (benefit) 75,699 Net income (loss) 120,394 Securitization adjustments Managed Additional adjustments As adjusted 788,056 125,697 662,359 395,860 1,603,849 438,417 1,165,432 1,333,673 2992 211,535 10,543 143,411 1,602,857 426,882 1,175,975 1,477,084 266,499 — 2266,499 2266,499 — 2168,241 474,841 448,616 923,457 559,123 2132,868 2474,841 98,242 2376,599 — 2301,109 — 546,858 546,858 559,123 — — — 196,093 75,699 120,394 2509,467 2192,745 2316,722 2313,374 2117,046 2196,328 Table 16 Discover’s Results for First Quarter, 2010: Managed Basis GAAP total loans Securitization adjustments Managed basis As adjusted adjustments As adjusted total loans Less: Guaranteed portion of student loans As adjusted total loans less: Guaranteed portion of student loans reserve rate February 28, 2010 (US$) November 30, 2009 (US$) 50,093,516 NA 50,093,516 50,093,516 23,625,084 27,235,288 50,860,372 26,226 50,854,146 21,910,889 48,182,627 21,274,453 49,579,693 Note: GAAP = Generally Accepted Accounting Principles large credit card securitizations The number of such firms is 11 (Target was also identified but we deleted it from the sample because it did not use sales accounting, rather financing accounting.) For each firm, we calculate the impact on opening retained earnings as of the date of adoption All firms adopted as of first quarter of 2010, except for Discover Discover has a different year-end and adopted as of the last quarter of 2009 CompuCredit chose Method as the transition method from sales accounting and therefore used fair values in its transition General Electric chose Method and therefore had to undo the impact of sales accounting by reconstructing the accounts as if financing had been used all along General Electric illustrates the issue of undoing previously recorded gains and recording them again in the future Because the additional loan loss provision for General Electric is below the reduction in retained earnings on adoption of the new Downloaded from jaf.sagepub.com at Taylor's University on December 22, 2012 Bryan and Lilien 95 Table 17 Effect on Retained Earnings of Transition From Sales Accounting to Financing Accounting for Largest Credit Card Securitizers Bank of America Discover Citigroup Capital One JP Morgan Chase General Electric Alliance Data Systems Cabelas PNC CompuCredit American Express Retained earnings opening balance Absolute impact Relative impact (%) 71,233 3,873 77,440 10,727 62,481 126,363 1,033 697 13,144 287 14,400 26,154 21,411 28,442 22,922 24,391 21,945 2400 289 292 34 21,800 29 236 211 227 27 22 239 213 21 239 213 pronouncement, we attribute the differences to reversal of securitizations gains Those gains will be recognized in the future and will be double counted as they are recognized again through income as investing income We show the findings of our sample in Table 17 For all the large credit card securitizers including General Electric, the cumulative effect of the adjustment is attributable to the additional provision for loan losses As these additional provisions are charged against retained earnings, they never pass through income These impacts, which we scale by opening retained earnings, range from a reduction of 1% for PNC to 39% for Alliance Data Systems The incentives to be conservative in recording the loan loss reserves are clear When overaccrued reserves are released, future income is enhanced, as confirmed by Eckblad (2010) Eckblad reports that the biggest U.S banks ‘‘virtually doubled their collective earnings in the third quarter just by injecting US$8.1 billion into net income from funds they had set aside to cover loan losses’’ (p C1) J P Morgan Chase’s third-quarter earnings of US$4.4 billion included US$1.7 billion from the bank’s loan loss reserves that were reversed Fitch Ratings (2010) supports this finding showing that card segment profitability was positive for all issuers through the first months of 2010, excluding Bank of America’s goodwill impairment charge Fitch states that provision expenses drove bottom line improvement, as revenues were generally lower, given portfolio contraction Conclusion The demise of the QSPE brings to light the question of how firms were allowed to assign valuations (without extensive accompanying disclosures) of component parts of a pool of receivables whose values exceeded the value of the pool as a whole Many interested observers also question (admittedly with the advantage of hindsight) how accounting standards could allow firms to transfer receivables to entities created by the firms themselves without more transparency and oversight of these vehicles In such a conflict of interest, the passivity of the transferee and independence between the transferor and transferee are called into question It is also of concern that the transferees had limited or no reporting requirements That notwithstanding, transferors could arguably have been in position to Downloaded from jaf.sagepub.com at Taylor's University on December 22, 2012 96 Journal of Accounting, Auditing & Finance alert investors about the diminished values of the receivables, even though these receivables were no longer on their (the transferors’) books The current set of choices for transitioning from sales accounting to financing accounting means that no party (transferor, transferee, investor) accepts the responsibility for diminished values of the receivables, at least as far as income recognition is concerned The ‘‘hit’’ is against the past, as if it had all along been the responsibility and concern of someone else Moreover, the charge against the past could easily and conceivably be overaccrued, precisely because there is no downside, but (large) potential upside Current earnings are not affected, but future earnings could be boosted as potentially overaccrued provisions are reversed Therefore, this may be the first context in which the same asset pools’ effects on income are not double counted, but rather triple counted Another question that needs to be addressed is how to justify such large differences between the non-GAAP numbers used for managing a business and corresponding GAAP numbers, because GAAP has the economic portrayal of a firm as one of its major goals Investors and creditors must be puzzled at the enormous difference between what standard setters have agreed upon as ‘‘proper’’ measurement and recognition versus what firms themselves are presenting In such an atmosphere, it is difficult and confusing for investors to try to interpret the adjusted numbers and arguably nearly impossible to understand ‘‘adjusted adjusted numbers.’’ Investors will also wonder, in assessing firms’ governance and incentive structures, whether precautions are instituted so that employees are not paid twice (or times) for the same earnings streams Finally, academics (certainly ourselves included) should be cautious in ascribing ‘‘usefulness’’ or ‘‘value relevance’’ to fair values The difficulty in measuring ‘‘fair’’ values requires a certain degree of guarded skepticism, as it is quite likely that the technology and methodology for estimating fair value to produce a number that is useful is limited More important, skepticism is especially warranted because these numbers that are estimated as fair emanate from an environment fraught with agency problems These issues are particularly timely, as an exposure draft issued June 24, 2010, ‘‘Revenue Recognition (Topic 605): Revenue From Contracts With Customers,’’ addresses revenue recognition, which involves substantial bifurcation and use of fair values As the imbroglio in the financial services sector slowly dies down, we, along with many other observers, are concerned that other potential flare-ups will occur We encourage the accounting profession, academicians, standard setters, and regulators to consider the issues and questions we raise and make adjustments as deemed necessary Examples of questions that arise from our study include the following: Should ‘‘cumulative effects’’ be resurrected for issues such as these described in the article (rather than charging prior earnings effects and loan loss provisions directly against Retained Earnings)? Should any and all conduits or other off-balance-sheet entities that the firm (transferor) creates directly or indirectly be reconsolidated (due to a natural agency conflict)? Should non-GAAP results be restricted to only certain allowable adjustments? Should securitization gains be fully justified and described in terms of how a tranched pool of cash flows can be valued at an amount that is greater than the pool as a whole? Should prior academic studies documenting relevance of fair values be reexamined with appropriate controls for agency environments? Declaration of Conflicting Interests The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/ or publication of this article Downloaded from jaf.sagepub.com at Taylor's University on December 22, 2012 Bryan and Lilien 97 Funding The author(s) received no financial support for the research, authorship, and/or publication of this article Notes Although Generally Accepted Accounting Principles (GAAP) requires derecognition of accounts 10 11 12 13 14 receivable and treatment of the funds raised as a sale, many consider the risk and rewards to the transferor continue to be the same as if the transaction was structured as a financing arrangement As the standards that are relevant to our article largely predate the codification of accounting standards and are referenced in other literature in a similar way, we continue to use the Statement of Financial Accounting Standards (SFAS) notation The retained tranches are accounted for using ‘‘a mixed attribute model.’’ The accounting depends on whether the newly created ‘‘investment securities’’ are designated as trading, heldto-maturity or available-for-sale We note that because the transactions are accounted for under the new rules as secured borrowings, the cash flows are presented as cash flows from financing activities rather than cash flows from operating or investing activities We show only the results for the largest players in this sector, but we collected data on the entire sector The additional firms provide no meaningful insights due to the small amounts involved However, for data for the additional firms, please contact the authors Where the transferee was deemed to be a Qualifying Special Purpose Entity (QSPE) and control over the receivables relinquished, the transferor would not further test for consolidation of the QSPE and the transferor would use sales accounting Concern with problems in QSPEs was not unknown to standard setters In Emerging Issues Task Force (EITF) Issue No 02-12 (Financial Accounting Standards Board [FASB], 2000, 2002) standard setters addressed ‘‘permitted activities’’ in the QSPE, but no consensus was reached, and in 2003, the EITF stopped working on the issue The first choice is to be used unless impractical By reviewing the comment letters submitted during the comment period on the Exposure Draft, we know that large credit card securitizers lobbied for the third choice, and the ‘‘Big Four’’ accounting firms in their responses were silent as to implications of these choices We assume in the first two transition choices that neither the transferor nor transferee made any cash collections from the outstanding receivables Credit card loan provisions are scoped out of SFAS No 114 (FASB, 1993) and continued to be accounted for under SFAS No In our sample, the loan loss provisions are considerably larger than the fair value adjustments recognized for the retained interests accounted for as trading or available-for-sale securities It is not immediately clear which party bears the burden of uncollectible receivables That is, when the receivables are written off, it is unclear whether the transferor will bear the burden or if there is any recourse to the investors in the beneficial interest In the context of our article, we are describing issues that we feel the accounting profession, standard setters, and regulators need to address We chose Discover simply as a reference point for our discussion and analysis but not suggest in any way that they entered into transactions or made disclosures that were improper We note that Discover does continue to show on its November 30, 2009, balance sheet held-to-maturity investments of US$2.34 billion and parenthetically a corresponding fair value of US$1.95 billion If the third party ultimately does in fact bear the loan losses, the transferor benefits by not having to repay the third party and also benefits by reversing all of its loan loss provisions Downloaded from jaf.sagepub.com at Taylor's University on December 22, 2012 98 Journal of Accounting, Auditing & Finance Irrespective of the fact that we observe the transferor booking the provisions, it is not transparent as to which party will ultimately bear the burden of any nonperforming assets 15 Under FASB Staff Position (FSP) 115-2 (FASB, 2009c), a firm no longer determines whether it has ‘‘intent and ability,’’ but rather determines whether it has the intent to sell or is more likely than not required to sell before it recovers its costs basis If so, the loss goes through income If the holder does not intend to sell and it is not more likely than not that it will have to sell before it recovers its cost basis, the holder disaggregates credit losses from other losses and puts through income only the credit losses References Barth, M., & Taylor, D (2010) In defense of fair value: Weighing the evidence on earnings management and asset securitizations Journal of Accounting and Economics, 49(1-2), 26-33 Chen, W., Liu C., & Ryan, S (2008) Characteristics of securitizations that determine issuers’ retention of the risks of the securitized assets Accounting Review, 83, 1181-1215 Dechow, P., & Shakespeare, C (2009) Do managers time securitization transactions to obtain accounting benefits? Accounting Review, 84, 99-132 Eckblad, M (2010, October 27) Banks turn their reserves to profit The Wall Street Journal, p C1 Financial Accounting Standards Board (1975, March) Accounting for contingencies (Statement of Financial Accounting Standards No 5) Norwalk, CT: Author Financial Accounting Standards Board (1993, May) Accounting by Creditors for Impairment of a Loan—An amendment of FASB Statement No and 15 Norwalk, CT: Author Financial Accounting Standards Board (2000, September) Accounting for transfers and servicing of financial assets and extinguishments of liabilities—A replacement of FASB statement No 125 (Statement of Financial Accounting Standards No 140) Norwalk, CT: Author Financial Accounting Standards Board (2006, September) Fair value measurement (Statement of Financial Accounting Standards No 157) Norwalk, CT: Author Financial Accounting Standards Board (2009a, June) Accounting for transfers of financial assets—An amendment of FASB Statement No 140 (Statement of Financial Accounting Standards No 166) Norwalk, CT: Author Financial Accounting Standards Board (2009b, June) Amendments to FASB interpretation No 46(R) (Statement of Financial Accounting Standards No 167) Norwalk, CT: Author Financial Accounting Standards Board (2009c, April) Recognition and interpretation of other-thantemporary impairments (Financial Accounting Standards Board Staff Position [FSP] 115-2) Norwalk, CT: Author Fitch Ratings (2010, November 3) Commercial Mortgage Special Report U.S CMBS Loss Study: 2009, New York, NY Landsman, W., Peasnell, K V., & Shakespeare, C (2008) Are asset securitizations sales or loans? Accounting Review, 83, 1251-1271 Ryan, S (2008) Accounting in and for the subprime crisis Accounting Review, 83, 1605-1638 Saha-Bubna, A (2010, September 21) Discover feels economy’s rebound The Wall Street Journal, p C3 Downloaded from jaf.sagepub.com at Taylor's University on December 22, 2012 [...]... 0.00 1. 71 7.00 10 .00 15 0,000 11 .92 1. 00 0.00 0.00 0.00 0.00 1. 00 9.00 0.53 31. 06% 9.90% 12 .03 75th 32*** 15 *** 2.23*** Credit Reputable Reputable Rating Auditor Underwriter Tenure Leverage Tangibility 0.09 1. 30 6. 71 9. 01 187,4 21 9.83 0.45 0. 01 0.00 0. 01 1.64 9, 517 9, 517 9, 517 9, 517 9, 517 9, 517 9, 517 9, 517 9, 517 9, 517 9, 517 Bond Size 0.27 6 .12 0.43 22.75% 9.05% 10 .48 M 9, 517 9, 517 9, 517 9, 517 9, 517 9, 517 ... Tangibility ROA 10 *** 11 *** 32*** 11 *** 11 *** 06*** 01 2.07*** 2 .11 *** 03*** 2. 01 13*** 2 .16 *** 2.30*** 09*** 2.33*** 2 .13 *** 01 22*** 16 *** 34*** 22*** 03*** 14 *** 2.04*** 2.32*** 10 *** 07*** 33*** 20*** 2.06*** 13 *** 2 .18 *** 2.33*** 62*** 2 .15 *** 2 .12 *** 2.29*** 2.42*** 07*** 2 .18 *** 42*** 17 *** 2.36*** 2.42*** 37*** 62*** 27*** 16 *** 14 *** 15 *** 15 *** 2.06*** 19 *** 10 *** 2.05*** 09*** 02*** 00 2. 01 01 2.07***... Yes 9, 517 47 2.96* 20.58*** 22.70 0.24*** 2.87 20.80*** 23.64 0.28* 1. 69 0. 01 1.63 0.04* 1. 79 0 .14 *** 11 .66 1. 34*** 12 .40 0.52 0. 01 0 .18 21. 32 21. 82*** 23.35 0.82 0.25*** 2.57 20.64 0. 01 0.55 1. 84 0.04*** 3.26 20.63 0.03 0.60 20.86 1. 47 0.09 0.43 20. 01 0.50 0.03* 20 .10 1. 89 0.42 0. 01 0.023 0 .15 0.79 1. 52 0.25 1. 69 1. 40 0.02 0.96 3.62 0.96*** 2.75 21. 29 0. 01 0.20 20.76 0. 01* * 2.50 21. 57 20. 01 20.05... 20. 01 1.20*** 20 .16 *** 20. 01 20.03*** 21. 58** 0.03 0.58 20 .19 4.22 25.23 20 .19 25.08 22.25 0. 61 20 .18 * 0.06* 2.03*** 0.02 20. 01 20. 01 20.09 0.28** (continued) 21. 75 20 .17 20.24 0.80 1. 02 1. 49*** 4.00 1. 31* ** 5.09 1. 88 0.09 0.72 20.25 21. 23 0. 01 0.35 0.03 0.49 3.33 20.43 20 .16 20.22 20.07 24.25*** 22.72 21. 32 21. 28 0.29 0.47 1. 36 0.05 0 .18 0 .19 1. 24 0.20** 2 .15 20.97 0.04** 2.24 0.04** 2.39 0. 01* **... information Final sample 60,337 (790) (49 ,14 3) (12 8) (759) 9, 517 Panel B: Sample distribution by year Year Firm-level observations 19 95 19 96 19 97 19 98 19 99 2000 20 01 2002 2003 2004 2005 2006 Total 54 82 10 6 202 12 6 64 18 1 62 330 19 6 19 9 16 9 1, 7 71 Bond issue-level observations 239 3 61 482 554 617 439 759 15 3 2,239 1, 5 51 1,330 793 9, 517 Note: Panel A of Table 1 presents the observation selection process... 0.03 3. 41 0.52 0. 01 0.07* 2.75 20. 01 20.03 20 .11 20.66 1. 14 1. 74*** 5.26 1. 35*** 4.82 20.65 20.03 20.08 0 .11 0.26 2.54*** 0.92 20.84 21. 78 20.35*** 22.68 3 .18 *** 3.09 1. 05 20 .19 * 21. 73 0.75 0.78 0 .13 0 .15 1. 13 20.48 20.33 t t t t t t Coefficient statistic Coefficient statistic Coefficient statistic Coefficient statistic Coefficient statistic Coefficient statistic Model 1 Reputable Auditor 20. 31* Reputable... earnings news Journal of Finance, 63, 2287-2323 Lee, C M C., & Ready, M J (19 91) Inferring trade direction from intraday data Journal of Finance, 46, 733-746 Merton, R C (19 87) A simple model of capital market equilibrium with incomplete information Journal of Finance, 42, 483- 510 Odean, T (19 98) Are investors reluctant to realize their losses? Journal of Finance, 53, 17 75 -17 98 Odean, T (19 99) Do investors... December 22, 2 012 24 .11 22.27 2. 81 20.80 2.20 3.20 2.80 25.65*** 21. 84** 0. 01* ** 20. 01 0.09** 2.75*** 0.36*** Yes Yes 9, 517 24 Reputable Underwriter Tenure Leverage Firm size Tangibility ROA Putable Callable Group 2 Yes Yes 9, 517 11 21. 85*** 0.73** 0. 01* * 0. 01 20.04** 1. 44** 0.35*** Coefficient Group 1 23.08 2.22 2.03 1. 35 22.39 2.54 3.08 z statistic Reputable Underwriter Maturity Group 2 Group 1 Bond Size... December 22, 2 012 33 Downloaded from jaf.sagepub.com at Taylor's University on December 22, 2 012 Model 2 Credit Spread Model 3 Maturity Model 4 Model 5 Bond Size Model 6 Yes 9, 517 03 4.77 7.65 1. 56 20.63 0.56 24.78 20.35 15 .70*** 7.86*** 2. 71 21. 82 0.08 20.35*** 20.07 Yes 9, 517 29 3.29* Yes 9, 517 17 0. 01 Yes 9, 517 48 28.08*** 0.26*** 0.22 7.02 20.02 20.05 0. 21* * 2.33 1. 68*** 4. 01 20.74 21. 31 0. 01 0.54 Note:... No of Covenants Panel B: Correlations Table 2 (continued) 2.02** Sinking Callable Putable Fund Subordinated 2 .15 *** 2.05*** 2.04*** 2.04*** 06*** 2.00 2.04*** 2. 01* 2. 01 39*** 2.53*** 37*** 01 02** Firm Size 32 Journal of Accounting, Auditing & Finance Table 3 Univariate Tests Panel A: Reputable versus ordinary auditor Credit Spread Maturity Bond Size M1 M0 Difference in M t statistic 1. 18 11 .94 10 .38