• unlimited access with print and download
    $ 37 00
  • read full document, no print or download, expires after 72 hours
    $ 4 99
More info
Unlimited access including download and printing, plus availability for reading and annotating in your in your Udini library.
  • Access to this article in your Udini library for 72 hours from purchase.
  • The article will not be available for download or print.
  • Upgrade to the full version of this document at a reduced price.
  • Your trial access payment is credited when purchasing the full version.
Buy
Continue searching

Sell-in versus sell-through revenue recognition: An examination of firm characteristics and financial information quality

ProQuest Dissertations and Theses, 2009
Dissertation
Author: Stephanie Jean Binger Rasmussen
Abstract:
This study examines revenue recognition methods used by high technology firms for sales to distributors. Revenue is either recognized when products are delivered to distributors (sell-in) or when distributors resell products to end-users (sell-through). This is the first empirical study to examine the firms that use these revenue recognition methods and the quality of financial information reported under the methods. I use a logistic regression to compare 479 firm-year observations in the computer and electronic equipment industries that use either the sell-in method or the sell-through method. I find that firms with higher growth opportunities and strong corporate governance are less likely to use the sell-in method. In addition, corporate governance strength moderates the association between use of the sell-in method and both capital requirements and management incentive compensation. Using ordinary least squares regression, I also examine two proxies for financial information quality: the ability of accounting information to predict future cash flows and the association between accounting information and stock returns. Results of these regressions suggest that financial information quality is higher under a deferred revenue recognition method (sell-through). Specifically, the ability of accounting information to predict future cash flows and the association between accounting information and returns are both higher for sell-through firms than for sell-in firms. The results of this study suggest that systematic differences exist between sell-in firms and sell-through firms and financial information quality differs between the two revenue recognition methods.

TABLE OF CONTENTS

Page ABSTRACT .............................................................................................................. iii DEDICATION .......................................................................................................... v ACKNOWLEDGEMENTS ...................................................................................... vi TABLE OF CONTENTS .......................................................................................... vii CHAPTER I INTRODUCTION ................................................................................ 1

II BACKGROUND AND PRIOR RESEARCH ..................................... 8 Revenue Recognition Practices of High Technology Industries .... 8 Financial Information Quality ........................................................ 10 Prior Research on Revenue Recognition ........................................ 11 III HYPOTHESES DEVELOPMENT ...................................................... 13

Costs and Benefits of the Revenue Recognition Methods ............. 13 Empirical Predictions ..................................................................... 15 Financial Information Quality ........................................................ 20

IV FIRM CHARACTERISTICS AND REVENUE RECOGNITION ..... 22

Research Design ............................................................................. 22 Sample Selection ............................................................................ 26 Descriptive Statistics and Correlations .......................................... 27 Multivariate Analyses .................................................................... 29 Supplemental Analysis ................................................................... 33

V FINANCIAL INFORMATION QUALITY AND REVENUE RECOGNITION .................................................................................... 35

Ability of Accounting Information to Predict Future Cash Flows . 35 Association between Accounting Information and Stock Returns . 39 Supplemental Analyses .................................................................. 43

viii

CHAPTER Page

VI CONCLUSION .................................................................................... 45

REFERENCES .......................................................................................................... 47 APPENDIX A ........................................................................................................... 55 APPENDIX B ........................................................................................................... 58 VITA ......................................................................................................................... 70

1

CHAPTER I INTRODUCTION This study examines revenue recognition methods used by high technology firms for sales to distributors. Distributors purchase products from manufacturers and resell the products to end-users. This activity is common within high technology industries, and current accounting standards allow firms some discretion on when to recognize revenue from sales to distributors. Two revenue recognition methods exist: the sell-in method and the sell-through method. I address two research questions in this study. First, what firm characteristics are associated with technology firms‘ choice of revenue recognition method for sales to distributors? Second, does the quality of financial information differ between the two revenue recognition methods? Under the sell-in method, firms recognize revenue when the product is delivered to the distributor (i.e. product is sold into the distribution channel). Under the sell- through method, firms defer revenue recognition until the distributor resells the product to an end customer (i.e. product is sold through the distribution channel). Sales to distributors usually meet the Staff Accounting Bulletin (SAB) No. 104 revenue recognition requirements that persuasive evidence of an arrangement exists and delivery has occurred. 1 The decision to use the sell-in or sell-through method generally depends upon the remaining two SAB 104 requirements: the final selling price is fixed or determinable and collectability is reasonably assured. Sales in high technology

This dissertation follows the style of The Accounting Review. 1 The revenue recognition principles contained in SAB 104 are relatively unchanged from SAB 101 (SEC 1999, 2003). The main purpose of SAB 104 was to rescind accounting guidance within SAB 101 that was superseded by the FASB‘s Emerging Issues Task Force (EITF) 00-21.

2

industries are often subject to rights of return and to pricing adjustments due to price reductions in the marketplace. A conservative interpretation of SAB 104 suggests that the final selling price for any distributor sales subject to pricing adjustments or rights of return are indeterminable. However, interpretive guidance within SAB 104 suggests that a selling price is determinable if product returns and pricing adjustments can be reasonably estimated. 2 The probability of collection depends on the fixed or determinable nature of the final selling price and if collections depend on the distributor reselling the products. Based on these factors, the sell-in method is typically considered the more aggressive method (Glass, Lewis & Co. 2004; Greenberg 2006). However, the discretion provided under SAB 104 allows high technology firms enough flexibility to justify using either revenue recognition method. 3

Understanding the characteristics of firms that use different revenue recognition methods and financial information quality under those methods is important for many reasons. First, revenue is arguably the most important component of earnings. Revenue is usually the largest item on the income statement and it is often viewed as a strong indicator of firm performance (Turner 2001). A former chairman of the SEC argued that early or premature revenue recognition is a fundamental problem in accounting (Levitt 1998), and misreported revenue is a leading cause of financial restatements (GAO

2 Guidance about fixed and determinable sales prices refers to Statement 48, para. 6 and 8, which state that revenue cannot be recognized if a firm is unable to make a reasonable estimate of product returns (FASB 1981). SAB 104 also directs users to SOP 97-2, para. 26 and 30-33, which states that prices on products sold to distributors are not fixed and determinable if the seller is unable to make reasonable estimates of pricing adjustments (AICPA 1997). 3 The first two disclosure examples in Appendix A suggest that sales arrangements with distributors are essentially identical for the firms identified, yet one firm uses the sell-in method and the other uses the sell-through method.

3

2007). 4 Overstated revenue has also been documented as the cause of at least 50 percent of Accounting and Auditing Enforcement Releases (AAER) (Feroz et al. 1991; Dechow et al. 1996; Dechow et al. 2007). Second, evidence on the effects of sell-in and sell-through methods on firm reporting quality is potentially useful to standard setters. The FASB has undertaken a joint project with the IASB to create a comprehensive revenue recognition standard (FASAC 2006). While it is unclear if the comprehensive revenue recognition standard will allow both the sell-in and sell-through methods, Lynn Turner, former SEC Chief Accountant, has expressed concerns about the sell-in method (Greenberg 2006): I have had to deal with the issue of whether you recognize revenue upon sell-in versus sell-through as an audit partner, a CFO and as a regulator, and now as an advisor to institutions. In all of these, I found nothing good about revenue recognition upon sell-in. Sooner or later, the urge to stuff the channel, especially when things are not going well and numbers for the next quarter are short, is very tempting.

The FASB‘s current Statement of Concepts (FASB 1978, para. 37) and a recent exposure draft of the proposed future conceptual framework (FASB 2008) both argue that financial reporting should provide information that capital providers and other parties can use to assess an entity‘s future net cash flows. Although current period cash flows may be the same under both the sell-in and sell-through methods if distributors settle accounts receivable prior to product resale, the ability of accounting information to predict future cash flows may differ between the two methods. It is also unclear whether

4 Cost/expense errors and revenue errors were the leading causes of restatements from January 1997- September 2005. In this period, 27.4 (27.2) percent of restatements were due to cost/expense (revenue) errors (GAO 2007).

4

sell-in or sell-through accounting information is most consistent with firm values. 5 To my knowledge, this is the first study to empirically examine financial information quality under the sell-in and sell-through methods. Finally, the evidence presented in this study may be useful for firms in industries that use the sell-in and sell-through methods for revenue recognition. Insights about the characteristics of firms that use different revenue recognition methods and the quality of financial information under the two methods may be helpful to firms that are examining their own revenue recognition practices and/or considering an accounting method change. In order to address my research questions, I study 479 unique firm-year observations in the computers and electronic equipment industries during 2001-2005. I classify firms in these industries as sell-in or sell-through based on their 10-K revenue recognition disclosures. I first investigate the characteristics of firms that use the sell-in and sell-through methods. Specifically, I test for associations between use of the sell-in method and proxies for capital requirements, management incentive compensation, growth opportunities, and corporate governance strength. I find that use of the sell-in method is negatively associated with growth opportunities and corporate governance strength. I also find that as capital requirements and management incentive compensation increases,

5 Prior research does examine the association between stock returns and accounting information under other revenue recognition practices (e.g. Altamuro et al. 2005; Zhang 2005; Srivastava 2008). These studies all examine firms that were required by standard changes to use less aggressive revenue recognition practices. The evidence presented in these studies suggests that the association between accounting information and stock returns was stronger when firms were allowed to accelerate revenue recognition than when they were required to delay it.

5

firms with strong corporate governance are less likely to use the sell-in method than firms with weak governance. Thus, corporate governance strength moderates the association between use of the sell-in method and other factors. Next, I examine whether financial information quality differs between the sell-in method and the sell-through method. It is not clear that one method is consistently more reliable than the other method. The sell-in method may suffer from errors in estimating product return and/or pricing adjustment accruals. Managers may also use the discretion allowed under this method to enhance performance through channel stuffing or accrual manipulation 6 . Meanwhile, reliability concerns exist for sell-through accounting if distributor inventory and resale data contain errors or are not updated on a timely basis. It is also unclear if financial statement users perceive differences in relevance between the two revenue recognition methods. The sell-in method provides timely information about expected future demand but not current end-user demand. On the other hand, the sell-through method more accurately reflects end-user demand and is a signal of conservative accounting practices. I use two proxies to test for financial information quality differences between the sell-in and sell-through methods: (1) the ability of accrual accounting information to predict future cash flows, and (2) the association between accounting information and contemporaneous stock returns. I find that sell-through firms‘ accrual accounting information is more highly associated with future cash flows than sell-in firms‘

6 Channel stuffing occurs when (1) manufacturers pull in and ship distributor orders originally scheduled to be delivered in the next accounting period or (2) distributor inventory levels significantly exceed the amount historically needed to service end customers.

6

information. These results are consistent with Forester (2008) who finds that the ability of accrual accounting information to predict future cash flows is better for firms using deferred revenue recognition practices. I also find that accounting information reported under the sell-through method is more strongly associated with contemporaneous stock returns than accounting information reported under the sell-in method. This suggests that the sell-through method provides more timely accounting information than the sell- in method. Taken together, the results of these two tests suggest that the sell-through method produces higher quality financial information than the sell-in method. This study contributes to the literature in a number of ways. First, this study examines revenue recognition practices that have not previously been examined: the sell-in method, which offers companies the opportunity to accelerate revenue recognition, and the sell-through method, under which revenues are likely to be recognized relatively conservatively. Second, this study provides additional evidence that use of conservative revenue recognition methods is associated with higher growth opportunities and strong corporate governance (Skinner 1993; Altamuro et al. 2005). In addition, corporate governance strength moderates the association between revenue recognition practices and certain firm characteristics. Finally, this study suggests that financial information quality is higher under a deferred revenue recognition practice. This result is important because prior research offers mixed evidence regarding financial information quality under aggressive and delayed revenue recognition methods (Altamuro et al. 2005; Zhang 2005; Forester 2008). The evidence presented in this study should be of interest to investors, practitioners, auditors, and regulators.

7

The remainder of this paper is organized as follows. Chapter II discusses background and prior research while Chapter III develops my hypotheses. Chapter IV examines the association between firm characteristics and revenue recognition practices. Chapter V examines financial information quality under the sell-in and sell-through methods. Chapter VI concludes.

8

CHAPTER II BACKGROUND AND PRIOR RESEARCH Revenue Recognition Practices of High Technology Industries The sell-in and sell-through methods differ with respect to the timing of revenue recognition for sales to distributors. Under the sell-in method, revenue and cost of goods sold are recognized upon product delivery to the distributor. Under the sell-through method, revenue is deferred until notification is received that the distributor has resold the product. Accounts receivable are typically recorded when the distributor receives the products, and the distributor often pays for the products before they are resold. Distributors typically have some limited right of return, but 10-K filings suggests that most distributors do not have unlimited return privileges on regular purchases. 7 I expect firms‘ revenue recognition method to be relatively sticky over time since a cumulative effects adjustment would be needed if firms change accounting methods. If distributor purchases equal distributor resales, sell-in revenue differs from sell- through revenue by the amount of return and pricing adjustment allowances required under the sell-in method. However, if distributor purchases exceed resales, then revenue is higher under the sell-in method than the sell-through method. For example, Apogee Technology initially reported fiscal 2003 sell-in product revenue of $9.3 million and later retroactively reported 2003 sell-through product revenue of $7.8 million when it

7 Distribution agreements between manufacturers and distributors typically do include clauses that allow the distributors to return any product on hand if the relationship between the two parties is terminated (e.g. Arrow Electronics 2004 10-K filing; Avnet 2005 10-K filing; Ingram Micro 2005 10-K filing). However, most manufacturers and distributors enter into agreements with the intent of maintaining a long-term relationship.

9

changed accounting methods (Apogee Technology 2004 10-K/A filing). 8 The difference is because distributor purchases were significantly higher than resales. This example indicates that revenue recognition practices can significantly affect reported operating performance. Current U.S. GAAP provides some revenue recognition guidance for product sales to distributors. As discussed earlier, revenue recognition under SAB 104 requires (1) persuasive evidence that an arrangement exists, (2) delivery has occurred, (3) final selling price is fixed or determinable, and (4) collectability is reasonably assured. SAB 104 also notes that product returns estimations, and thus determination of a final selling price, for sales to distributors may be difficult due to the following factors: channel stuffing, difficulty in observing distributor inventory and resale data, and the significance of a distributor to the seller‘s business.

9 Estimates of future pricing adjustments may also be difficult, and SAB 104 refers financial statement preparers to SOP 97-2 for guidance on this issue. 10

The distributor is considered the customer for revenue recognition purposes but is not the end-user of the products. Distributors attempt to stock products they can resell, and they purchase inventory based on existing customer orders and expectations

8 Apogee adopted the sell-through method after an investigation by its audit committee found that the firm‘s use of the sell-in revenue recognition method did not comply with U.S. GAAP. 9 It is important to note that it would also be difficult for a firm to use the sell-through revenue recognition method if distributor resale and inventory data are not easily obtainable. 10 High technology firms offer pricing adjustments to their distributors in order to compensate for price reductions in the marketplace or to incentivize sales of certain products (Lee et al. 2000; CSFB 2004). Since the exact amount of pricing adjustments is often not known until the distributor resells the product, firms use historical resale information to estimate pricing adjustment.

10

about future orders.

11 If actual distributor resales significantly vary from forecasted orders and speculative beliefs, sell-in revenue will not provide timely information about future end-user demand.

Accounting for revenue from sales to distributors is important for high technology firms because these firms depend heavily on distributors to service customers. An equity research report suggest that distributors service more than 25 percent of global semiconductor/electronic component sales (CSFB 2004), and manufacturer 10-K filings suggest that 50 percent or more of their sales can go through distributors (e.g. Fairchild Semiconductor 2004 10-K filing; Cypress Semiconductor 2005 10-K filing). Distributors provide (1) access to an additional ―sales force,‖ (2) aggregation and service of small orders, and (3) reduced collection risk (CSFB 2004). Financial Information Quality The quality of financial accounting information should be of interest to all parties that create or use financial statements. Earnings quality is often used as an indicator of overall financial information quality in prior studies (Schipper and Vincent 2003; Francis et al. 2006). The FASB‘s Conceptual Framework implies that decision usefulness is the appropriate benchmark to assess the effectiveness and quality of accounting information (Concepts Statement No. 2, FASB (1980), paras. 30 and 32). However, financial statement users often define decision usefulness differently.

11 One equity research report suggests that the composition of electronic component distributors‘ inventory is as follows: 25 percent to support existing customer orders, 50 percent to support expected future customer orders, and 25 percent as speculative inventory to support unanticipated demand (CSFB 2004).

11

Prior studies use accounting-based and market-based measures of earnings quality (see Francis et al. 2006 for a thorough discussion). 12 Accounting-based measures assume that higher quality earnings allow for better estimation of future cash flows, earnings, or earnings components. In general, more persistent, more predictive, and less variable earnings are assumed to be of higher quality (e.g. FASB 1980; Penman and Zhang 2002; Zhang 2005). Earnings are also assumed to be higher quality when accrual estimation errors are smaller (e.g. Dechow and Dichev 2002). Market-based measures assume that higher quality accounting information better represents firm value and reflects the good and bad news in stock returns in a timely manner. Studies using returns-earnings regressions view stronger associations as evidence of more relevant and reliable accounting information, and thus higher earnings quality (e.g. Barth, Beaver, and Landsman 2001). Other studies measure the timeliness and conservatism in earnings using earnings-returns regressions and view more timely and more conservative earnings as being of higher quality (e.g. Basu 1997; Ball et al. 2000). Prior Research on Revenue Recognition There is limited research on firms‘ revenue recognition methods and their impact on financial information quality. Prior research suggests that firms with external financing needs, financial covenants, or weaker corporate governance are more likely to accelerate revenue recognition (Bowen et al. 2002; Marquardt and Weidman 2004;

12 Additional ex-post indicators of earnings quality include financial restatements, bankruptcies, litigation, and discontinuities around earnings targets (Degeorge et al. 1999; Anderson and Yohn 2002; Ecker et al. 2006).

12

Altamuro et al. 2005). 13 Other studies examine firms that adopted accounting standards intended to delay revenue recognition (Altamuro et al. 2005; Zhang 2005; Forester 2008; Srivastava 2008). In general, these studies examine short time periods after standard changes and find that accelerated revenue recognition results in more timely and relevant accounting information. 14 However, Forester (2008) examines a longer period and finds that the initial decline in earnings informativeness is due to a temporary disturbance caused by deferred revenues resulting from SAB 101 adoption. After these deferred revenues are recognized, Forester (2008) finds that the deferred revenue recognition method improves earnings informativeness in later periods.

13 Prior research also finds that more than 50 percent of AAERs are due to overstated revenue (Feroz et al. 1991; Dechow et al. 1996; Dechow et al. 2007). Revenue recognition errors are one of the leading causes of restatements from 1997-2006 (GAO 2007) and are associated with more negative stock price reactions and a higher likelihood of litigation than other restatements (Anderson and Yohn 2002; Wu 2003; Palmrose and Scholz 2004). Other research finds that firms manipulate revenue to meet or beat earnings benchmarks (Caylor 2008; Stubben 2006). 14 Zhang (2005) also finds that accelerated revenue recognition results in less reliable revenue for her sample (i.e. larger accounts receivable errors and less reduced revenue predictability).

13

CHAPTER III HYPOTHESES DEVELOPMENT The first objective of this study is to investigate characteristics of firms that use the sell-in and sell-through methods. Based on the fact that both methods are used in practice and the assumption that SAB 104 generally offers enough discretion for high technology firms to justify using either method, I expect that firms use the method they perceive offers the greatest net economic benefit. I use prior accounting research and anecdotal evidence to identify potential costs and benefits of the sell- and the sell- through methods. I then offer hypotheses about firm characteristics that I expect to be associated with use of the revenue recognition methods. Costs and Benefits of the Revenue Recognition Methods

Several commentators characterize the sell-in method as more aggressive than the sell-through method (Glass, Lewis & Co. 2004; Greenberg 2006). However, the sell- in method does offer benefits. Because the sell-in method recognizes revenue upon delivery to distributors, it provides a more timely reflection of actual business transactions. Texas Instruments gives this reason for using the sell-in method (Greenberg 2006). Sell-in revenue recognition also provides information about expected product demand (i.e. future distributor resales). Potential costs of using the sell-in method arise from the estimations and discretion allowed under this method. The requirement to maintain product return and

14

pricing adjustment accruals creates the possibility of unintentional estimation errors.

15

In addition, managers can manipulate performance by channel stuffing and/or adjusting product return and pricing adjustment allowances. Both of these factors increase the risk of accounting misstatements (Glass, Lewis & Co. 2004). For example, Symbol Technologies restated two years of prior financial statements due to a variety of revenue recognition issues, including selling prices that were later deemed undeterminable because of pricing adjustments subsequently awarded to distributors (Symbol Technologies 2002 10-K filing). This company switched to the sell-through method as part of its financial restatement process. The benefits of using the sell-through method relate to its conservative nature. This method more accurately reflects end-user demand and offers no incentive to stuff the distribution channel. 16 Accordingly, firms may use the sell-through method to signal the quality of their financial information. Levine and Hughes (2005) model a setting where conservative accounting choices provide positive signals about future cash flows. Conservative accounting should also reduce the risk of shareholder litigation, which is typically associated with overstated rather than understated net assets and earnings (Kellog 1984; St. Pierre and Anderson 1984; Watts 2003). Obtaining the benefits of the sell-through method does come at a cost. First, firms using the sell-through method depend on the reliability of distributor resale and

15 Some firms cite estimation difficulties due to frequent price changes and technological obsolescence as reasons for using the sell-through method (e.g. Intel 2007 10-K filing; Micron 2007 10-K filing; Supertex 2007 10-K filing). 16 Increased focus on end-user demand and reduction of distributor inventory were both mentioned by ON Semiconductor and International Rectifier when they announced plans to use the sell-through method (ON Semiconductor press release 4/25/01; International Rectifier conference call 8/4/08).

15

inventory data. 17 Chipalkatti et al. (2007) note that it may be difficult to obtain distributor data, remove data errors, validate the data, and convert data received from multiple distributors into one consistent format. 18 In order to deal with these issues, the sell-through revenue recognition process requires additional internal controls beyond those controls used for revenue recognition of non-distributor customer sales. Second, practitioners have stated that it is difficult to find sell-through accounting resources. Sell-through firms admit to: (1) advising other firms on sell-through accounting practices, and (2) using spreadsheets and home grown systems to facilitate sell-through accounting due to limited off the shelf software products that meet their needs. Empirical Predictions The costs and benefits discussion in the previous section leads to a number of predictions about associations between firm characteristics and use of the sell-in and sell-through methods. The first characteristic I expect to be associated with the revenue recognition method used is a firm‘s capital requirements. Firms with existing capital or need of new capital may use the sell-through method in order to signal a commitment to more conservative accounting. Prior research finds that lenders benefit from conservative accounting by receiving more timely signals of default risk, and that lenders reward borrowers using conservative accounting with lower interest rates (e.g. Ahmed et al. 2002; Zhang 2008). In addition, firms undergoing an initial public offering

17 Distributor data issues also affect sell-in firms‘ product return and pricing adjustment estimates. However, all revenue recognition for sell-through firms depends on distributor data. I view problems with distributor data to be a greater risk for sell-through firms than for sell-in firms. 18 Texas Instruments cites its lack of confidence in Asian distributor data as one reason it uses the sell-in method (Greenberg 2006), and a recent KPMG (2006) study indicates that 20 percent of resale reports from channel partners may contain missing data or errors.

16

have been found to report more conservatively than private firms suggesting that conservative accounting is valued by equity providers (Ball and Shivakumar 2008). Firms with ample capital may use the sell-through method because they attempted to obtain the benefits just described when they issued debt or equity in prior periods. However, firms with existing capital or need of new capital may instead use the sell-in method because it offers greater opportunity to manipulate financial performance through accrual manipulation and/or channel stuffing activities. Sweeney (1994) finds that firms with existing debt implement income increasing accounting changes in order to avoid debt covenant violations. Other research suggests that firms needing new capital manage earnings by accelerating revenue recognition prior to issuing debt or equity (Bowen et al. 2002; Marquardt and Weidman 2004). In both cases, firms attempt to mislead capital providers by reporting better financial performance than they otherwise would. Due to the competing evidence with respect to the association between capital requirements and accounting methods, my first hypothesis is non- directional: H1: Use of the sell-in method is systematically related to firms’ capital requirements. The second characteristic I expect to be associated with firms‘ revenue recognition method is the level of incentive compensation available to management. Prior research on management compensation and accounting practices finds mixed results. One stream of this research suggests that executives manage earnings in order to increase current period compensation. Cheng and Warfield (2005) find that as equity

Full document contains 79 pages
Abstract: This study examines revenue recognition methods used by high technology firms for sales to distributors. Revenue is either recognized when products are delivered to distributors (sell-in) or when distributors resell products to end-users (sell-through). This is the first empirical study to examine the firms that use these revenue recognition methods and the quality of financial information reported under the methods. I use a logistic regression to compare 479 firm-year observations in the computer and electronic equipment industries that use either the sell-in method or the sell-through method. I find that firms with higher growth opportunities and strong corporate governance are less likely to use the sell-in method. In addition, corporate governance strength moderates the association between use of the sell-in method and both capital requirements and management incentive compensation. Using ordinary least squares regression, I also examine two proxies for financial information quality: the ability of accounting information to predict future cash flows and the association between accounting information and stock returns. Results of these regressions suggest that financial information quality is higher under a deferred revenue recognition method (sell-through). Specifically, the ability of accounting information to predict future cash flows and the association between accounting information and returns are both higher for sell-through firms than for sell-in firms. The results of this study suggest that systematic differences exist between sell-in firms and sell-through firms and financial information quality differs between the two revenue recognition methods.