Transition to the New Revenue Standard: A Study of Firms

University of Missouri, St. Louis University of Missouri, St. Louis

IRL @ UMSL IRL @ UMSL

Dissertations UMSL Graduate Works

4-7-2023

Transition to the New Revenue Standard: A Study of Firms’ Transition to the New Revenue Standard: A Study of Firms’

Decisions to Adopt ASC 606 Early Decisions to Adopt ASC 606 Early

Irene Y. Zhang

University of Missouri-St. Louis

, [email protected]

Follow this and additional works at: https://irl.umsl.edu/dissertation

Part of the Accounting Commons

Recommended Citation Recommended Citation

Zhang, Irene Y., "Transition to the New Revenue Standard: A Study of Firms’ Decisions to Adopt ASC 606

Early" (2023).

Dissertations

. 1314.

https://irl.umsl.edu/dissertation/1314

This Dissertation is brought to you for free and open access by the UMSL Graduate Works at IRL @ UMSL. It has

been accepted for inclusion in Dissertations by an authorized administrator of IRL @ UMSL. For more information,

please contact [email protected].

Transition to the New Revenue Standard:

A Study of Firms’ Decisions to Adopt ASC 606 Early

Irene Y. Zhang

Master of Accountancy, Kansas State University, 2007

A Dissertation Submitted to The Graduate School at the University of Missouri

– St. Louis in partial fulfillment of the requirements for the degree of

Doctor of Business Administration with an emphasis in Accounting

May 2023

Advisory Committee

Jennifer W. Chen, Ph.D.

Chairperson

Tom Kozloski, Ph.D.

Michele Meckfessel, Ph.D.

Abstract

The Financial Accounting Standards Board has issued a new revenue standard

through a series of updates since May 2014 (collectively, ASC 606). The standard

allowed firms to adopt it a year early before the required effective date. This study

provides evidence of the adoption effects of ASC 606 on firms’ financials and examines

the relationship between the likelihood of early adopting ASC 606 and firm-specific

characteristics. The results show that early adopters are distributed across 12 industries

and 40% of them are concentrated in the computer programming and software industry.

Early adopters in the software industry reported the largest cumulative adoption effects in

dollars on retained earnings. Overall, the majority of the sample firms (83%) reported

positive adoption effects on financial statements and the adoption of ASC 606 increased

their retained earnings by 2.95% on average. The results also show that most early

adopters chose the full retrospective method while the majority of industry-size-matched

non-early adopting peers chose the modified retrospective method. Finally, the

multivariate regression analyses suggest that early adopters are associated with decreased

earnings in the year before adoption and more favorable adoption effects than non-early

adopting peers.

KEYWORDS: Revenue Recognition, ASC 606, Early Adoption, Accounting Choice,

Positive Accounting Theory

ACKNOWLEDGMENTS

I would like to express my sincerest gratitude to every member of my dissertation

committee, Dr. Jennifer W. Chen, Dr. Tom Kozloski, and Dr. Michele Meckfessel for

providing great insight and feedback throughout the research project. They are wonderful

accounting researchers and educators. I appreciate the opportunity to learn from each of

them.

I am particularly grateful to my dissertation chair, Dr. Chen. Her consistent

encouragement, support, and overall insights have made this journey an inspiring

experience for me. I want to thank all the DBA professors and staff for their leadership

and support in my doctoral studies. I have developed research interests and skills that I

will continue to explore in the future. I also want to thank my fellow DBA classmates. I

truly appreciate your friendship and enjoy the whole learning process with you in the

coursework, class discussions, and team projects.

Last but not least, I want to thank my husband, Long Qiao, and my children, Ray

and Rainer, for your love, patience, support, and sacrifices made over the last few years

so that I can pursue my DBA dream. Without you, this research would not have been

possible. I dedicate this dissertation to my family and all my loved ones.

4 
 
 
   
Table of Contents 
Chapter 1: Introduction .................................................................................................................... 5 
Chapter 2: Literature Review ......................................................................................................... 15 
  2.1 History of US Accounting Authorities and Pronouncements ……………………….15 
  2.2 Background on Revenue Standard-Setting ………………………….…....................16 
  2.3 ASC 606 New Revenue Standard…………………………………………………..  21 
  2.4 Transition Requirements and Choices ………………………………………………27 
  2.5 Existing Research on ASC 606………………………………………………………30 
  2.6 Research Interest……………………………………………………………………..33 
  2.7 Hypotheses Development…..………………………………………………………..39 
Chapter 3: Research Method .......................................................................................................... 44 
  3.1 Sample Selection….………………………………………………………………….44 
  3.2 Empirical Model and Measurement….………………………………………………50 
Chapter 4: Results .......................................................................................................................... 57 
  4.1 Descriptive Statistics…………...…………………………………………………….57 
  4.2 Test of Hypotheses..………………………………………………………………….64 
Chapter 5: Discussion and Conclusion .......................................................................................... 75 
          5.1 Implications for Research ........................................................................................... 76 
          5.2 Implications for Practice ............................................................................................. 80 
          5.3 Limitations and Future Studies ................................................................................... 82 
          5.4 General Conclusions ................................................................................................... 83 
References ...................................................................................................................................... 84 
Appendix A. ................................................................................................................................... 94 
 

New Revenue Standard ASC606

Chapter 1: Introduction

Revenue is a key measure of a firm’s financial performance and a starting point to

analyze a company’s value, trends, and prospects. It affects key financial ratios, related metrics

(e.g. EBITDA), and contracts that are based on these measures such as incentive compensation

plans and debt covenants. As an indicator of a firm’s potential value and management’s

stewardship effectiveness, this top-line number on the income statement is monitored closely by

internal and external users of financial statements such as managers, investors, creditors, and

market regulators (Papa, 2016; PWC, 2019).

Despite its importance, revenue recognition has always been complex and challenging.

Standard-setters in the past followed a rules-based approach to accommodate the complexity by

issuing industry- and transaction-specific guidance. These rules developed by various standard-

setting bodies were excessive and hard to implement, which resulted in “different accounting for

economically similar transactions” (FASB, ASU2014-09: p.1). Given its critical importance and

difficulties in applications, it is not surprising that revenue is the most scrutinized number by

auditors, investors, and regulators. Studies of the SEC enforcement actions indicate that

improper revenue recognition was a major reason for financial restatements and a frequent

subject of accounting fraud (Carmichael, 2019; Jones & Pagach, 2013; Sherman & Young,

2016).

Over the years, regulators and stakeholders have raised concerns over revenue guidance's

weakness, inconsistency, and non-comparability and requested improvement (SEC, 2003). To

address the concerns, the Financial Accounting Standards Board (FASB) has worked

collaboratively with the International Accounting Standards Board (IASB) since 2002, and

New Revenue Standard ASC606

issued a new universal revenue standard – Accounting Standards Codification (ASC) 606,

Revenue from Contracts with Customers, on May 28, 2014.

ASC 606 replaced the rules-based industry-specific guidance prescribed in ASC 605

Revenue Recognition with a new principles-based five-step model to streamline revenue

reporting. This five-step model is highly conceptual. It starts with identifying a bona fide

contract and distinct performance obligations (what), then determines the amount of revenue to

recognize for each obligation (how much), and ends with determining the timing of revenue

recognition based on the pattern of transfer of control of goods/services to customers (when).

The new guidance affects almost all entities (public and non-public) that enter into contracts with

customers to transfer goods or services unless they are subject to other standards (e.g., insurance

contracts, lease contracts, and financial instruments). It represents a significant change in

accounting principles and requires increased discretion and judgment under the new principles-

based framework (Ahn et al., 2021; Peters, 2018; Reddy, 2018; Thorps et al., 2020; Tzuo, 2017).

The new standard allows firms to make two accounting choices in the initial adoption:

adoption timing and transition method. In terms of adoption timing, firms have the choice to

voluntarily adopt it as early as one year before the required effective date or defer compliance to

the mandatory effective date, December 15, 2017. The standard also allows firms to choose

either the full retrospective method (FRM) or the modified retrospective method (MRM) to

report the GAAP changes due to the application of the new standard. Under the FRM approach,

ASC 606 is applied retrospectively to each reporting period on the comparative financial

statements, and the cumulative true-ups are made at the earliest reporting period. The FRM

method enables a line-by-line comparison of financial results before and after the adoption, but

this method may require a substantial overhaul of accounting systems to perform dual reporting

New Revenue Standard ASC606

using two sets of accounting data, which may not be easily available (Tysiac, 2014). In

comparison, the MRM provides a practical expedient approach where the cumulative

adjustments are applied at the date of adoption and the comparative years are not restated. This

approach could limit the information needed to understand the adoption effect of ASC 606

(Papa, 2016; Peters, 2018). In sum, the adoption timing and transition methods chosen may

affect the time and the level of effort to implement the changes. The adoption decisions may also

affect the outcome of the financial statements presented. Companies need to evaluate the impact

of adoption choices, seek information about how other companies adopt the standard, and

determine what meaningful information to disclose before making the adoption decisions.

The implementation of the new standard presents significant challenges for many

companies. This is especially true for those that have multiple deliverables such as software and

high-tech firms, for aerospace and defense firms that have multi-year contracts with milestone

payments, or for firms with innovative business models that involve varying levels of credit risks

and fulfillment responsibilities (Deloitte, 2017; Nehoray et al., 2018). For example, Intellectual

Property (IP) - related business models are becoming more pervasive in the technology,

pharmaceutical, and entertainment industries, but accounting for IP license revenue is difficult

due to the complexity involved in determining the nature of licenses

. To address stakeholders’

concerns about limited implementation guidance in the application of the highly conceptual

revenue model, the FASB added additional concepts (functional IP or symbolic IP) in its

clarifying guidance to help firms determine whether revenue should be recognized over time or

at a point in time (Papa, 2016). As another example, Oracle, a software company that provides an

Entities must analyze whether a transaction is a sale or licensing of IP, whether the IP is a distinct performance

obligation, and the nature of the license to determine the timing of revenue recognition (Bellomy & Driscoll, 2020).

New Revenue Standard ASC606

automated revenue management solution for customers, restated its 2018 Q3 earnings in its

initial application of ASC 606 which included a $500 million reduction in revenue. The

reduction was attributed to a misapplication of the new rules, causing confusion and frustration

among analysts and investors. This example is concerning considering that Oracle is the firm that

provides a revenue management system to ensure compliance with ASC 606 (Reddy, 2018). Not

surprisingly, an article in Compliance Week reports that “24 percent of non-reliance notices

(alert investors not to rely on financial statements) are caused by the problems with the ASC 606

adoption in 2018. And 2 percent of late filings in 2018 have cited revenue recognition problems

as the reason for tardy filings.”(Whitehouse, 2019).

Besides figuring out the details of the compliance, firms need to dedicate a significant

amount of time, resources, and expertise to review the contract portfolio affected, update or

overhaul the system, seek advice from auditors/lawyers and compensation committees, and

evaluate the impact of the standard on their financial statements. Many firms found that the

implementation process was much more complex and time-consuming than they had expected

(Peters, 2018). According to a survey conducted by Ernst & Young in 2018, the average cost of

transition to the new standard is estimated to be $3.3 million per company, and 88% of

companies said that they found it challenging to compile data needed for the transition and more

than 80% said they may have to rely on manual work to comply with new requirements

(Shumsky, 2019).

In response to the practical difficulties and challenges of transition to the new standard,

the FASB postponed the original effective date for public companies to December 15, 2017, with

New Revenue Standard ASC606

one-year early adoption permitted (FASB, ASU2015-14)

. In addition, the FASB issued a series

of updates to clarify measurement and presentation issues as well as some practical expedients to

facilitate the transition and implementation of the standard. For example, ASU 2016-08 Revenue

from Contracts with Customers: Principal versus Agent Considerations, provides guidance on

whether revenue should be recognized on a gross or net basis based on the nature of the promise

to provide goods or services to a customer (i.e., a principal or agent destination). ASU 2016-10

Revenue from Contracts with Customers: Identifying Performance Obligations and Licensing,

provides guidance on identifying performance obligations and accounting for licenses of

intellectual property. ASU2014-09 and its amendments (Collectively, ASC 606) are the new

revenue standard in the study.

Despite the complexity and challenges in implementing the changes of the new standard,

a few companies elected to adopt it before the required date (i.e., early adopters). The type of

early adopters varies significantly from large-cap firms such as Microsoft ($857 billion) and

Google ($828 billion) to small companies such as Amerityre ($844,000) (Peters, 2018). It is not

clear if these early adopters exhibit different characteristics from their non-early adopting peers,

what the adoption effects are on financial statements, and how the adoption effects affected their

adoption timing and method decisions. As early adopters, these companies likely had no peer

companies as references regarding the extent and format of new disclosure requirements. This

could have invited additional scrutiny from regulators and market participants, which may have

increased the legal risk of reporting. For example, Peters (2018) indicated that nearly one-third of

ASU 2014-09 defines that for public companies, “this Update is effective for annual reporting periods beginning

after December 15, 2016, including interim periods within that reporting period.” Companies faced obstacles in

implementing ASU 2014-09 and requested deferral in the adoption. The FASB responded to the requests by issuing

ASU 2015-14 and postponed the effective date of ASU 2014-09 to December 15, 2017. Thus, a public company that

adopted ASU 2014-09 with an annual reporting period beginning between December 16, 2016 and December 15,

2017, is an early adopter of ASC 606.

New Revenue Standard ASC606

the early adopters received revenue recognition-related comment letters from the SEC, and some

received as many as six rounds of comments from the SEC, questioning their judgments and

analyses used in applying the five steps in revenue recognition and related presentations. Given

the complexity and challenges of adopting ASC 606, it is unclear what the motivations are for

their early adoption decisions.

One potential consideration in firms’ choice in adopting ASC606 is the standard’s impact

on financial statements

. The new revenue model emphasizes the transfer of control rather than

the satisfaction of performance obligations prescribed under ASC 605 when determining whether

revenue should be recognized at a point in time or over time. This control-based model may lead

to accelerated or deferred revenue recognition, or revenue reclassification under certain

situations. For example, telecommunication companies often offer bundled products at deep

discounts in exchange for multi-period service contracts (e.g., a free cell phone and monthly

subscription services). In the past, equipment and service revenue were treated as a single

performance obligation with revenue recognized over time as service is provided. ASC 606 is

likely to treat them as a distinct performance obligation, resulting in equipment revenue being

recognized upfront at the point of sale and service revenue being recognized over time. AT&T

disclosed in its 2017 annual report that they expect a positive impact on short-term financial

results, and its significantly affected items are the re-classification of service to equipment

revenues and deferral of commission expenses. In contrast, some construction companies and

consulting companies may have to defer revenue recognition when the percentage of completion

The SEC’s initial review of adoption impacts and disclosures indicated that ASC 606 is more likely to affect the

timing of revenue recognition, but has no real effect on cash flows associated with a contract (Peters, 2018)

New Revenue Standard ASC606

method is replaced by the control-based model, where the transfer of control occurs upon the

final delivery of products or services (Hepp, 2018).

The effects of ASC 606 on financial reporting are likely to vary for firms in different

industries due to the fundamental differences in business models and practices, operating

environments, prior GAAP rules, and contract designs

(Choi, Kim & Wang, 2022). Even within

the same industry, the impact of ASC 606 varies because not all contracts in the industry are

necessarily economically similar. To the extent that a company chose a different revenue

accounting method than its peers under ASC 605, the impacts as a result of the new guidance

would be different as well. So far, the nature, extent, and magnitudes of the adoption effects on

financial statements are under-examined (Arms & Bercik, 2015; Marco & Ovuka, 2017;

McKenna, 2017).

In addition, it is not clear why some firms chose to be early adopters knowing that the

FASB had postponed the compliance of ASC 606 due to its complexity and challenges in initial

applications, and they could be subject to a heightened risk of errors and increased scrutiny from

regulators and investors as first movers. Therefore, it is an empirical question to explore the

adoption effects of this standard and examine if there is a systematic difference between early

and non-early adopters in firm characteristics and adoption effects to explain their voluntary

adoption behaviors. This is also a research area that the FASB called for in the revenue research

webinar in March 2021

The changes in accounting treatment on contract acquisition costs and the new disclosure requirements are

expected to affect all industries and entities across the board (Jones & Pagach, 2013; Kral & Levy, 2019).

https://fasb.org/Page/PageContent?PageId=/projects/academics.html#section-2

New Revenue Standard ASC606

This study aims to provide evidence on the impact of ASC 606 on firms’ financials with a

focus on early adopters. Specifically, I examine two research questions: (1) What is the adoption

effect of ASC 606 on the financial statements of early adopters versus non-early adopting peers?

(2) Is there evidence of systematic differences in the characteristics of firms that were early

adopters versus non-early adopting peers?

As discussed earlier, adopting ASC 606 early is a significant accounting choice firms

have to make because they had to commit great resources and tackle many challenges. This

group of firms is likely to expect more benefits from the changes in accounting by overcoming

time and resource constraints. Focusing on this group will provide evidence of the impacts of

ASC 606 and complement concurrent studies that have mixed findings regarding the impact of

ASC 606 on the stock market. By adopting a new standard early, firms may use the one-time

earnings management opportunity to signal their ability and quality of reporting, smooth

transitory increase or decrease in earnings, and provide the stock market with information

regarding their long-term expectations (Eakin & Gramlich, 2000).

ASC 606 provides an interesting setting for early adoption research in the accounting

choice literature. This is because ASC 606 directly affects the top-line number in addition to the

bottom-line number on income statements. Prior accounting changes in early adoption studies

affected only the bottom-line numbers of the income statement. For example, Ayres (1986)

examined the characteristics of firms early adopting SFAS 52 - Accounting for Foreign Currency

Translations, where the gains and losses from foreign currency translation are reported under

owners’ equity rather than the income statement under the previous standard. Sami and Welsh

(1992) studied the characteristics of firms early adopting SFAS 87 - Employers’ Accounting for

Pensions, where the major effect of the new standard is the method used to measure net periodic

New Revenue Standard ASC606

pension cost and recognize the minimum liability provision for underfunded plans. Gujarathi and

Hoskin (2003) studied the characteristics of firms early adopting SFAS 96 – Accounting for

Income Taxes, where the new standard changed accounting for income tax recognition and

accrual. Unlike prior accounting changes studied in the prior research, ASC 606 directly affects

the top-line number, which is critical to users in evaluating firms’ performance, trends, and

prospects. Given the importance of ASC 606, my study contributes to both ASC 606 literature

and the accounting choice literature that are described below.

ASC 606 literature is new and growing. Extant research on ASC 606 focuses on

examining the market consequences of the standard focusing on materially affected firms, and

their findings are mixed. For example, Lee and Lee (2020) found that ASC 606 decreased

earnings predictability while Chung and Chuwonganant (2019) found that ASC 606 improves the

informativeness of earnings. Complementing the concurrent studies that focus on the market

effects of ASC 606, this study chooses to examine the direct impact of ASC 606 on firms’

financials and examine whether early adopters have certain firm-specific characteristics and

benefits from their voluntary adoption decisions. The analyses will provide insight into the

reporting benefits of the standard and how the benefits vary with the characteristics of firms,

which in turn will further our understanding of the impact of the new standard.

The accounting choice literature suggested that voluntary adoption decisions have

important implications for firms and users of financial statements. Prior studies document that

early adopters consist of firms that have unique characteristics and tend to be strategic in making

adoption decisions. They have economic incentives to take advantage of the unique earnings

management opportunity to adopt a new standard early before required (Ali & Kumar, 1994;

New Revenue Standard ASC606

Ayres, 1986; Eakin & Gramlich, 2000; Gujarathi & Hoskin, 2003; Langer & Lev, 1993; Sami &

Welsh, 1992; Simon & Costigan, 1996; Trombley, 1989; Zmijewski & Hagerman, 1981).

This study will contribute to the literature on ASC 606 and the voluntary adoption

literature in the following ways. First, this study will enhance users’ understanding of the new

revenue standard and its financial reporting impact to make well-informed decisions. Second,

this study may be interesting to practitioners in identifying earnings management opportunities

for the best interest of stakeholders. Third, the findings may be interesting to standard-setters to

understand the implications of accounting choice to determine an optimal level of discretion

allowed when issuing new rules (Fields et al., 2001; Watts & Zimmerman, 1990). Lastly, this

research may provide useful information for the FASB’s post-implementation review of ASC

606 (FASB, 2020-07-29)

The remainder of the paper proceeds as follows. Chapter 2 discusses the background and

prior research in revenue standards and voluntary accounting choices followed by hypothesis

development. Chapter 3 explains the sample selection process and the research method. Chapter

4 presents the quantitative evidence of summary statistics and logistic regressions with

discussions of the results. Chapter 5 discusses the research implications, contributions, and future

research opportunities.

https://www.fasb.org/Page/PageContent?PageId=/pir/pir.html&isstaticpage=true#section-2

New Revenue Standard ASC606

Chapter 2: Literature Review

2.1. History of US Accounting Authorities and Pronouncements

The promulgation of financial accounting standards in the U.S. is a political process

influenced by numerous parties, including the Securities Exchange Commission (SEC), standard-

setting bodies such as the FASB, the corporate sector, public accountants, and users of financial

statements. The SEC was given the authority to prescribe accounting standards in the Securities

Act of 1933 but has chosen to delegate this responsibility to the private sector

The earliest accounting standard-setting body delegated by the SEC can be traced back to

the predecessor of the American Institute of Certified Public Accountants (AICPA) – the

American Institute of Accountants (AIA), which established a subcommittee - Committee on

Accounting Procedure (CAP) in 1938 to develop accounting standards. The CAP issued 51

accounting standards known as Accounting Research Bulletins (ARBs) during its twenty-one

years of existence. In 1959, The Accounting Principles Board (APB) replaced the CAP and

issued 31 APB opinions and four statements during its brief existence. In 1973, the FASB

replaced the APB and has been responsible for establishing and interpreting financial accounting

and reporting standards for public, private, and not-for-profit organizations since then. The

FASB pronouncements are considered as having substantial authoritative support by the SEC in

its Accounting Series Release (ASR) 150 - Statement of Policy on Establishment and

Improvement of Accounting Principles and Standards, and by the AICPA in its Code of

Professional Conduct under Rule 203 – Accounting Principles (Storey and Storey, 1998).

In 1978, the FASB established the Conceptual Framework for Financial Accounting and

Reporting which provides the basic objectives and fundamentals underlying financial reporting.

https://www.sec.gov/news/speech/1989/011089ruder.pdf

New Revenue Standard ASC606

The framework is not authoritative but is used as guidance to develop accounting principles and

analyze complex or emerging practice issues in the absence of applicable authoritative

pronouncements. Over the years, the FASB has issued eight Statements of Financial Accounting

Concepts (SFAC) that provide a broad overview of accounting concepts (Drew, 2021), and

numerous Statements of Financial Accounting Standards (SFAS) that become the generally

accepted accounting principles once published. These generally accepted accounting principles

govern the financial reporting of publicly traded companies and are overseen by the SEC. Since

2009, both SFAC and SFAS have been superseded by the FASB Accounting Standards

Codification (ASC). The codification is updated via Accounting Standards Updates (ASUs) and

the FASB Conceptual Statements.

2.2 Background on Revenue Standard-Setting

Revenue literature in the U.S. includes broad conceptual statements as well as industry-

specific rules that were developed over the years by the standard setters that existed at the time.

SFAC No. 6 defines revenues as inflows or enhancements of assets or settlements of liabilities

from delivering goods or services from an entity’s main operations. The SFAC No. 5 includes

core concepts of revenue recognition and measurement, where recognition is defined as a process

of formally recording an item into the financial statements, and measurement is choosing a

reliable method to quantify the item recognized (Storey & Storey, 1998). It specifies that revenue

is not recognized until it is “realized or realizable, and earned” (SFAC No.5: page 30). While

SFAC No. 5 provides general guidelines for revenue recognition, ASC 605, and numerous

industry- or transaction-specific requirements (e.g., SFAS 66 – Accounting for Sales of Real

Estate or ASC 985-605 - Software Revenue Recognition) provide detailed authoritative guidance

for revenue recording. ASC 605 requires four criteria for revenue recognition: (1) persuasive

New Revenue Standard ASC606

evidence of an arrangement exists, (2) delivery of goods or services has occurred, (3) the selling

price is fixed or determinable, and (4) collectability is reasonably assured. In 1999, the SEC

issued the Staff Accounting Bulletin (SAB) No. 101 (amended by SAB 104 in 2003) that

provides interpretive guidance on the four criteria that firms must meet to recognize revenue

(SEC, 1999).

The previous GAAP rules under ASC 605 Revenue Recognition are perceived more as a

rules-based accounting approach that comprises over 200 industry- or transaction-specific rules.

These rules are lengthy, complex, and difficult to implement with numerous restrictions,

exceptions, and bright-line tests (McCarthy & McCarthy, 2014). The advantages of a rules-based

system include increased transparency and verifiability in financial reporting, reduced ambiguity

in compliance, and reduced litigation risks (Benston et al., 2006; Donelson et al., 2012; Lee &

Lee, 2020). However, a rules-based method has been criticized for being overly prescriptive and

conflicting in some areas, resulting in a lack of reporting discretion and lack of consistency in

revenue recognition. Schipper (2003) suggested that detailed guidance in GAAP is intended to

increase comparability, but when the guidance is too strict with little room for discretion, the

rules-based reporting system becomes undesirable as it leads to technical compliance and

achieves only surface comparability. Moreover, voluminous guidelines and numerous

restrictions/exceptions can cause unnecessary complications in preparing financial statements

and managers may try to structure transactions to circumvent the true spirit of standards,

resulting in a violation of a critical accounting concept - substance over form. This concept is

particularly relevant in revenue recognition for reliable financial reporting (SEC, 2003). The

SEC indicated that a rules-based accounting system induced companies to circumvent

accounting rules to meet the letter but not the intent of standards – “an act of compliance rather

New Revenue Standard ASC606

than an act of communication” (SEC, 2003: p. 5). It was given as one of the reasons for a series

of high-profile revenue misreporting scandals in the early 2000s.

The shortcomings of revenue recognition requirements were one of the reasons that high-

tech and software firms regularly use non-GAAP measures and complained that previous

guidance did not fully reflect the value of their new product launches. For example, Twitter

reported a net loss of $521 million under GAAP in 2015, but a net income of $276 million using

non-GAAP measures (Sultanoglu, 2017). The differences are attributed to the stringent vendor-

specific objective evidence (VSOE) requirements under the old subscription accounting rules

Software firms often provide multiple deliverables (e.g., hardware, software, hosting,

maintenance, and support services) that are delivered at varying times but are often offered under

a total fee arrangement. Under the old rules, revenue for an individual service item could be

recognized separately only if a company could establish vendor-specific objective evidence

(VSOE) for the item - the fair market value if the item is sold separately. Otherwise, some or

even all of the multiple arrangements’ associated revenue would have to be deferred. However,

VSOE analysis is based on stringent rules and tested by different methods. In practice, it is

extremely challenging for firms to establish VSOE, and it is sometimes impossible to measure

the fair value of an element that is not sold separately (Regan & Regan, 2007). These complex

accounting rules were criticized for causing a significant understatement of revenue, the

inconsistent accounting treatment for economically similar transactions, and the incomparability

of financial results. Unsurprisingly, firms such as Apple, Microsoft, Amazon, Facebook, and

SOP (Statement of Position) 91-1 specifies that software-license revenue is recognized upon delivery and service

revenue is recognized proportionally over time. SOP 97-2 provides further guidance regarding revenue recognition

timing and the amount of deferred revenue for delivered and undelivered elements. It requires firms to establish

VSOE for each element in multiple deliverables before revenue is individually recognized (Choi, Kim & Wang,

2022).

New Revenue Standard ASC606

Twitter lobbied for changes in the revenue rules and the elimination of the VSOE requirements

(Leone, 2009).

Despite a large number of rules, there were still transactions that had no direct

authoritative guidance under GAAP, and some requirements dated back to various authoritative

predecessors of AICPA which resulted in inconsistent accounting treatment for economically

similar transactions. Investors had a hard time comparing firms’ financial performance across

sectors without knowing the nuances of different revenue rules. With the rapid development of

innovative business models and more complex transactions, some rules became obsolete and

ineffective in reflecting the underlying economic substance of transactions (FASB, ASU 2014-

09; SEC, 2003).

Additionally, there were issues and concerns about revenue recognition requirements in

IFRS (International Financial Reporting Standards) as well. Previous revenue guidance under

IFRS is viewed largely as a principles-based approach that provides general guidelines for a wide

range of transactions and diverse business models. It is intended to reduce complexities and

potential manipulations of rules in financial reporting (Schipper, 2003). However, the overly-

broad standards under the legacy International Accounting Standards (IAS) 18 Revenue and IAS

11 Construction Contracts and related interpretations have been criticized for providing limited

guidance for complex revenue transactions such as multiple-element arrangements, inadequate

disclosures, and insufficient structure to frame professional judgments. As a result, some IFRS

firms (e.g., software) had to refer to US GAAP rules for reporting revenue. IFRS’s rules have

been criticized as having caused inconsistent practices for revenue recognition, less trustworthy

and auditable procedures, increased likelihood of disagreements on accounting treatments, and a

significant loss of comparability among reporting entities (Benston et al, 2006; Bjornsen &

New Revenue Standard ASC606

Fornaro, 2019; Guillaume & Pierre, 2016; McCarthy & McCarthy, 2014; Papa, 2016; Peters,

2016). As a result, there was also a call for new revenue standards in the IFRS regime of

standards.

Last but not least, disclosure requirements under previous GAAP and IFRS were “limited

and lacked cohesion” (FASB, ASU 2014-09, p.8). They were limited to generic boilerplate

descriptions of a firm’s accounting policies and their effects on revenue, however, investors and

users of financial statements indicated that the disclosures were insufficient for analyzing a

firm’s performance and prospects.

The deficiencies in both US GAAP and IFRS led to the boards’ joint decisions to issue

the Norwalk agreement in 2002 and the Memorandum of Understanding in 2006 to develop a

global revenue standard that can be applied to all contracts with customers regardless of

industry- and transaction-specific patterns. The SEC indicated that neither GAAP nor IFRS

reached an optimal balance in the various trade-offs inherent in revenue standard setting (e.g.,

relevance and reliability, or costs and benefits of information production). They suggested a

move toward an objectives-oriented approach in developing the new revenue standards based on

a high-quality theoretical framework. The new standards should have clearly stated accounting

objectives, sufficient and operational rules, minimum exceptions, and minimum bright-line tests

that require meeting prescribed thresholds (SEC, 2003).

After working collaboratively for over a decade, the FASB and IASB finally issued a

unified revenue standard Revenue from Contracts with Customer on May 28, 2014 (i.e., ASC

Topic 606 for US firms or IFRS 15 for international firms). The length of time taken to complete

the revenue project underscores the complexity of the topic. This revenue standard supersedes all

existing revenue rules in GAAP and IFRS and significantly reduces the number of requirements

New Revenue Standard ASC606

in recognizing revenue. It aims to remove weaknesses, inconsistencies, and complexities in

previous requirements, streamline and harmonize revenue recognition practices, enhance

disclosures, and provide a more robust framework that allows for better comparability and

decision-usefulness of financial statements (FASB, ASU 2014-09). The boards believe that the

long-term benefits will outweigh the short-term costs of implementation.

This research focuses on U.S. public companies adopting ASC 606, thus the discussion of

IFRS 15 is outside the scope of this study.

2.3. ASC 606 New Revenue Standard

In the U.S., ASC 606 supersedes ASC 605 Revenue Recognition and affects almost all

entities (public and non-public) that enter into contracts with customers to transfer goods or

services unless they are subject to other rules (e.g. insurance contracts, lease contracts, and

financial instruments). Given the large percentage of contracts with customers in most

businesses, this new guidance is expected to bring the most influential accounting changes to

corporations since the Sarbanes-Oxley Act of 2002 (Thorps et al., 2020; Tzuo, 2017).

The new revenue standard moves from a rules-based approach to a principles-based

approach and consists of the following five steps:

Table 1. The Five-Step Model of ASC 606

Step 1:

Step 2:

Step 3:

Step 4:

Step 5:

Identify the

contract(s)

with a

customer

Identify

performance

obligations

(POs)

Determine the

transaction

price

Allocate the

transaction

price to each

Recognize

revenue

when/as a PO

is satisfied

Step 1. ASC 606 defines a contract as an agreement between two or more parties that has

enforceable rights and obligations, has commercial substance, and meets a collectability

New Revenue Standard ASC606

threshold which is crucial in determining whether a contract exists. It can be written, oral, or

implied by business practices. The central requirements for contract existence are intended to

prevent fraudulent or abusive contract practices, such as fictitious contracts, side letter

agreements (SLA), round-trip transactions, bill-and-hold arrangements, or channel stuffing

(Carmichael, 2019; Thorps et al., 2020).

Step 2. Performance obligations are enforceable promises to transfer goods or services to a

customer, and each good or service that can be separately sold or identified is considered a

distinct performance obligation to be recognized separately. If they have substantially the same

pattern of transfer, they are accounted for as a series of bundled goods or services (combined

output) to transfer to the customer.

Step 3. Transaction price refers to the amount of consideration an entity is entitled to receive in

exchange for transferring promised goods or services to a customer, excluding the amounts

received on behalf of third parties (e.g., sales taxes). The amount of considerations can be fixed

or variable. Variable considerations such as discounts, rebates, refunds, performance

bonuses/penalties, or incentives are contingent upon the occurrence of uncertain future events

and need to be included in the transaction price at contract inception if they can be reasonably

estimated (use either the expected value method or the most likely amount method) and are not

expected to be reversed. Significant financing components also need to be considered when

determining the transaction price.

Step 4. When the total transaction price involves multiple performance obligations, the

transaction price needs to be allocated to each distinct performance obligation identified in step 3

based on its stand-alone selling price directly observable or estimated (e.g., expected cost plus

margin).

New Revenue Standard ASC606

Step 5. Revenue is recognized when a performance obligation is satisfied, i.e. a customer has

substantially obtained control of goods or services. Performance obligations can be satisfied at a

point in time or over time by measuring the progress of completion.

This 5-step model replaced excessive, disparate, and rigid industry-specific rules with an

overarching framework to standardize the revenue reporting process that allows more

management discretion and judgment. The core principle underlying ASC 606 is that an entity

recognizes revenue to reflect the transfer of promised goods or services to a customer in the

amount to which it is entitled. The first four steps help identify a seller’s promises to customers

and the amount of price associated with each promise, and the last step helps identify the pattern

of transfer of promised goods or services to determine whether revenue should be recognized at a

point in time or over time.

ASC 606 primarily affects the timing, measurement, and disclosures of revenue, as

explained below:

First, ASC 606 changes the timing of revenue recognition. This is because the new

revenue model shifts away from the prior risks and rewards model (“earned, realized, or

realizable”) and focuses on the transfer of control as the criteria to determine the timing of

revenue recognition, where control is defined as “the ability to direct the use of and obtain

substantially all of the remaining benefits from the asset” (FASB, ASU 2014-09: page 28). The

determination of control is evaluated from the customers’ perspective based on indicators of

transfer of control such as physical possession, legal title, and customer acceptance. In practice,

control of an asset may have been passed to a customer while the seller still carries risks related

to the asset, thus revenue may be recognized under ASC 606 but not under ASC 605. In other

words, the possibility of transferring assets early before the associated risks could lead to

New Revenue Standard ASC606

accelerated revenue recognition. For example, large aerospace & defense (A&D) companies

such as Boeing, Lockheed Martin Corp, and Raytheon Corp. make customized goods for

customers. Under ASC 605, revenue was recognized when the title of goods was transferred to a

customer upon shipping or delivery. ASC 606 is likely to accelerate revenue reporting because

the seller is creating an asset that has no alternative use, which means that the control of goods

may be transferred to customers at the stage of work in process. For U.S. government contracts,

contractors usually do not have an alternative use for customized goods. The customer has

substantial control of the work-in-process inventory as aerospace & defense contracts allow the

customer to unilaterally terminate the contract for convenience. Therefore, revenue would be

recognized earlier under ASC 606 (McKenna, 2017; Wilks & Ellsworth, 2018).

In addition, the new standard requires firms to capitalize on incremental costs of

obtaining a contract (e.g., sales commissions, contingent legal fees, and travel costs) with a

customer rather than expense them if the costs are expected to be recovered. And ASC 606

breaks down performance obligations into more granular units (Ali & Tseng, 2022), which eases

requirements in software revenue. For example, ASC 605 only allowed the breakdown of the

product and support revenue, but ASC 606 allows the breakdown of the sale of software into

multiple performance obligations

. All these changes affect the timing of revenue recognition.

Second, ASC 606 affects the measurement of revenue recognition. When measuring the

amount of revenue, the new standard requires that variable consideration (e.g., performance

bonus, returns, rebates, usage-based charges), significant financing components (payments in

advance more than a year), and non-cash consideration be evaluated and estimated in

https://www.performio.co/insight/what-is-asc-

606#:~:text=While%20many%20of%20the%20changes,revenue%20to%20be%20broken%20up.

New Revenue Standard ASC606

determining the transaction price. Many retailers assume that the new standard probably has little

impact on their revenue recognition practices as revenue is typically recognized at the point of

sale. However, the new standard may significantly affect the accounting treatment for gift cards,

customer incentives, customer loyalty programs, and rights of return (Kishi, 2020).

Measuring variable considerations is challenging in practice as it involves a lot of

uncertainty and contingent events, which requires managers to apply increased judgment on

when control of goods or services transfers to customers. For example, a customer loyalty

program is where a past sales contract gives a customer an option to purchase goods at a lower

price in the future. Under ASC 606, the customer’s option to purchase goods or services could

become a separate performance obligation to be recognized as part of the original contract. This

could affect the allocation between current and future revenues, assets, and liabilities (Hepp,

2018). Arora (2019) assessed the impact of ASC606 on eight US airlines and found that the

aggregated income of these airline firms increased by 2% and the frequent flyer program (FFP)

liability increased by 76.93% due to the adoption of new rules. Estimates of the FFP value

represent one of the most uncertain areas in the airline industry as the valuation method changes

from the incremental cost method to the deferred revenue method in ASC 606. The liabilities of

loyalty points recorded under the incremental cost method is likely to be significantly lower than

those recorded under the deferred revenue method which uses the stand-alone redemption prices

rather than the historical costs to estimate the FV of liabilities. As expected, Delta Airlines

increased its FFP liability by 2.2 billion due to the change of valuation method under ASC 606 as

of the date of initial application (2018 Annual Report, Note 1). The highly volatile nature of

airfares indicates that the FV of mileage credits may constantly change. Therefore, variable

New Revenue Standard ASC606

consideration represents a major area of uncertainty and requires more judgment under ASC 606

(Gordon et al., 2018).

Third, the new standard significantly increases the amount of information companies are

required to disclose regarding the nature, amount, timing, and uncertainty of revenue and cash

flows arising from contracts with customers. The new disclosure requirements are more detailed

and comprehensive than previous standards. “It is as if your teacher isn’t just demanding that you

show your work, but also that you write an in-depth essay explaining the approach you chose,

why you chose it, what assumptions you made, what tools you used, and what processes you

followed to ensure nothing would go wrong.” (Knachel, 2017: page 1).

Under ASC 606, firms must disclose revenue policies and explain the analysis and

decisions made when applying the 5-step model, including more extensive disclosures on

separate revenue streams, deferred customer contract acquisition and fulfillment costs, changes

in contract assets and liabilities, descriptions of transaction prices allocated to remaining

performance obligations and significant judgments made in the timing and amount of revenue

recognition (Olsen & Weirich 2010; Streaser et al., 2014; Yeaton, 2015).

As an example, AT&T disclosed in its 2018 annual report that “As of December 31,

2018, the aggregate amount of the transaction price allocated to remaining performance

obligations was $39,871 million of which we expect to recognize approximately 55% next year

and 80% cumulatively over the next two years, with the balance recognized thereafter” (AT&T

2018 Annual Report, Note 5). It appears that the length of textual content related to ASC 606

disclosures has increased significantly for many companies, especially those that have a material

impact from the adoption in measurement and presentation (Lee & Lee, 2020).

New Revenue Standard ASC606

2.4. Transition Requirements and Choices

Initially, ASC 606 was effective for public companies with reporting periods beginning

after December 15, 2016, including interim periods within that reporting period. Due to

significant concerns about implementation complexities and transition challenges, the FASB

postponed the effective date by one year. Thus, ASC 606 is effective for US public firms with

reporting periods beginning after December 15, 2017. Most public companies adopted ASC 606

in the year of 2018 (see Table 2). For non-public firms, it was first postponed from December 15,

2017 to December 15, 2018 (FASB, ASU 2015-14). Then, the FASB further extended the

effective date for certain non-public firms by another year, noting the practical difficulties and

challenges of transition to new standards amplified by the business and market disruptions

caused by the coronavirus pandemic (FASB, ASU 2020-05).

To assist with the transition to the new standard, the FASB allows firms to make two

choices in the year of transition: adoption timing and transition method. In terms of adoption

timing, firms can elect to adopt it early before the effective date (early adopters) or wait till it

becomes mandatory (non-early adopters). In terms of the transition method, firms can choose

either the full retrospective method (FRM) or the modified retrospective method (MRM) in the

initial adoption. The FRM requires retroactive restatements of all prior reporting periods

presented on comparative financial statements as if ASC 606 were effective for those periods;

The MRM requires only an adjustment to beginning retained earnings to recognize the

cumulative effects of the new standard at the date of initial application, and additional

disclosures about amounts and explanations of significant changes made to financial statements

in the transition year. Prior periods’ amounts are not adjusted and are still reported under historic

accounting guidance (see Table 2).

New Revenue Standard ASC606

Table 2: Transition Timing and Method

Transition Method

2016

2017

2018

FRM

Comparative years restated as if ASC 606

was effective and cumulative adjustments

made at the earliest reporting period

Required adoption of

ASC 606

MRM

ASC 605

Required adoption of

ASC 606 and cumulative

adjustments made at the

adoption year

Originally, the FRM method was the only method proposed in the initial 2010 and 2011

exposure drafts as it allows a comparison of financial results across comparative periods and

provides consistent revenue trend data and peer comparisons to investors, but feedback from

entities indicated that this method could be too costly and burdensome to implement (Peters,

2016). KPMG’s Accounting Change Survey (2016) indicated that most companies were

concerned about the data availability, the competing priorities, the resource constraints, and the

time needed to have the new systems in place to accommodate the complexities in the FRM

transition method. As a compromise, the MRM method is allowed as a practical expedient

approach to ease the transition process, which was especially popular for companies that

expected a minimal impact on trends. Under the MRM, the comparative years would not be

restated, but additional disclosures that explain significant changes in the line items of financial

statements, the reasons for these changes, and a numerical reconciliation of key accounts

reported under ASC 605 and ASC 606 would be required (Kral & Levy, 2019).

In a report published by Financial Executives International (FEI) in November 2017,

15% of Fortune 500 companies disclosed that they would use the full retrospective method, 71%

disclosed that they would use the modified retrospective method, and the rest did not disclose

their selections (Macro & Ovuka, 2017).

New Revenue Standard ASC606

Hunsaker (2018) indicated that the accounting choice of adoption timing and transition

method may affect the outcome and presentation of financial statements. Under certain

circumstances, ASC 606 may cause firms to appear to “lose revenue” if ASC 606 accelerates the

timing of revenue recognition for certain contracts that are still outstanding under ASC 605 (see

Table 2A), which may have important implications for firms and users of financial statements.

Table 2A: Illustration of Transition Timing and Method*

ASC 605

2016

2017

2018

Total

Revenue

$500

$1,500

Beg. Retained Earnings

$500

$1,000

MRM Adoption of ASC 606: January 1, 2018

2016

2017

2018

Total

Revenue

$500

$250

$1,250

Beg. Retained Earnings

$500

$1,250

FRM Adoption of ASC 606: January 1, 2018

2016

2017

2018

Total

Revenue

$1,000

$250

$1,500

Beg. Retained Earnings

$1,000

$1,250

* A software firm offers a bundled product and service for $1,500 (software: $750; 3 years' service: $750). Under ASC 605,

the revenue of $1,500 is recognized pro rata over 3 years (i.e., $500/yr). ASC 606 divides the contract into two performance

obligations (software and service) and accelerates software revenue upon delivery (i.e., $750 software revenue + $250 service

revenue = $1,000 in year 1) - adapted from Hunsker (2018)

If a firm uses the MRM to adopt ASC 606, the accelerated revenue will bypass the

income statement and affect the current-year retained earnings because the MRM does not

require a recast of income statements in pre-adoption years. As a result, the revenue in the

adoption period will be less under ASC 606 than that reported under ASC 605. If a firm uses the

FRM, the cumulative effect of changes would be reflected in the earliest reporting period (i.e.,

2016). And any impact on revenue would be spread out over multiple years rather than the

New Revenue Standard ASC606

current reporting year only. Gujarathi and Hoskin (2003) indicate that the ability to bypass

income statements represents an opportunity for firms to achieve desired financial results in

different reporting periods. The transition choices allow firms to choose a method of adoption

that suits their preference for the reported numbers on comparative financial statements. Thus,

the management decisions of adoption timing and method are relevant factors affecting the

financial results in the year of adoption.

2.5. Existing Research on ASC 606

ASC 606 is introducing significant changes to revenue reporting and its impact is likely

to go beyond firms’ financials to influence other dimensions of businesses, such as customer

contracting practices, information technology systems and processes, compensation plans, debt

covenants, and tax positions. However, the specific impacts of ASC 606 are not well understood.

According to a survey conducted by Deloitte in 2018, only 11% of the participants were familiar

with ASC 606 and only 8% assessed the impact of the new guidance on their revenue streams

(Nehoray et al., 2018). Market observers predicted that the new guidance may have more impact

on the software, aerospace and defense, airline, and telecommunication industries (Arms &

Bercik, 2015; Marco & Ovuka, 2017; Mckenna, 2017), but there is a lack of evidence on the

nature, extent, and magnitudes of its impacts on firms’ financial statements (Ali & Tseng, 2022).

The literature on ASC 606 is new and growing. Concurrent studies that examine the

impacts of ASC 606 on the market focus on materially affected firms, and their results are mixed

and at times contradictory as to whether ASC 606 has improved the informativeness of earnings.

For example, Lee and Lee (2020) examined the effect of ASC 606 on earnings predictability and

the debt market using a difference-in-difference design. They collected the sample firms Russell

3000 firms on Audit Analytics and separated the sample into a treatment group (firms reporting a

New Revenue Standard ASC606

material impact of ASC 606 based on firms’ disclosures) and a control group (the rest). Their

underlying assumption is that companies that have less predictable earnings are likely to have

larger forecast errors and distributions. They found that the new standard is associated with

increased analysts’ forecast errors and increased discretionary working capital accruals,

indicating that the new standard is associated with decreased earnings predictability and

increased opportunistic behaviors. The decreased earnings predictability could reduce the

reliability and usefulness of earnings in the debt market as creditors were more willing to use

capital-based covenants rather than earnings-based covenants as observed in the study.

In contrast, Chung and Chuwonganant (2019) evaluated the effect of ASC 606 on market

reactions to earnings announcements. They conjecture that the enhanced disclosure requirements

in ASC 606 improve the quality and quantity of information regarding firms’ revenue, thus, they

would increase the informativeness of earnings. Their study found that the new standard is

associated with increased market efficiency, stock liquidity, and decreased information

asymmetry between traders, suggesting that ASC 606 increases the informativeness of earnings,

and the effect is more significant for industries with larger recurring revenues.

Rutledge et al. (2016) indicated that the new standard is expected to produce more

relevant financial information with increased comparability, but at the same time, it also provides

more room for management estimates and judgment which may reduce the informativeness of

earnings, suggesting that the direction of the market effect is not clear.

Glaze et al. (2021) examined whether quarterly reports issued concurrently with earnings

news bring useful information to investors under the ASC 606 setting. They assume that a new

accounting standard is associated with increased uncertainty which may suppress the market

response. McClane (2019) indicates that financial statements may become less useful due to the

New Revenue Standard ASC606

heavy volume of mandatory disclosures, causing low understandability of financial statements

and potential information overload. Their studies found that the quarterly reports issued

concurrently with earnings news are associated with increased trading because these interim

reports contain more detailed and value-relevant information about earnings which helps reduce

ASC606-related uncertainty and help investors understand and respond to the earnings surprise.

Ahn et al. (2021) examined the disclosure requirements under SAB 74 in the context of

ASC 606

. ASB 74 requires firms to disclose the potential adoption effects of a new accounting

standard in the quarters preceding the implementation. They found that variations exist among

firms within the materially impacted group, and the SAB 74 disclosures are associated with

reduced information asymmetry and other financial reporting issues, especially for companies

with more complex adoptions in ASC 606.

Ferreira (2020) collected data from Compustat and used a difference-in-difference design

to examine the market effects of implementing the new revenue standard. Assuming that the

converged guidance increases the comparability of financial statements, investors are likely to

incur fewer information costs to compare firms’ performance across industries or capital markets

because they do not have to know the nuances of industry-specific rules under ASC 605. As a

result, the number of informed investors would increase and stock prices would become more

informative. A more informative stock price leads to higher liquidity. Their results provide

evidence that ASC 606 is associated with increased liquidity through an improvement in

reporting precision and comparability of financial statements.

The SEC Staff Accounting Bulletin No. 74 (SAB 74) requires that when a recently issued accounting standard has

not yet been adopted, a registrant disclose the potential effects of the future adoption in its interim and annual SEC

filings to provide insight into a registrant’s preparedness to a new accounting standard (SEC, SAB Topic 11.M)

New Revenue Standard ASC606

Hu (2021) used a difference-in-difference design to examine whether the new standard

has improved comparability between US firms and IFRS firms that report in the US market. The

author noted that comparability only improved in the industries that are more affected by the new

standard, in particular, the effect is more pronounced in the telecommunication and software

industries.

In summary, most of these studies mainly tested the market reactions to ASC 606

focusing on materially affected firms, and their findings are mixed (see Table 3). Although it is

important to study the effects of ASC 606 on market activities, the study of adoption effects on

financial statements and the adoption choices in ASC 606 will complement the market-focused

studies.

Table 3. ASC 606 Literature

Research Topic - ASC 606

Findings

Lee & Lee (2020)

Earnings quality

Earnings predictability (-)*

Chung et al (2019)

Market efficiency

Informativeness of earnings & liquidity (+)

Glaze et al (2021)

10-Q disclosures

Value relevance of earning & trading (+)

Ahn et al (2021)

Preadoption disclosures (SAB 74)

Information asymmetry (-)

Ferreira (2020)

Liquidity effect

Value relevance & liquidity (+)

Hu (2021)

Comparability (GAAP & IFRS)

Comparability in telecom. & software (+)

* (+) or (-) indicates an increased or decreased effect

2.6 Research Interest

As discussed above, existing studies on ASC 606 focus on the information perspective to

examine the relationship between the new standard and its market effect, using tests of market

efficiency and value relevance for new accounting information. These studies provide evidence

of the market’s response to this mandatory accounting change. However, the voluntary aspect of

the adoption choices has not been examined. For example, why did some firms choose to adopt

New Revenue Standard ASC606

the standard early? Are firm characteristics and adoption effects associated with early adoption

decisions?

The adoption decision for ASC 606 is a significant accounting choice. As discussed

earlier, the costs and efforts involved in ASC 606 applications were so significant that not only

the original two-year preparation periods (i.e., the time between the issuance date and the

original effective date) were deemed insufficient, but also an additional year of an extension was

needed (Ernst & Young, 2018; FASB, ASU 2015-04).

As first movers, early adopters of a significant accounting standard could potentially be

subject to a higher risk of errors, higher litigation risk, and stricter scrutiny from regulators. For

example, nearly one-third of early adopters received revenue recognition-related comment letters

from the SEC and were questioned about their judgments, analyses, and disclosure in applying

the standard. Thus, early adopters provide an interesting setting to understand the impact of ASC

606 and the potential motivations behind the adoption decisions. Specifically, this study asks and

examines the following questions: (1) What is the adoption effect of ASC 606 on the financial

statements of early adopters versus non-early adopting peers? (2) Is there evidence of systematic

differences in the characteristics of firms that were early adopters versus non-early adopting

peers?

Accounting choice literature suggests that managers have economic incentives to make

accounting choices to obtain desired financial results. Most of these studies focused on the

contracting perspective in positive accounting theory to explain and predict accounting practices.

Assuming information asymmetry in an imperfect market, the theory predicts that managers

select accounting methods to maximize their utility which is highly associated with explicit and

implicit contracting costs. These contracting costs arise in internal transactions as well as

New Revenue Standard ASC606

external transactions and consist of transaction costs, agency costs, political costs, bankruptcy

costs, etc. It is hypothesized that firms have incentives to minimize transaction costs (e.g.,

brokerage fees and information production costs), agency costs (e.g. costs of compensation and

debt contracts), and political costs (e.g., taxes and regulation costs) to maximize firm value and

self-interest (Watts & Zimmerman, 1990).

Fields et al. (2001) reviewed and summarized prior accounting choice studies and defined

an accounting choice as “any decision whose primary purpose is to influence the output of the

accounting system.” (Fields et al., 2001: page 256). This broad definition includes accounting

choices that have real effects on cash flows such as production or investment decisions (e.g.,

reducing R&D expenditures), structuring transactions in certain ways (e.g., operating vs capital

lease), or choices of accounting methods (e.g., FIFO vs. LIFO). It also includes accounting

choices in the timing of adoption of a new standard that does not affect cash flows but has effects

on reported income presented on the comparative financial statements. Research shows that

managing the timing of the adoption of a new accounting standard is a strategy to manage

earnings to obtain desired results. To the extent that reported income affects a firm’s economic

value through its implications for future performance, firms have incentives to use income-

increasing or income-decreasing strategies to influence reported income (Fields, 2001; Han &

Hsiao, 2017; Zmijewski & Hagerman, 1981).

Most accounting choice studies assume that accounting numbers matter because they

play an important role in contractual relations (relations among management, shareholders, and

creditors), information dissemination (relations between informed managers and less-informed

investors), and public policy implications. For example, Fields et al. (2001) summarized the

importance of reported numbers in three areas: agency costs - managers choose accounting

New Revenue Standard ASC606

methods to maximize managerial compensation and avoid debt covenant violations, information

asymmetries – managers communicate private information through accounting choice to

influence stock prices, and political costs - managers select accounting methods to avoid

potential regulations or reduce/defer taxes. In general, managers make accounting choices to

influence the output of the accounting system that is beneficial to themselves and their firms.

However, there has not been a consensus on what motivates managers to choose one accounting

alternative over another, and what the determinants and economic implications of these

accounting choices are to users of financial statements. This is because managerial motivations,

determinants, and implications of different accounting choices vary by the specific context.

The implication of making accounting choices to influence users of financial statements

is consistent with accounting research on earnings management (Fields et al., 2001; Watts &

Zimmerman; 1990). Revenue/earnings, the most scrutinized numbers in a company’s financial

statements, provide a very informative signal to stakeholders that facilitate resource allocation

and wealth redistribution in capital markets. It is not surprising that financial managers have a

keen interest in how they are presented in financial statements and will exercise discretion to

obtain desired financial results.

Healy & Wahlen (1999) suggested that earnings management is a pervasive phenomenon

under the accrual basis accounting framework in GAAP, which allows management discretion in

the choice and application of accounting methods, estimates, and disclosures that influence

external reporting. They identified three major motivations managers have to manage reported

earnings: capital market motivations, contracting motivations, and regulatory motivations. (1)

Under capital market motivation, public companies feel pressure to report earnings that meet or

exceed analysts’ expectations. (2) Under the contracting motivation, accounting-based

New Revenue Standard ASC606

compensation and lending contracts induce managers to report higher earnings to increase

compensation and job security and mitigate potential violations of debt covenants by maintaining

some accounting ratios. (3) Under the regulatory motivation, managers have incentives to report

lower earnings to reduce or defer tax and avoid being a target of an anti-trust investigation. In

practice, whether managers exercise discretion to report higher or lower income depends on the

dominant factors underlying multiple motivations and goals.

In summary, the accounting choice literature and earnings management literature

identified common factors that drive managers’ accounting decisions to maximize firm value and

self-interest. Researchers also indicated that not all accounting choices involve opportunistic

earnings management (Fields et al., 2001; Healy & Wahlen, 1999; Watts & Zimmerman; 1990).

This study aims to identify the characteristics and motivations of early adopters in the

context of ASC 606 to complement existing ASC 606 studies and offer additional insights to

further our understanding of the new guidance. ASC 606 provides an interesting setting for early

adoption research in the accounting choice literature. This is because most previous studies

assumed that the new accounting standards had a positive effect on income or only selected firms

that had a positive adoption effect on the reported income in the sample (Gujarathi & Hoskin,

2003; Simon & Costigan, 1996). ASC 606 presents a more comprehensive setting to re-examine

the positive adoption choices literature as ASC 606 is likely to have a varying degree of effect on

firms’ financials. Unlike prior accounting changes, ASC 606 directly affects the top-line number

on the income statement. The informative nature and implications of revenue numbers to

investors make ASC 606 a unique standard to examine.

Prior literature on adoption timing choice in new accounting regulations documents that

early adopters consist of firms that have unique characteristics and tend to be strategic in making

New Revenue Standard ASC606

adoption decisions. The reporting flexibility allowed in transition choices provides firms

opportunities to communicate firm performance to investors in certain ways and influence

contractual outcomes predicted by positive accounting theory (Ali & Kumar, 1994; Ayres, 1986;

Gujarathi & Hoskin, 2003; Langer & Lev, 1993; Sami & Welsh, 1992; Simon & Costigan, 1996;

Trombley, 1989; Zmijewski & Hagerman, 1981).

However, the evidence on firm characteristics of early adopters is mixed in the extant

literature. For example, Zmijewski & Hagerman (1981) found that firms’ income strategies are

associated with factors including firm size, industry concentration, capital structure, systematic

risk, and debt constraints. Ayres (1986) found that early adopters of SFAS 52 – Accounting for

Foreign Currency Translation were smaller firms with lower earnings before adoption, more

constraints on dividend payouts, and less insider ownership. Trombley (1989) found that early

adopters of SFAS 86 – Accounting for Costs of Computer Software adoption decisions were

associated with auditors’ preference for the treatment of software development costs. Sami and

Welsh (1992) found that early adopters of SFAS 87 – Accounting for Pensions were large firms

with more fully funded pension obligations. Simon and Costigan (1996) found that early

adopters of SFAS 96 – Accounting for Income Taxes were associated with declining earnings

and larger deferred tax liabilities before the adoption.

ASC 606, provides a potential opportunity for earnings management in the year of

transition as it is likely to affect the timing and measurement of revenue recognition, but the

directions and magnitude of changes are not clear. Accordingly, this study examines the adoption

effects of this standard and tests if there is a systematic difference between early and non-early

adopters in firm characteristics to explain their voluntary adoption behaviors.

New Revenue Standard ASC606

2.7. Hypotheses Development

Managers evaluate their choices in debt management to maximize the value of their

stakeholders. Creditors are concerned about firms’ abilities to repay the loans when they come

due and usually seek the protection of their interest by requiring certain restrictive rules in debt

agreements to limit wealth transfer between debt and equity stakeholders. For example, they may

require borrowers to maintain certain financial leverage ratios or generate sufficient non-

restrictive earnings before distributing dividends. If these restrictive covenants are not met, firms

face negative consequences of technical default. The creditors may request immediate repayment

or firms may incur additional costs to renegotiate agreements with lenders. In addition, lenders

may increase interest rates on loans and notes following a violation or impose additional control

of firms' assets or activities by adding new covenant restrictions. Researchers documented that

the average costs of a technical violation of accounting-based debt covenants range between 1.2

percent and 2 percent of stockholders’ equity (Beneish & Eric, 1993). Thus, managers could

have incentives to increase reported earnings to ease constraints on debt covenants. Managers

could also have incentives to maintain an optimal level of debt to take advantage of positive

financial leverage (i.e., return on assets exceeds after-tax interest costs) to increase the rate of

return on equity for the benefit of stockholders (Garrison et al., 2021).

Positive accounting theory predicts that managers are more likely to increase reported

earnings when a firm approaches its debt covenant limits to reduce the likelihood of technical

violation and its associated costs. If a firm is far away from its covenant limits, the incentive to

adopt an income-increasing method would be less than a firm that is closer to the limit. In other

words, the probability of choosing an income-increasing accounting strategy is positively related

New Revenue Standard ASC606

to firms’ closeness to debt covenant constraints, and the effect is more pronounced when the

effects on earnings and equity are high (Ali & Kumar, 1994).

As discussed in Chapter 2, the new standard affects the timing and measurement of

revenue recognition. The changes due to the adoption of the new standard may affect covenant

slack – the difference between the required threshold value and the actual value of the covenant

measure. If these covenant slacks are one of the factors managers consider in their decision to

adopt ASC 606 early, then the closer firms are to the covenant threshold, the more likely firms

would take advantage of the opportunities to reduce the pressures imposed by debt covenants.

Research also suggests that firms are more likely to choose income-increasing strategies rather

than reduce dividend payments when they are closer to dividend restrictions. This is because a

missed dividend may send a warning signal to investors who expect to receive dividend

payments periodically.

Therefore, firms that are close to violating debt covenants are more likely to adopt ASC

606 early, if the adoption effects on reported financials are positive and the magnitude of the

adoption effects helps firms reduce the likelihood of debt covenant violations. If the adoption

effect is negative, firms are less likely to adopt it early if they approach the covenant threshold

because the adverse adoption effect would increase their chance of covenant violations. Rather

than adopting the standard early, they would wait for the mandatory adoption to have more time

to renegotiate contracts with third parties. This suggests that the likelihood of adopting the

standard early is decreasing in the magnitude of a negative adoption effect. Taken together, the

likelihood of adopting ASC 606 early increases with the adoption effects of ASC 606 as firms

are getting closer to their covenant threshold. This gives rise to my first hypothesis:

New Revenue Standard ASC606

H1: There is a positive relationship between the adoption effect of ASC 606 and the

propensity of firms that approach their debt constraints to adopt the standard early.

Prior research has identified that the consideration of stock prices may influence

managers’ accounting choices. A consistent trend of revenue growth is usually associated with

higher stock prices (Healy & Wahlen, 1999). The former chairman of the SEC, Arthur Levitt

said that in a market where missing an earnings projection by a penny could cause firms to lose

millions of dollars in their market value, it is not surprising that managers are motivated to game

the system to project a smooth earnings path and avoid negative stock price reactions. The

former chairman also criticized firms for using opportunistic revenue practices to influence

contractual outcomes that erode the quality of financial reporting, such as premature revenue

recognition, “cookie jar reserves” (banking income for the future in good years), “big bath”

charges (flushing more costs in bad years so that earnings can be inflated in the future), or

abusive practices of materiality (intentionally misstate earnings within a self-defined materiality

threshold) (Levitt, 1998). Healy (1985) suggested that managers are more likely to inflate

earnings when earnings are close to market expectations and deflate earnings when earnings are

way above or below market expectations.

Positive accounting theory predicts that self-interested managers have incentives to

exercise their discretion over accounting numbers to maximize firm value and compensation

(Watts & Zimmerman, 1990). Management compensation plans usually include base salaries,

performance-based bonuses, stock options, and deferred compensations that are tied closely to

firms’ stock performance in the market. These accounting-based and stock market-based

incentive plans are structured as a part of solutions for agency problems to align the interests of

principals (shareholders) with agents (managers). Managers of firms with incentive

New Revenue Standard ASC606

compensation plans are more likely to use accounting alternatives to increase reported earnings

to present a better picture of their financials if the pre-adoption financial performance is below

the market expectations. Studies also have suggested that managers may seek a consistent trend

of revenue growth and avoid reporting income below the previous year’s level even absent an

income-based incentive compensation plan (Trombley, 1989). This is because the trend of

revenue directly affects stock prices and the perception of management performance.

The adoption of ASC 606 could affect firms’ reported revenue and thus earnings,

although the effect is likely to vary across firms and industries. The adoption timing choice

provides firms with a one-time earnings management opportunity to influence reported revenue

numbers and earnings. If a firm has a declining financial performance before adoption and the

adoption results in a positive effect on revenue and earnings, it is more likely for the firm to

adopt the standard early if the favorable impact helps the firm get closer to or exceed the targeted

return. Firms with a declining pre-adoption performance are less likely to adopt the standard

early if the adoption results in a negative impact on earnings because the unfavorable adoption

effect would make their financials appear even worse, holding everything else constant. In

contrast, firms with improving pre-adoption performance are less likely to adopt the standard

because a favorable adoption effect is less needed to help them meet the earnings target while an

unfavorable adoption effect may make them miss the earnings target if the size of changes is

large enough. In sum, firms with improving pre-adoption performance are less likely to adopt the

standard early considering the benefit of having more time to comply with such a complex and

comprehensive standard and increased attention and scrutiny from the public and regulators.

Therefore, the likelihood of adopting ASC 606 early increases with the adoption effects as firms’

pre-adoption financial performance deteriorates. This discussion leads to my second hypothesis:

New Revenue Standard ASC606

H2: There is a negative relationship between the adoption effect of ASC 606 and the

propensity of firms with an increasing pre-adoption financial performance to adopt the standard

early.

New Revenue Standard ASC606

Chapter 3: Research Method

3.1 Sample Selection

In May 2014, the FASB issued ASU 2014-09, “Revenue from Contracts with

Customers”, which supersedes the revenue recognition requirements in Revenue Recognition

under ASC 605. The FASB subsequently issued several amendments to the initial guidance

(collectively, ASC 606) in response to shareholders’ concerns about the practical difficulties and

challenges of transition to the new standard. Among these amendments, ASU2015-14 defers the

mandatory effective date of the new standard for public entities from annual reporting periods

beginning after December 15, 2016, to annual reporting periods beginning after December 15,

2017, but allows companies to adopt it a year early. This means that a firm is an early adopter of

ASC 606 if it adopted the standard with annual reporting periods beginning between December

16, 2016, and December 15, 2017. Otherwise, a non-early adopter if it adopted the standard

outside this window frame.

Table 4 summarizes the sample collection process of early adopters and non-early

adopting peers of ASC 606 in this study. I began the sample selection by choosing the Data Year

– Fiscal variable with a date range from 01/2014 to 12/2020 and searching the entire Compustate

North America Fundamental Annual database and all variable types. Then, I required the data

with the Foreign Incorporate Code (FIC) variable - USA and Currency Code variable – USD.

This resulted in a total of 55,824 firm years. The item “FIC” identifies the country in which the

company is incorporated or legally registered. This study is interested in US firms that adopted

ASC606, thus all foreign-incorporated firms were removed from the sample even though some

of them adopted ASC 606 or IFRS 15 and trade stocks in the US market.

New Revenue Standard ASC606

Table 4. Sample Selection

Data Selection Process

Data Output

(1) In Compustat, select Data Year - Fiscal: 01/2014-12/2020,

and require FIC = ‘USA’ and Currency = “USD”

55,824 firm-years

(2) Remove firms with SIC codes between 6000 and 6999

(Division H: Finance, Insurance, and Real Estate)

30,248 firm-years

(3) Filter data and identify firms reported as materially affected

by ASC606 by requiring ACCTCHG = ‘ASU14-09’

1,807 firms

(4) Identify firms that adopted ASC606 with accounting periods

beginning between 12/16/2016 and 12/15/2017 (i.e., ending

between 12/15/2017 and 12/14/2018 in Compustat)

50 early adopters and

1,757 non-early adopters

(5) Identify early adopters that have required data available and

disclosed the cumulative adoption effects in footnotes

20 firms

(6) Select the non-early adopting peers that have required data

available by matching the four-digit SIC code and closest total

assets with early adopters identified in step (5)

20 firms

Next, I removed firms with the SIC codes between 6000 and 6999 (Division H: Finance,

Insurance, and Real Estate) using the Standard Industrial Classification (SIC) Manual provided

by the Department of Labor – Occupational Safety and Health Administration

. Firms in this

regulated industry were excluded because they have unique industry characteristics and their

incentives to adopt early may be different from other industries. This step resulted in 30,248 firm

years.

I then followed the prior literature on adoption timing choices to focus on early adoption

samples that were materially affected by the new standard (Ali & Kumar, 1994; Simon &

Costigan, 1996). To identify materially affected firms, I relied on the accounting change variable

https://www.osha.gov/data/sic-manual

New Revenue Standard ASC606

(ACCTCHG) equal to “ASU2014-09” in Compustat. According to Compustat’s variable

explanations, the item “ACCTCHG” captures the effective date of accounting changes that “have

a substantive impact on the measurement and presentation of financial data, or which require

significant new disclosures”. To understand how the “substantive impact” is captured in this

variable, I communicated with the S&P Global Client Support team and they confirmed that the

variable ACCTCHG was set to “ASU2014-09” in the adoption year based on firms’ disclosures

of quantitative adoption effects and the corresponding accounting change footnotes. In other

words, if the adoption of ASC 606 had no or immaterial effect on a firm’s financial statements

including its presentation, or a firm did not disclose the adoption effects in footnotes, this

accounting change variable in Compustat was not set to “ASU2014-09”. For example, Apple,

Inc. disclosed that it adopted ASC 606 in the first quarter of 2019 using the FRM method and the

adoption did not have a material impact on its financial statements (2019 Annual Report, Note

1), the variable ACCTCHG was not set to “ASU2014-09” in the adoption year or any other firm-

years for Apple Inc. in Compustat. Thus, I identified firms that were materially affected by

ASU2014-09 based on the ACCTCHG variable in Compustat. This step resulted in a sample of

1,807 US firms.

I further separated the sample into two groups: early adopters and non-early adopters.

Specifically, I sorted the 1,807 firms by FY end date (“datadate” in Compustat) and identified an

initial sample of 50 firms that met the early adoption requirements - firms that adopted ASC 606

with a fiscal year beginning between December 16, 2016, and December 15, 2017

. In this

Compustat provides fiscal year ending data instead of the fiscal year beginning data, thus the criteria for

identifying early or non-early adopters are based on firms’ fiscal year ending dates (i.e., ending between December

15, 2017, and December 14, 2018) in the sample selection process.

New Revenue Standard ASC606

sample of 1,807 materially affected firms, 50 firms are early adopters and the rest of the 1,757

firms are non-early adopters.

Table 5A below provides the industry distribution of the 1,807 firms identified above

based on the two-digit SIC codes.

Table 5A. Industry Distribution of Early Adopters vs Non-Early Adopters

SIC*

Industry

Early

Adopters**

Freq.

Non-Early

Adopters**

Freq.

Total

Oil and Gas Extraction

Food and Kindred Products

Chemicals and Allied Products

16%

240

14%

248

Industrial and Commercial Mach.

and Computer Equip

Electronic and Other Electrical

Equipment

127

131

Transportation Equipment

Measuring, Analyzing, and

Controlling Instruments

12%

102

108

Sanitary Services

112

113

Wholesale Trade - Durable Goods

Apparel and Accessory Stores

Miscellaneous Retail

Computer Programming, Data

Processing, Prepackaged Software

40%

343

20%

363

Other Group***

515

29%

515

Total

100%

1757

100%

1807

* Standard Industry Classification (SIC): https://www.osha.gov/data/sic-manual

** Identified by "ASU2014-09" in Compustat

*** The rest of the 515 non-early adopters are combined into the “other group” (see appendix A for details)

The results indicate that the 50 early adopters are distributed across 12 industries and

40% of them are concentrated in the software industry, followed by the chemicals and allied

products industry (16%), then the measuring, analyzing, and controlling instruments industry

(12%). As a comparison, these 12 industries represent about 71% of the industry membership in

non-early adopters. The rest of the 29% of the industry membership in the non-early adopters'

New Revenue Standard ASC606

group was combined and reported in the “other group”

. The industry distribution of the non-

early adopters shows that 20% of them are in the software industry, 14% in the chemicals and

allied products industry, 7% in the electronic and other electrical equipment, and 6% in the

measuring, analyzing, and controlling instruments industry. Overall, the industry membership of

early adopters is similar to that of non-early adopters except for the software industry (SIC: 73),

which has a relatively higher percentage of firms choosing to be early adopters (40% of early

adopters vs 20% of non-early adopters).

Table 5B presents the mean values of key variables of interest comparing early adopters

with non-early adopters of ASC 606 in the year of adoption (year t in this study)

Table 5B. Preliminary Statistics of Early vs Non-early Adopters of ASC 606

# of

Firms

LEV

ROE

ROA

Total

Assets

Total

Liab.

Net

Sales

EBIT

Early

1.012

(2.450)

(0.315)

16,206

10,116

9,060

1,563

804

Non-Early

1752

0.726

(0.840)

(0.212)

7,278

4,949

4,575

535

285

All Adopters

1802

0.734

(0.885)

(0.215)

7,526

5,092

4,699

563

299

Note: All numbers are in millions, except N, LEV, ROE, and ROA

ASC606 adopters are identified by Compustat Variable Descriptor: ASU2014-09

LEV = Total Liabilities/Total Assets

ROE = Pretax Income/Absolute Value of Total Equity

ROA= Pretax Income/Total Assets

After removing five non-early adopters that have missing values in total assets and total

liabilities in Compustat, the number of non-early adopters dropped from 1,757 firms in Table 5A

to 1,752 firms in Table 5B. The results indicate that the average size of early adopters (50 firms)

is larger than that of non-early adopters (1,752 firms) in terms of total assets, total net sales,

In the other group, no single industry’s membership frequency exceeds 2.7% based on the first two digits of the

SIC code.

Year t instead of year t-1 statistics are provided due to data limitation of early adopters in year t-1.

New Revenue Standard ASC606

earnings before interests and taxes (EBIT), and net income (NI). The average leverage ratio of

early adopters is higher than that of non-early adopters, while the profitability ratios (ROA and

ROE) of early adopters are lower than those of non-early adopters. These preliminary descriptive

statistics are consistent with the expectations that early adopters are more likely to have higher

debt constraints and lower financial performance before adoption.

To collect all the variables for my empirical model, I checked the data availability of the

50 early adopters in the adoption year (year t), and the preadoption years (t-1 and t-2). I found that

13 firms missed financial information in preadoption years t-1 and t-2, and 8 firms missed financial

information in year t-2. This is mainly because these firms issued and completed initial public

offers (IPOs) after adopting ASC606, therefore their preadoption years’ financial statements are

not available to the public. This also means that their preadoption cumulative RE effects are not

available for the empirical model. In addition, one of the early adopters had a fresh-start accounting

upon emergence from bankruptcy in the adoption year and eight firms did not provide adequate

information to determine the cumulative adoption effects which is a variable of interest in the

empirical model

. Thus, these 30 firms were removed from the initial 50 early adopters, and the

final sample consists of 20 early adopters that present all the data required for the empirical model.

After identifying the sample of early adopters, I selected a control sample of 20 non-early

adopting peers matched by industry and size with early adopters in year t-1. Specifically, I first

removed firms that have missing values under the required variables (total assets, total liabilities,

retained earnings, and pre-tax income) from the 1,757 non-early adopters. I then sorted the data

by the four-digit Standard Industry Classification Code (SIC) and the closet total assets to select

Five firms incurred significant operating losses since inception.

New Revenue Standard ASC606

the control group of 20 non-early adopting peers. The purpose of matching is to control for

industry and size differences between early adopters and non-early adopters.

Finally, I manually collected the cumulative adoption effects of sample firms from Item 8

Financial Statements and Supplementary Data on Form 10-K from the SEC’s Edgar Database.

3.2 Empirical Model and Measurement

I tested H1 and H2 by estimating the following binary logistic regression model:

(EarlyAdopt

=1) = α + β

SIZE

t-1

+ β

LEV

t-1

+ β

GROWTH

t-1

+ β

ROE

t-1

+ β

ΔRE

t-1

LEV

t-1

* ΔRE

t-1

+ β

Growth

t-1

* ΔRE

t-1

+ ɛ

Table 6. Variable Definitions

Variable

Variable Definitions

Compustat

Sources

EarlyAdopt

An indicator variable that equals one if a firm

is an early adopter of ASC 606 in year t and

zero otherwise.

ACCTCHG:

ASU2014-09

A&K (1994);

G&H (2003)

SIZE

t-1

A predictor variable that measures firms' size

by total sales (= natural log of Total Sales

t-1

)

Ln (SALE

t-1

)

Z&H (1981);

S&W (1992)

LEV

t-1

A predictor variable that measures firms'

constraints on debt covenants (= Total Debt

t-1

Total Assets

t-1

)

t-1

/AT

t-1

Z&H (1981);

B&P (1993);

G&H (2003)

GROWTH

t-1

A predictor variable that measures the rate of

growth of pretax earnings (= Pretax Income

t-1

Pretax Income

t-2

) /Absolute value of Pretax

Income

t-2

)

(PI

t-1

- PI

t-2

) /|PI

t-2

G&H (2003);

S&W (1992)

ROE

t-1

A predictor variable that measures firms'

profitability in year t-1 (= Pretax Income

t-1

Absolute Value of Total Equity

t-1

)

t-1

/|TEQ

t-1

A&K(1994);

G&H (2003);

S&W (1992)

ΔRE

t-1

A predictor variable that measures firms'

cumulative adoption effect [(= RE Effect

t-1

Total Assets

t-1

) *100]

Hand-collected

from 10Ks / AT

t-1

G&H (2003)

All independent variables are specified in year t-1 because the adoption decision is made in year t-1.

New Revenue Standard ASC606

H1 states that there is a positive relationship between the adoption effect of ASC 606 and

the propensity of firms approaching debt constraints in adopting the standard early. Specifically,

I expect the probability of firms being early adopters to be increasing with the interactive effect

of leverage (LEV

t-1

) and the adoption effects (ΔRE

t-1

). Leverage (LEV

t-1

) is measured as Total

Debt

t-1

divided by Total Assets

t-1

, which captures firms’ closeness to debt covenant constraints

(i.e., LEV

t-1

increases as firms approach their debt constraints). The adoption effect of ASC 606

(ΔRE

t-1

) is measured as the cumulative adoption effects on opening retained earnings divided by

the beginning total assets in percentage in the year of adoption. A larger ΔRE

t-1

indicates a

greater adoption effect of ASC 606 on firms’ financial statements. Thus, a finding of a positive

on LEV

t-1

* ΔRE

t-1

provides support for H1.

In prior literature, the most frequently used measures of debt constraints when modeling

early adoption decisions of a new accounting standard are debt-to-equity or debt-to-assets ratios

and dividend payout ratios. However, dividend payout ratios have limited data availability in the

sample. The majority of the sample firms did not pay out dividends and only 14 firms (35% of

40 firms) paid cash dividends in year t-1. More specifically, only 5 early adopters and 9 non-

early adopters paid cash dividends. The 14 dividend-paying firms tend to be larger and more

profitable than those that did not pay cash dividends

. This is consistent with the findings from

Fama and French (2001), who showed that publicly traded firms have become less likely to pay

dividends as the perceived benefits of dividends have declined over time, and firms are more

inclined to reinvest in themselves with increased growth opportunities or prefer stock

The 14 firms that paid cash dividends have average total assets of $74,489 million, average net income of $2,967

million, and average retained earnings of $11,301 million, while those that did not pay dividends have average total

assets of $7,597 million, average net income of $693 million, and average retained earnings of $3,900 million.

New Revenue Standard ASC606

repurchases which may result in high returns through increases in stock prices. This phenomenon

raises the concern of whether dividend-paying ratios capture the underlying concept of debt

constraints in the sample. Considering both the data limitation in the sample and the documented

reduction in dividend-paying firms in the literature, I chose to use the debt-to-assets ratio (i.e.,

LEV

t-1

) as the proxy for debt constraints in the empirical model.

The inclusion of the interaction term (LEV

t-1

* ΔRE

t-1

) in the empirical model also

followed the research design discussion in the accounting choice literature. Prior studies usually

included firm characteristics as baseline variables for inferences of economic differences in

contracting costs and political costs that influence firms’ accounting choices. However, Ali and

Kumar (1994) argued that many of these studies either left out the variables that capture the

economic consequence of financial statement effects in accounting decisions or only included

them as separate independent variables along with measures of firm-specific variables. They

recommended the inclusion of the interactive effects of firm characteristics and financial

statement effects of the accounting choice in the model to enhance the power of explanations and

mitigate concerns about omitted variables. Gujarathi and Hoskin (2003) modeled the early

adoption decision in SFAS 96 – Accounting for Income Taxes and showed that the interactive

effects model has higher explanatory power in terms of adjusted R

when comparing models

with and without interactive terms.

The adoption effect of ASC 606 (ΔRE

t-1

) is regarded as a balance sheet effect in the

accounting choice literature. Gujarathi and Hoskin (2003) suggested that the balance sheet effect

is more important than the income effect in accounting changes as the balance sheet effect

captures the overall change to equity due to the adoption of a new standard, while the income

effect only captures the current year adoption effect on the income statement. In addition, the

New Revenue Standard ASC606

focus on the balance sheet effect is because the adoption year income effect is not available for

firms that used the full retrospective method. Under the FRM method, revenues on comparative

financial statements were restated or reported under ASC 606, but the revenue under ASC 605 in

the adoption year is not available, which makes it infeasible to calculate the adoption year

income effect (i.e., the difference between ASC 606 revenue and ASC 605 revenue). Regardless

of the transition method, firms disclosed cumulative adoption effects that affected the opening

retained earnings in the year of transition (i.e., adoption year balance sheet effect). Therefore, I

used the cumulative adoption effects on retained earnings to capture the adoption effect (ΔRE

t-1

)

in this study.

H2 states that there is a negative relationship between the adoption effect of ASC 606 and

the propensity of firms with an increasing pre-adoption financial performance to adopt the

standard early. Firms’ preadoption financial performance is captured by the variable Growth

t-1

which is the change of pretax income from year t-2 to year t-1 divided by the absolute value of

pretax income in year t-2. To test H2, my interest lies in the sign of the coefficient of β

on the

interactive term

Growth

t-1

* ΔRE

t-1

. I expect the probability of firms being early adopters to be

decreasing with the interactive effect of firms’ adoption effect and increasing preadoption

performance. Thus, a finding of a negative β

on Growth

t-1

* ΔRE

t-1

provides support for H2.

I followed the prior literature that used the rate of growth (Growth

t-1

) and profitability

(ROE

t-1

) to capture the management compensation and stock market incentives (Gujarathi &

Hoskin, 2003; Sami & Welsh, 1992). The two variables (Growth

t-1

and ROE

t-1

) provide different

measures of financial performance. Specifically, ROE is a measure of a firm’s profitability and

efficiency of operations in a specific period. It is calculated using pre-tax income at year t-1

divided by the absolute value of total equity at year t-1. Pretax income rather than net income is

New Revenue Standard ASC606

used in the calculation of ROE to avoid potential tax effects on earnings from the 2017 Tax Cuts

and Jobs Act. In practice, ROE is often considered an important metric for analyzing stocks.

Whether a firm’s ROE is high or low depends on what is normal in the industry and what it is

relative to its competitors. In comparison, growth rates capture a trend and measure the

percentage change in income over multiple periods. Two firms with the same ROE may have

very different growth rates. Studies show that earnings growth is considered one of the most

important elements in stock selections. A stable and consistent trend of earnings growth is

usually associated with higher stock prices (Gujarathi & Hoskin, 2003)

In addition, firm size is included as a control variable in the model, and no predictions are

made regarding the direction of the association between the firm size characteristics and early

adoption. Positive accounting theory suggests that the effect of firm size on accounting choices is

often associated with political costs. Large firms have higher political costs than small firms

because they receive more public attention and are more likely to be the target of potential anti-

trust lawsuits as the public tends to associate high income with monopoly rents (Watts &

Zimmerman, 1990). Small firms have lower political costs because they are followed by fewer

investors and analysts and tend to have more private debt with banks (e.g., notes payable) than

public debt with investors (e.g., bonds payable). Thus, large firms have incentives to defer

compliance with a mandatory accounting standard that has a favorable and larger effect on their

earnings if they are meeting analysts’ expectations. However, the new standard imposes new

information production costs on firms such as system upgrading costs, bookkeeping costs,

training costs, and consulting costs. Ceteris paribus, large firms are more likely to be early

adopters because they have more human and economic resources and abilities to implement the

changes, and the information costs related to adoption would be much lower for large firms than

New Revenue Standard ASC606

small firms relative to their sizes. Therefore, political cost theory and information production

cost theory give an opposite prediction regarding the direction of the association between the

firm size characteristics and early adoption. Most prior research documented that firm size is an

important factor in adoption decisions but provided mixed findings (Simon & Costigan, 1996).

Some researchers suggested that firm size can be a proxy variable for factors other than political

costs and information production costs (Sami & Welsh, 1992). Thus, size is included in the

model as a control variable. In the literature, the firm size is usually measured by total assets or

total sales. I chose to use total sales to reduce correlations among independent variables because

both LEV

t-1

and ΔRE

t-1

variables are scaled by total assets in the model.

Prior literature documented that LEV

t-1

, Growth

t-1

, and ROE

t-1

, are important factors that

are associated with firms’ adoption decisions, therefore, they are included as control variables in

the model. I expect that early adopters of ASC 606 have higher preadoption financial leverage

and lower preadoption profitability and growth rates than non-early adopting peers. ΔRE

t-1

also included as a control variable in the model and I expect that early adopters of ASC 606 are

more likely to have larger and more favorable adoption effects than non-early adopting peers.

Since the dependent variable of my model is a dichotomous categorical variable that has

two possible outcomes (1= early adopter and 0 = non-early adopter) and independent variables

are measured at scale or ratio level, I used binary logistic regression to predict the probability of

the target group (early adopters) relative to the control group (non-early adopters) based on the

value of the predictors. Least Square regression is not appropriate in this study because it

assumes a constant relationship between independent variables (X) and dependent variables (Y)

where X and Y would extend infinitely to both directions (linear function). However, since the

dependent variable in my model (the probabilities of events occurring) is bound by 0 or 1, the

New Revenue Standard ASC606

logistic regression model (non-linear function) provides a better line of fit as it recognizes a

different predicting power over the entire range of predictors (S-shaped curve). According to

Meyers, Gamst & Guarino (2017), logistic regression uses maximum likelihood estimation to

obtain predicted probabilities and makes fewer assumptions than linear regression. It does not

require normal distribution of predictor variables and homoscedasticity of variances as linear

regression does when making inferences about the coefficients. Thus, a binary logistic regression

was employed for the empirical model in this study.

New Revenue Standard ASC606

Chapter 4: Results

This study investigated the motivations behind firms’ early adoption decisions of ASC

606 and addressed two research questions: (1) What is the adoption effect of ASC 606 on the

financial statements of early adopters versus non-early adopting peers? (2) Is there evidence of

systematic differences in the characteristics of firms that were early adopters versus non-early

adopting peers?

4.1 Descriptive Statistics

I first examined the industry pattern of early adopters using the two-digit SIC codes and

presented the results in Table 7A.

Table 7A. Industry Distribution of Early Adopters (n = 20)

SIC

Industry

# of

Firms

Frequencies

Chemicals and Allied Products

Industrial and Commercial Machinery and Computer Equipment

10%

Electronic and Other Electrical Equipment and Components

20%

Transportation Equipment

10%

Measuring, Analyzing, and Controlling Instruments

15%

Sanitary Services

Wholesale Trade - Durable Goods

Computer Programming, Data Processing, and Prepackaged Software

30%

Total

100%

The results show that the early adopters in the final sample are distributed across eight

industries and about 30% of them are in the computer programing and software industry (SIC:

73), followed by 20% in the electronic and other electrical equipment industry (SIC: 36), and

15% in the measuring, analyzing, and controlling instruments industry (SIC: 38). This industry

Due to inadequate data in Compustat and the 10-Ks, the final sample includes 20 early adopters and 20 non-early

adopting peers matched by industry and size. Thus, non-early adopters have the same industry pattern as the early

adopters in the final sample.

Standard Industry Classification (SIC): https://www.osha.gov/data/sic-manual

New Revenue Standard ASC606

distribution is similar to the industry distribution in the initial 50 early adopters (see Table 5A).

More specifically, the 30% early adopters in the final sample are consistent with the 40% early

adopters in the initial sample of 50 early adopters (see Table 5A). This higher proportion of early

adopters in the software industry may be related to their specific industry characteristics and

business models. As discussed earlier, software firms often have multiple deliverables and ASC

606 breaks down performance obligations into more granular units, resulting in certain types of

income being shifted to earlier periods. For example, Microsoft (SIC: 7372) disclosed in its

annual report (6/30/2018) that it elected to early adopt the standard effective July 1, 2017, using

the full retrospective method. The new revenue standard affects the company’s accounting for

software license revenue (e.g., Windows 10 and certain multi-year commercial software

subscriptions). Under ASC 606, they recognize license revenue at the time of delivery rather than

ratably over the subscription period under ASC 605. This finding is consistent with the

expectation of industry experts that the software firms are probably more affected by the new

standard than other industries and consistent with their lobbying efforts for the change of revenue

standards.

The results also show that 20% of early adopters are in the electronic and other electrical

equipment industry (SIC: 36). This phenomenon may be related to the accounting changes for

their distributor revenue in this industry. For instance, Power Integrations Inc. (SIC: 3674)

disclosed in its annual report (12/31/2017) that it elected to early adopt ASC 606 effective

January 1, 2017, using the full retrospective method. ASC 606 changes its accounting for

distributor and reseller revenue, where the firm recognizes revenue upon shipment and transfer

of control to distributors (“sell-in” model) in ASC 606, rather than deferring revenue recognition

until goods are sold to end customers (“sell-through” model) in ASC 605.

New Revenue Standard ASC606

Then, I examined the average adoption effects of ASC 606 by industry and presented the

cumulative adoption effects in dollars (REEFFECT) and percentage changes (ΔRE) in Table

Table 7B. Average Adoption Effects of ASC 606 by Industry (n = 40)

SIC

Industry

# of Firms

REEFFECT

(in millions)*

ΔRE (in %)**

Chemicals and Allied Products

$21.76

2.58

Industrial and Commercial Machinery and

Computer Equipment

$6.23

26.67

Electronic and Other Electrical Equipment and

Components

$10.26

2.44

Transportation Equipment

-$170.75

-0.32

Measuring, Analyzing, and Controlling

Instruments

$32.40

-3.23

Sanitary Services

$0.35

0.72

Wholesale Trade - Durable Goods

$4.06

0.48

Computer Programming, Data Processing, and

Prepackaged Software

$1,346.30

0.42

Total

$395.66

2.95

*REEFFECT = Cumulative Adoption Effect of ASC 606 as of the Date of Adoption ( in millions)

** ΔRE = (Cumulative RE effect

t-1

/ Total Assets

t-1

)

*100

The results indicate that the new standard affected industries differently in both

magnitude and direction. For example, the computer programming and software firms (SIC: 73)

reported an average increase of retained earnings by $1,346.30 million (0.42% increase of RE),

while the transportation equipment firms (SIC: 37) reported an average decrease of retained

earnings by $170.75 million (0.32% decrease in RE) due to the adoption of ASC606 in the year

of transition. Although the average dollar amount adoption effect for the industrial and

The non-early adopting peers are selected based on the closest size in the same industry to early adopters.

New Revenue Standard ASC606

commercial machinery firms (SIC: 35) is only $6.23 million, they reported the highest adoption

impact in terms of a percentage change in RE (26.67%). This is mainly because these firms are

relatively small (average total assets: $260.36 million) compared with the entire sample firms

(average total assets: $31,009.64 million). Of the sample firms, 33 out of 40 firms (83%)

experienced a positive change to stockholders’ equity due to the application of ASC 606 in the

year of transition.

Next, I examined the descriptive statistics of the variables of interest in the study and

presented the results in Table 8A.

Table 8A. Descriptive Statistics (n = 40)

Variable

Mean

Std Dev

25%Q

Med.

75%Q

REEFFECT

t-1

395.66

2411.68

0.66

5.59

34.31

SIZE

t-1

6.68

3.24

4.86

6.58

8.84

LEV

t-1

0.60

0.41

0.26

0.61

0.80

GROWTH

t-1

- 0.16

1.15

-0.61

-0.05

.21

ROE

t-1

-0.39

6.80

-0.14

0.17

0.27

ΔRE

t-1

2.95

17.62

0.08

0.72

3.00

Variable Definitions:

REEFFECT

t-1

= Cumulative Adoption Effect of ASC 606 as of the Date of Adoption ( in million dollars)

SIZE

t-1

= Natural log of Total Sales

t-1

(t = Adoption Year)

LEV

t-1

= Total Debt

t-1

/Total Assets

t-1

GROWTH

t-1

= (Pretax Income

t-1

- Pretax Income

t-2

) /Absolute value of Pretax Income

t-2

ROE

t-1

= Pretax Income

t-1

/ Absolute value of Total Equity

t-1

ΔRE

t-1

= (Cumulative RE effect

t-1

/ Total Assets

t-1

)

*100

The results show that the average dollar amount adoption effect (REEFFECT

t-1

) of ASC

606 on firms’ financials is $395.66 million and the median value is $5.59 million. The range of

the adoption effects is widely spread in the sample. This is because Microsoft (early-adopter)

reported a $15,121 million increase in retained earnings and General Motors (non-early adopter)

New Revenue Standard ASC606

reported a $1,336 million decrease in retained earnings as of the date of transition due to the

adoption of the new standard

To adjust for the effect of firm size on the adoption effects, I scaled the cumulative

adoption effects (REEFFECT

t-1

) by total assets to obtain the variable ΔRE

t-1

. The results in

Table 8A show that sample firms on average experienced a 2.95% increase in retained earnings

due to the adoption of the new standard in the year of transition.

After that, I compared the adoption effects of ASC 606 between early and non-early

adopters and reported the summary statistics by group (early vs non-early group) in Table 8B.

Table 8B. Descriptive Statistics (20 Early Adopters vs 20 Non-Early Adopting Peers)

Variable

Sample

Mean

Std Dev

25%Q

Med.

75%Q

REEFFECT

t-1

Early

746.89

3387.38

0.45

15.94

41.55

Non-Early

44.42

452.18

0.89

4.29

28.60

SIZE

t-1

Early

6.28

3.86

4.00

6.56

9.56

Non-Early

7.08

2.50

5.51

6.58

8.84

LEV

t-1

Early

0.65

0.52

0.25

0.64

0.85

Non-Early

0.56

0.28

0.54

0.80

GROWTH

t-1

Early

-0.49

0.99

-1.08

-0.35

0.21

Non-Early

0.17

1.23

-0.47

0.04

0.29

ROE

t-1

Early

-2.23

7.63

-0.56

-0.02

0.22

Non-Early

1.45

5.43

-0.09

0.12

0.41

ΔRE

t-1

Early

4.90

25.06

0.03

1.76

3.84

Non-Early

1.00

1.23

0.13

0.57

1.49

Variable Definitions:

REEFFECT

t-1

= Cumulative Adoption Effect of ASC 606 as of the Date of Adoption ( in million dollars)

SIZE

t-1

= Natural log of Total Sales

t-1

(t = Adoption Year)

LEV

t-1

= Total Debt

t-1

/Total Assets

t-1

Per Microsoft (Annual Report, 6/30/2018), the adoption of the new standard resulted in an increase in revenue,

income taxes, accounts receivable, and other long-term assets and a decrease in unearned revenue due to the upfront

recognition of license revenue. For some complex commercial license contracts, the revenue recognition treatment

depends on the contract-specific terms. General Motors disclosed in its annual report (12/31/2018) that the new

standard affected their accounting for sales incentives and certain transfers to daily rental companies. Sales

incentives are recorded at the time of sale rather than at the later of sale or announcement, resulting in the shifting of

incentive spending to earlier periods. In addition, some transfers were previously accounted for as operating leases

under ASC 605 but are accounted for as sales under ASC 606 if ownership of vehicles is not expected to be

transferred back.

New Revenue Standard ASC606

GROWTH

t-1

= (Pretax Income

t-1

- Pretax Income

t-2

) /Absolute value of Pretax Income

t-2

ROE

t-1

= Pretax Income

t-1

/ Absolute value of Total Equity

t-1

ΔRE

t-1

= (Cumulative RE effect

t-1

/ Total Assets

t-1

)

*100

The results show that the average dollar amount adoption effects of early adopters

($746.89 million) are much larger than those of non-early adopting peers ($44.42 million). The

median value of adoption effects for early adopters ($15.94 million) is also larger than that of

non-early adopters ($4.29 million). This suggests that the adoption of ASC 606 produced a more

favorable financial effect on early adopters than on non-early adopting peers. The scaled

adoption effects (ΔRE

t-1

) indicate that early adopters reported an average of 4.90% increase in

retained earnings while non-early adopters reported an average of 1.00% increase in retained

earnings. The median value of the ΔRE

t-1

effect is consistent with the results of the mean values.

That is, a 1.76% increase of ΔRE

t-1

for early adopters is higher than a 0.57% increase of ΔRE

t-1

for non-early adopters.

In addition, early adopters appear to be more leveraged and have lower pre-adoption

performance than non-early adopting peers. As shown in Table 8B, the average debt-to-assets

ratio (LEV

t-1

) is 65% for early adopters and 56% for non-early adopters; the pre-adoption

profitability (ROE

t-1

) is -2.23 for early adopters and 1.45 for non-early adopters; and the pre-

adoption rate of growth is -49% for early adopters and 17% for non-early adopters. Furthermore,

60% of early adopters experienced a loss in pretax income (PI

t-1

< 0), while 35% of non-early

adopters had a loss in pretax income. These results are consistent with the expectation that firms

with higher debt constraints, lower preadoption financial performance, and larger adoption

effects are more likely to be early adopters as predicted in the hypotheses. This evidence

suggests that managers potentially made their adoption timing choices based on firm

characteristics and adoption effects on firms’ financials.

New Revenue Standard ASC606

Finally, I examined the percentage of sample firms in the choices of transition methods

and presented the results in Table 9.

Table 9. Frequencies of Transition Method Choice in Sample Firms (n=40)

Early Adopters

Non-Early Adopters

Overall

Transition Method

# of Firms

Full Retrospective

65%

15%

40%

Modified Retrospective

35%

85%

60%

Total

100%

The results show that 65% of early adopters chose the full retrospective method while

85% of non-early adopters chose the modified retrospective method. This is interesting

considering that the full retrospective method is far more challenging and resource intensive to

implement. It might indicate that these early adopters have relatively lower information

production costs and value the benefit of the availability of historical trend information via

restated comparative financial statements. Another reason that might contribute to this finding is

that some early adopters were planning to issue IPOs. If they use the modified respective

method, revenue reported would appear to be lower in the transition period if ASC 606

accelerates the timing of revenue recognition as the accelerated revenue would bypass the

income statement and be recorded as a cumulative adjustment to retained earnings at the date of

adoption (see Table 2A). This issue of accelerated revenue not being shown on the income

statement may be a consideration for firms that were planning to go public since investors use

revenue, net income, and revenue growth rate when analyzing firms’ profitability and prospects.

The full retrospective method may help mitigate the perception that a firm’s revenue growth is

decreasing, depending on various factors such as the date to issue IPOs, the number of

New Revenue Standard ASC606

outstanding contracts (contracts that started before adoption), or new contacts (contracts that

started during the adoption year), etc.

In summary, descriptive statistics suggest that most early adopters chose the full

retrospective method while the majority of non-early adopters chose the modified retrospective

method, and on average, the cumulative adoption effects of early adopters are higher than those

of the non-early adopting peers in both dollar amounts and percentage changes. In the next

section, I tested the two hypotheses by estimating the empirical model as proposed earlier.

4.2 Tests of Hypotheses

To answer the second research question whether there is evidence of systematic

differences in the characteristics of firms that were early adopters versus non-early adopting

peers, I first examined the statistical significance between the group means of the variables of

interest using the independent samples t-test, and the results were presented in Table 10 below.

Table 10. Independent Samples T-Test: Comparing Group Mean Differences

Variable

Early

Adopters

(N=20)

Non-Early

Adopters

(N=20)

Group

Means*

T-

statistics

One-tailed

P-value

Effect

Sizes**

Levene's

Test (Sig.)

SIZE

t-1

6.284

7.083

-0.799

-0.777

0.221

0.246

0.145

LEV

t-1

0.648

0.555

0.093

0.707

0.242

0.224

0.255

GROWTH

t-1

-0.489

0.170

-0.659

-1.865

0.035

-0.590

0.992

ROE

t-1

-2.229

1.445

-3.674

-1.755

0.044

-0.555

0.426

ΔRE

t-1

4.899

1.003

3.896

0.695

0.248

0.220

0.031

Variable Definitions:

REEFFECT

t-1

= Cumulative Adoption Effect of ASC 606 as of the Date of Adoption ( in million dollars)

SIZE

t-1

= Natural log of Total Sales

t-1

(t = Adoption Year)

LEV

t-1

= Total Debt

t-1

/Total Assets

t-1

GROWTH

t-1

= (Pretax Income

t-1

- Pretax Income

t-2

) /Absolute value of Pretax Income

t-2

ROE

t-1

= Pretax Income

t-1

/ Absolute value of Total Equity

t-1

ΔRE

t-1

= (Cumulative RE effect

t-1

/ Total Assets

t-1

)

*100

New Revenue Standard ASC606

The results show that the mean differences between early adopters and non-early adopters

in Growth

t-1

(t = -1.865, p = 0.035) and ROE

t-1

(t = -1.755, p =0.044) are statistically significant

at the 5% level, and both are consistent with the expectations that less profitable firms or firms

with deteriorating preadoption performance are more likely to be motivated to adopt ASC 606

early. Although the mean differences in other indicator variables (SIZE

t-1

, LEV

t-1

, ΔRE

t-1

) are

not statistically significant, the direction of the differences in LEV

t-1

and ΔRE

t-1

follows the

conjecture that early adopters were subjected to tighter contractual constraints and experienced a

larger impact of adoption.

To further understand the magnitude of the mean difference, I used Cohen’s d as a

reference which provides point estimates for effect sizes in standard deviation (SD) units. The

guidelines of Cohen’s d suggest that the effect size is small if Cohen’s d is less than 0.2 and large

if it is over 0.8. The results indicate that the effect size of the group mean differences among the

variables is between 0.2 and 0.6. In particular, the two preadoption performance indicator

variables Growth

t-1

and ROE

t-1

exhibit relatively larger effect sizes with Cohen’s d at -0.590

and 0.555, respectively. The results suggest that preadoption financial performance may be an

important factor in managers’ consideration of the adoption timing decisions of ASC 606.

One concern with using the independent samples t-test is that it assumes homogeneous

variances in the population. Levene’s test for equality of variances addresses whether this

assumption is violated by comparing the variability of the group means in the sample. If it is

significant, then this assumption is violated. The results indicate that this assumption is not

violated for all the variables except for ΔRE

t-1

with Levene’s test significant at the 5% level.

Since the mean difference in ΔRE

t-1

is not statistically significant based on the t-test (p = 0.248),

New Revenue Standard ASC606

the comparison between the early adopters and non-early adopters suggests that these two groups

differ in their preadoption performance as shown by Growth

t-1

and ROE

t-1

in Table 10.

Compared with the independent samples t-test, logistic regression makes fewer

assumptions. It does not require normal distribution of predictor variables, or homoscedasticity

of variances when making inferences about the coefficients. Logistic regression requires an

absence of multicollinearity, independence of errors (each observation is independent of other

observations), and a categorical binary dependent variable (two possible outcomes) where the

focus category/target group is coded as 1 to be consistent with the odds statement in SPSS.

According to Meyers, Gamst, and Guarino (2017), collinearity or multicollinearity exists when

two or more predictors are strongly correlated. The general rule of thumb of 0.7 or more is used

to determine if a collinearity or multicollinearity problem exists.

Table 11A reports the Pearson correlations between the independent variables.

Table 11A. Pearson Correlations

Variable

SIZE

t-1

LEV

t-1

GROWTH

t-1

ROE

t-1

ΔRE

t-1

SIZE

t-1

Pearson Correlation

LEV

t-1

Pearson Correlation

-0.23

Sig. (2-tailed)

0.889

GROWTH

t-1

Pearson Correlation

0.235

-0.242

Sig. (2-tailed)

0.144

0.132

ROE

t-1

Pearson Correlation

0.481**

-0.062

0.254

Sig. (2-tailed)

0.002

0.702

0.113

ΔRE

t-1

Pearson Correlation

-0.186

.505**

-0.402*

0.135

Sig. (2-tailed)

0.249

0.001

0.010

0.408

Variable Definitions:

REEFFECT

t-1

= Cumulative Adoption Effect of ASC 606 as of the Date of Adoption ( in million dollars)

SIZE

t-1

= Natural log of Total Sales

t-1

(t = Adoption Year)

New Revenue Standard ASC606

LEV

t-1

= Total Debt

t-1

/Total Assets

t-1

GROWTH

t-1

= (Pretax Income

t-1

- Pretax Income

t-2

) /Absolute value of Pretax Income

t-2

ROE

t-1

= Pretax Income

t-1

/ Absolute value of Total Equity

t-1

ΔRE

t-1

= (Cumulative RE effect

t-1

/ Total Assets

t-1

)

*100

The results show that the correlation between LEV

t-1

and ΔRE

t-1

is the highest among the

variables. They are positively and significantly correlated (r = .505, p = .01). A potential reason

for the relatively strong correlation between LEV

t-1

and ΔRE

t-1

might be that both variables are

scaled by total assets

. An alternative option is to use market value as the scaler, but a

preliminary examination of the logistic regression reveals that scaling adoption effects by total

assets is a better modeling choice than by market value because the Nagelkerke pseudo-R

value

and the predictive accuracy in the classification table are both higher when total assets are used

as scalers. The correlation matrix also indicates that Size

t-1

is positively and significantly

corrected with ROE

t-1

(r = .481, p = .01), and Growth

t-1

is negatively and significantly

correlated with ΔRE

t-1

(r = -.402, p = .05). They are all in the expected direction.

In summary, the correlations among independent variables in the empirical model are less

than 0.7, thus, collinearity problems are not a serious concern. Pearson correlations also reveal

that total assets and total sales are strongly and significantly related (r = 0.935, p = 0.01).

Because both LEV

t-1

and ΔRE

t-1

are scaled by total assets, I chose to use total sales as a proxy

variable for firm size to keep the correlations among independent variables as low as possible.

Although multicollinearity is not concerning based on correlation statistics, I chose to

conduct formal testing on multicollinearity to provide additional assurance. Specifically, I tested

One potential option to reduce the correlation is to scale the cumulative adoption effects by the beginning market

value in the transition year to reduce the influence of total assets, however, the market value of two early adopters in

year t-1 is not available in Compustat. If I eliminated them, I had to remove their matched non-early adopters, which

would further reduce the sample size and thus the staitiscal power of the empirical tests.

New Revenue Standard ASC606

multicollinearity by regressing each independent variable on all other independent variables

using the ordinary least square (OLS) method, and present the results in Table 11B.

Table 11B. Tests of Multicollinearity (OLS Regression: N=40)

Model 1

Tolerance

VIF

Model 3

Tolerance

VIF

LEV

t-1

0.728

1.374

SIZE

t-1

0.678

1.476

GROWTH

t-1

0.742

1.349

LEV

t-1

0.700

1.428

ROE

t-1

0.850

1.176

ROE

t-1

0.678

1.475

ΔRE

t-1

0.602

1.662

ΔRE

t-1

0.633

1.581

DV: SIZE

t-1

(adjusted R

= 0.245, F = 0.007)

DV: Growth

t-1

(adjusted R

= 0.174, F = 0.029)

Model 2

Tolerance

VIF

Model 4

Tolerance

VIF

SIZE

t-1

0.704

1.420

SIZE

t-1

0.926

1.080

GROWTH

t-1

0.742

1.349

LEV

t-1

0.736

1.358

ROE

t-1

0.654

1.529

GROWTH

t-1

0.808

1.237

ΔRE

t-1

0.726

1.377

ΔRE

t-1

0.650

1.539

DV: LEV

t-1

(adjusted R

= 0.220, F = 0.012)

DV: ROE

t-1

(adjusted R

= 0.307, F = 0.002)

Model 5

Tolerance

VIF

SIZE

t-1

0.753

1.327

LEV

t-1

0.940

1.064

GROWTH

t-1

0.867

1.153

ROE

t-1

0.747

1.339

DV: RE

t-1

(adjusted R

= 0.397, F = 0.001)

Variable Definitions:

SIZE

t-1

= Natural log of Total Sales

t-1

(t = Adoption Year)

LEV

t-1

= Total Debt

t-1

/Total Assets

t-1

GROWTH

t-1

= (Pretax Income

t-1

- Pretax Income

t-2

) /Absolute value of Pretax Income

t-2

ROE

t-1

= Pretax Income

t-1

/ Absolute value of Total Equity

t-1

ΔRE

t-1

= (Cumulative RE effect

t-1

/ Total Assets

t-1

)

*100

Meyers, Gamst, and Guarino (2017) suggest that multicollinearity exists if multiple

correlations are high, that is, the adjusted R

is over 0.6. None of the adjusted R

in the models

above exceeds 0.4 (see Table 11B), suggesting that multicollinearity is less of a concern. In

addition, SPSS provides some indices that help identify multicollinearity problems. For example,

tolerance (1- R

)

is the amount of variance in a predictor not explained by the other predictors.

New Revenue Standard ASC606

Or a related statistic, variance inflation factors (VIF = 1/tolerance), also helps identify

multicollinearity problems. According to Meyers, Gamst, and Guarino (2017), tolerances less

than 0.4 (or VIF > 2.5) indicate that variables may be problematic and tolerances less than 0.1 (or

VIF > 10) suggest a severe multicollinearity problem. Based on the tolerance or the VIF values

in Table 11B, multicollinearity is not a concern.

The results indicate that a multivariate test is an appropriate approach to estimate the

simultaneous effect of independent variables on the early adoption decision of ASC 606, after

taking both the correlations between independent variables (as shown in Table 10A) and the

analysis of multicollinearity (as shown in Table 11B) into consideration.

I first ran the univariate logistic regression using the 20 early adopters and 20 industry-

size-matched non-early adopters and presented the results in Table 12A.

Table 12A. Univariate Logistic Regression Analysis (N=40)

Variable*

Expected

Sign

Coefficients

Value

Exp

(B)

Chi-

Square

Sig

Pseudo-

H&L

Test Sig

Predictive

Accuracy

SIZE

t-1

-0.080

0.432

0.924

0.427

0.021

0.182

50.0%

LEV

t-1

(+)

0.581

0.479

1.788

0.467

0.018

0.614

55.0%

GROWTH

t-1

(-)

-0.676

0.101

0.509

0.050

0.122

0.650

60.0%

ROE

t-1

(-)

-1.760

0.092

0.172

0.003

0.260

0.532

60.0%

ΔRE

t-1

(+)

0.014

0.503

1.014

0.467

0.018

0.201

62.5%

a Null hypothesis of the omnibus tests of model coefficients: the coefficients are zero

b.Nagelkerke pseudo-R2 explains the percentage change of DV explained by IV in the model

c. Null hypothesis of Hosmer and Lemeshow Test: the predictions match the observed values

d. The probability of a correct prediction (percentage correctness)

Variable Definitions:

SIZE

t-1

= Natural log of Total Sales

t-1

(t = Adoption Year)

LEV

t-1

= Total Debt

t-1

/Total Assets

t-1

GROWTH

t-1

= (Pretax Income

t-1

- Pretax Income

t-2

) /Absolute value of Pretax Income

t-2

ROE

t-1

= Pretax Income

t-1

/ Absolute value of Total Equity

t-1

ΔRE

t-1

= (Cumulative RE effect

t-1

/ Total Assets

t-1

)

*100

New Revenue Standard ASC606

The results show that the coefficients on all the variables of interest have the predicted

signs, but only Growth

t-1

(χ2[1] =3.851, p =0.050) and ROE

t-1

(χ2[1] =8.687, p = 0.003) are

negatively and statistically significantly related to early adoption decision at a 10% level. The

odds ratio, a robust measure of effect-size statistics, calculates the odds of an event happening to

one group relative to another. An odds ratio higher than 1 indicates that the treatment group is

more likely to experience an event than the control group. However, an odds ratio below 1 is not

directly interpretable (McHugh, 2009). In practice, researchers suggest that we calculate its

reciprocal if the odds ratio is below 1 to make the interpretation more intuitive

The odds ratio for Growth

t-1

(b = -0.676, p = .101; Exp(B) = 0.509, 1/Exp(B) = 1.96)

indicates that firms with low preadoption growth rates are 1.96 times more likely to be early

adopters than firms with high preadoption growth rates. The Nagelkerke pseudo-R

in the

Growth

t-1

only model is 0.122, indicating that Growth

t-1

alone explains approximately 12.2% of

the variance in adoption timing decisions. The classification accuracy based on Growth

t-1

alone

is 60%, which is slightly higher than the model without any predictors in it (50% accuracy).

Similarly, the odds ratio for ROE

t-1

(b = - 1.760, p = .092; Exp(B) = 0.172, 1/Exp(B) =

5.81) indicates that firms with a lower rate of return are 5.81 times more likely to be early

adopters than firms with a higher rate of return. Its Nagelkerke pseudo-R

is 0.260, indicating

that the ROE

t-1

alone explains approximately 26% of the variance in adoption timing decisions.

The classification accuracy based on ROE

t-1

is 60%, slightly higher than the model with

constant only (50% accuracy).

In addition, Size

t-1

(b =-0.080, p = .432; Exp(B) = 0.924; R

=0.021), LEV

t-1

(b = .581,

p = .479; Exp(B) = 1.788; R

=0.018), and ΔRE

t-1

(b =0.014, p = .503; Exp(B) = 1.014; R

http://onbiostatistics.blogspot.com/2012/02/how-to-interpret-odds-ratios-that-

are.html#:~:text=%E2%80%9CAn%20OR%20of%20less%20than,is%20not%20the%20OR%20result.

New Revenue Standard ASC606

=0.018) are not statistically significant in the univariate tests. Size

t-1

and LEV

t-1

appear to make

no significant contribution to the prediction of outcome based on their respectively low

Nagelkerke pseudo-R

. However, the classification accuracy based on ΔRE

t-1

alone is 62.5%,

suggesting that the adoption effects are an important factor in predicting the adoption timing

decisions (i.e., early vs non-early adopters).

In addition, the Hosmer and Lemeshow Tests assess whether the predicted probabilities

are statistically different from the observed probabilities and its null hypothesis is that there is no

difference between the predictions and observed values. The results show that none of them are

statistically significant, indicating that there is an acceptable match between the predicted and

observed probabilities.

Next, I ran the multivariate logistic regression modeling the binary variable of the early

adoption decision of ASC 606 and presented the results in Table 12B.

Table 12B. Multivariate Logistic Regression Analysis

Model 1 (Baseline)

Model 2 (Interaction)

Variable*

Expected

Sign

Coefficients

Value

Exp

(B)

Coefficients

Value

Exp

(B)

SIZE

t-1

0.197

0.292

1.218

0.309

0.123

1.362

LEV

t-1

(+)

-0.902

0.589

0.406

-0.568

0.759

0.566

GROWTH

t-1

(-)

-0.146

0.775

0.864

-0.248

0.687

0.780

ROE

t-1

(-)

-2.603

0.099

0.074

-1.931

0.223

0.145

ΔRE

t-1

(+)

0.047

0.414

1.048

0.856

0.091

2.353

LEV

t-1

*ΔRE

t-1

(+)

-0.741

0.101

0.476

GROWTH

t-1

*ΔRE

t-1

(-)

-0.276

0.106

0.759

Chi-Square Sig

0.059

0.055

Nagelkerke R2

0.311

0.389

H&L Test Sig

0.576

0.102

Predictive Accuracy

60%

75%

a Null hypothesis of the omnibus tests of model coefficients: the coefficients are zero

b.Nagelkerke pseudo-R2 explains the percentage change of DV explained by IV in the model

c. Null hypothesis of Hosmer and Lemeshow Test: the predictions match the observed values

d. The probability of a correct prediction (percentage correctness)

New Revenue Standard ASC606

Variable Definitions:

SIZE

t-1

= Natural log of Total Sales

t-1

(t = Adoption Year)

LEV

t-1

= Total Debt

t-1

/Total Assets

t-1

GROWTH

t-1

= (Pretax Income

t-1

- Pretax Income

t-2

) /Absolute value of Pretax Income

t-2

ROE

t-1

= Pretax Income

t-1

/ Absolute value of Total Equity

t-1

ΔRE

t-1

= (Cumulative RE effect

t-1

/ Total Assets

t-1

)

*100

The early adopters as the target group was coded as 1 and non-early adopters as the

reference group was coded as 0. Two models were used to test the predictive power of the

independent variables. The predictor variables in Model 1 (baseline model) include firm

characteristics (SIZE

t-1

, LEV

t-1

, Growth

t-1

, ROE

t-1

) and adoption effects of ASC 606 (ΔRE

t-1

In addition to all the variables in Model 1, Model 2 (interaction model) adds two additional terms

(LEV

t-1

*ΔRE

t-1

and GROWTH

t-1

*ΔRE

t-1

) to capture the interactive effects of firm

characteristics and magnitude of adoption effects on firms’ adoption timing decisions.

The results in Model 1 show that coefficients on Growth

t-1

, ROE

t-1,

and ΔRE

t-1

are

consistent with the predicted signs, but ROE

t-1

is the only variable that is statistically

significantly associated with early adoption decisions at a 10% level ( b = -2.603, p = .099 ). The

odds ratio for ROE

t-1

(Exp(B) = 0.074; 1/Exp(B) = 13.51 ) indicates that firms with a lower rate

of return in year t-1 are 13.51 times more likely to be early adopters than firms with a higher rate

of return. This result is consistent with the findings in the univariate test for ROE

t-1

(see Table

12A)

The Nagelkerke pseudo-R2 indicates that the baseline model accounts for approximately

31.1% of the total variance in the dependent variable, which is higher than all the univariate

models (the highest Nagelkerke pseudo-R

is 26% for the ROE

t-1

model).

The Hosmer and Lemshow test is 0.576, indicating that the predicted frequencies are not

statistically different from the observed frequencies. Based on a classification threshold predicted

probability of the target group of 0.5 in the constant-only model, results of the logistic analysis

New Revenue Standard ASC606

indicate that the baseline model provides a statistically significant prediction of early adopters

(χ2[5] =10.619, p = 0.059). However, the overall correct prediction rate is 60% (i.e., a correct

prediction rate of 60% for early adopters and 60% for non-early adopters), indicating no big

improvement from the predictive accuracy of univariate models (see Table 12A). Taken together,

there is some evidence from the baseline model suggesting that the preadoption financial

performance is associated with firms’ early adoption decisions.

The interaction model (model 2) expands the baseline model by including two additional

terms (LEV

t-1

*ΔRE

t-1

and GROWTH

t-1

*ΔRE

t-1

), which provides formal testing of H1 and H2.

The impacts of the inclusion of the two interactive terms in model 2 on model statistics are

noticeable. As seen in Table 12B, there is a significant increase in the prediction accuracy based

on a cutoff value of 0.5 for the constant-only model. The overall correct prediction rate increases

from 60% in Model 1 (baseline model) to 75% in Model 2 (interaction model). More

specifically, the correct prediction rate for early adopters increases from 60% in the baseline

model to 85% in the interaction model. The correct prediction rate for non-early adopters

increases from 60% in the baseline model to 65% for non-early adopters.

The test of the null hypothesis that each of the coefficients in the model is equal to zero

yields a chi-square value (χ2[7] =13.803, p = 0.055), indicating that the interaction model

provides a statistically significant prediction of early adopters at a 10% level. The Nagelkerke

pseudo-R

indicates that the interaction model accounts for approximately 38.9% of the total

variance in the dependent variable, which is higher than the baseline model (the Nagelkerke

pseudo-R

= 31.1%) and the univariate model (the Nagelkerke pseudo-R

= 26%, the highest

among all the univariate models), indicating that the interaction model has higher predictive

power for early adopters.

New Revenue Standard ASC606

Moreover, the coefficients on Growth

t-1

(b = -.248, p = .687; Exp(B) = 0.780) and ROE

t-1

(b = -1.931, p = .223; Exp(B) = 0.145) have the predicted signs but are not statistically

significant. ΔRE

t-1

is statistically significant at a 10% level (b =0.856, p = .091; Exp(B) =

2.353) and its positive coefficient is in the direction predicted. The multivariate results provide

evidence consistent with the prediction that firms are more likely to adopt ASC 606 early if this

standard has produced more favorable effects on firms' financials.

H1 predicts a positive relationship between the adoption effect of ASC 606 and the

propensity of firms approaching debt constraints in adopting the standard early. This hypothesis

would be supported by a positive coefficient β

on the first interactive term LEV

t-1

*ΔRE

t-1

Although the coefficient is not positive, it is not statistically significant (b = -.741, p = .101;

Exp(B) = 0.476). Thus, the result does not provide support for the hypothesis of debt constraint

considerations (H1) in early adoption decisions. This result is different from the findings in the

independent samples test (see Table 10) and the univariate test on leverage (see Table 12A)

which provides preliminary evidence that early adopters are more leveraged than non-early

adopters.

A potential reason for the insignificant coefficient of β

could be attributed to the

unusually high leverage ratio in a few of the sample firms. For example, LEV

t-1

is 85% for early

adopters and 80% for non-early adopters at the 75% percentile as shown in Table 8B. I checked

the sample and found that a few firms’ debt-to-asset ratios are close to or even over 1, indicating

that these firms were mostly funded by debt, thus putting them at an extremely high risk of

defaulting on debt. These companies are more likely to have a lower degree of flexibility in

managing their legal obligations as they are very closely monitored by their creditors, which

could influence their reporting choices. To get an idea of the influence of the extremely

New Revenue Standard ASC606

leveraged firms, I excluded four observations (two early adopters and two non-early adopting

peers) with LEV

t-1

over 1 and re-ran the regression. I find a positive but not statistically

significant coefficient on LEV

t-1

*ΔRE

t-1

(b =0.261, p = .827; Exp(B) = 1.298 ). Moreover, the

prediction accuracy dropped to 69.4% when the sample size became 36.

H2 predicts a negative relationship between the adoption effect of ASC 606 and the

propensity of firms with an increasing pre-adoption financial performance to adopt the standard

early. The coefficient β

on the second interactive term GROWTH

t-1

*ΔRE

t-1

is negative as

predicted with a statistical significance slightly above the 10% level (b = -0.276, p = .106;

Exp(B) = 0.759). This finding is also consistent with the evidence from the independent samples

test (see Table 10) and the univariate test on preadoption financial performance (see Table 12A),

which show that early adopters have statistically significantly lower growth rates than non-early

adopters. In addition, the coefficient on ΔRE

t-1

is statistically significant and positive at a 10%

level (b =0.856, p = .091; Exp(B) = 2.353) and consistent with its predicted sign. Thus, the

multivariate results provide evidence supporting hypothesis 2 that there is a negative relationship

between the adoption effect of ASC 606 and the propensity of firms with an increasing pre-

adoption financial performance to adopt the standard early. Based on the results of the omnibus

tests of model coefficients ( χ2[7] =13.803, p = 0.055), the Nagelkerke pseudo-R

(38.9%),

and

the classification accuracy (75%), the interaction model improves the predicting power for early

adopters of ASC 606 compared with the baseline model.

New Revenue Standard ASC606

Chapter 5: Discussion and Conclusion

As discussed in previous chapters, ASC 606 has brought the most influential accounting

changes since the Sarbanes-Oxley Act (SOX) of 2002. The changes by ASC606 call for more

research to better understand the impacts and consequences of this standard. Most concurrent

research on ASC 606 took an information perspective and examined the relationship between the

new revenue standard and its market effect, focusing on the tests of market efficiency and value

relevance for new accounting information. The results from these market-based studies provided

mixed findings on whether the new standard improves the comparability of financial statements,

earnings informativeness, earnings predictability, and stock liquidity.

This study complemented existing research on ASC 606 by taking a contracting

perspective and drawing from the positive accounting theory to examine the adoption effect on

firms' financial statements and the timing of the adoption decisions. This study provided

evidence of the impact of ASC 606 on firms’ financials and examined whether the likelihood of

adopting ASC 606 early is associated with firm-specific characteristics and adoption effects on

firms’ financial statements. Collectively, the market-based approach and the contracting costs-

based approach complemented each other to enhance our understanding of ASC 606 and its

impact.

5.1 Implications for Research

Using a sample of 20 early adopters and 20 non-early adopting peers, I found that the

adoption effects of ASC 606 on firms’ financials vary by industry. Firms in the computer

programing and software industry reported the largest dollar amount increase in retained

earnings while the transportation industry reported the largest dollar amount decrease in retained

earnings in the year of adoption. Overall, the majority of the sample firms (83%) reported

New Revenue Standard ASC606

positive adoption effects on financial statements, which on average increased opening retained

earnings by 2.95% as of the date of adopting ASC 606.

I also found that both the mean and median adoption effects of early adopters are higher

than those of their non-early adopting peers. The multivariate test of the interaction model shows

that the adoption effects (ΔRE

t-1

) are statistically significant and positively associated with the

likelihood of adopting ASC 606 early. These results are consistent with the expectation that early

adopters of ASC 606 had more favorable adoption effects than their non-early adopting peers.

The positive adoption effects could be a potential consideration for firms’ adoption timing

decisions since it could help reduce contracting costs associated with restrictive debt covenants

and meet market expectations of earnings and growth.

Positive accounting theory suggests that managers are more likely to increase reported

earnings to reduce the risk of a technical default when a firm is close to its debt covenant limits

(Watts & Zimmerman; 1990). In hypothesis 1, I predict that there is a positive relationship

between the adoption effect of ASC 606 and the propensity of firms that approach debt limits to

adopt the standard early. This hypothesis predicts the coefficient on the interactive term LEV

*ΔRE

t-1

to be positive. Although there is evidence that early adopters are more leveraged and

report larger cumulative adoption effects than non-early adopters in univariate analysis, I did not

find a statistically significant association between the adoption timing decision and the

interaction of leverage and adoption effects in the multivariate analysis. This result might be

because some sample firms are highly leveraged (debt-to-asset ratios close to or above 1). These

firms are more sensitive to economic and earnings declines and are subject to a higher default

risk or bankruptcy risk. They are monitored closely by creditors and have limited flexibility to

New Revenue Standard ASC606

use income-increasing strategies from accounting changes to influence the leverage ratios, which

could influence the reporting choices that managers can make.

Positive accounting theory also suggests that managers have incentives to manage

accounting-based earnings to meet market expectations, justify increased compensations, or

increase job security (Watts & Zimmerman; 1990). This is because a consistent trend of revenue

growth affects the perception of management performance and stock prices. If the revenue

growth rate is below internal or external expectations, managers are more likely to choose

income-increasing accounting changes to get closer to or meet the targeted growth rate. In

hypothesis 2, I predict that there is a negative relationship between the adoption effect of ASC

606 and the propensity of firms with an increasing preadoption performance to adopt the

standard early. This hypothesis predicts the coefficient on the interaction term GROWTH

*ΔRE

t-1

to be negative. I found the coefficient on GROWTH

t-1

*ΔRE

t-1

negative and just right

above the significance level of 10% (p = 0.106). This finding is consistent with the evidence

from the independent samples tests and the univariate tests, which show that GROWTH

t-1

and

ROE

t-1

are negatively and significantly related to the likelihood of early adoption. This finding is

also consistent with the evidence from the multivariate test which shows that ΔRE

t-1

is positively

and significantly related to the likelihood of adopting ASC 606 early. Taken together, this study

provides evidence supporting hypothesis 2 that early adopters of ASC 606 are associated with

decreased earnings in the year before adoption and more favorable adoption effects than non-

early adopting peers.

Finally, positive theory suggests that firm size is a proxy variable for political costs and

information production costs. Large firms are associated with higher political costs but lower

information production costs. Prior literature provides mixed findings on whether large firms are

New Revenue Standard ASC606

more or less likely to be early adopters. The results of my study show that SIZE

t-1

is positively

related to the likelihood of early adoption, but it is not statistically significant. Statistics show

that “24 percent of non-reliance notices (alert investors not to rely on financial statements) are

caused by the problems with the ASC 606 adoption in 2018. And 2 percent of late filings in 2018

cited revenue recognition problems as the reason for tardy filings” (Whitehouse, 2019). The

result about firm size is not surprising as it indicates that large firms are likely to have the

resources and expertise to adopt ASC 606 earlier than small firms considering the complexity of

the new standard and its enhanced disclosure requirements, the time and level of effort needed to

implement it.

This study has made the following contributions to the literature. First, this paper

contributes to the broad literature studying the impact of the new revenue standard. While most

existing studies focused on the impact of the new standard on market efficiency and value

relevance, limited research examined the direct impact of ASC 606 on firms’ financials. One

paper (Arora, 2019) assessed the impact of ASC 606 on the airline industry and found that the

new standard increased aggregated net income of eight airline firms by about 2%. Another paper

(Choi, Kim & Wang, 2022) found that software firms are more affected by ASC 606 than

electronic computer firms and ASC 606 improves financial statement comparability and

informativeness for software firms. Both papers focused on specific industries in their

examination of the financial reporting impact of ASC 606. This study focuses on early adopters

but does not limit the sample to any specific industry. By comparing the adopting effects

between early and non-early adopters, this study helps us understand the potential incentives

underlying financial reporting. Thus, this study highlights the variations in managerial responses

to ASC 606. These variations could be circumstances-dependent (e.g., how leveraged the firm is

New Revenue Standard ASC606

or how profitable it is before adoption) and may provide insight into mixed findings of the ASC

606 effects in existing market-based studies. Second, this study tests positive accounting theory

in the context of ASC 606 and finds differences between early and non-early adopting peers in

both the adoption effects and their preadoption financial performance. Third, this study provides

additional evidence for the employment of the interactive effects model in the accounting choice

literature, which echoes the recommendation in prior research (Ali & Kumar, 1994; Gujarathi &

Hoskin, 2003). Finally, this study provides information about ASC 606 effects in response to the

calls from the FASB for its ongoing post-implementation review of the new revenue standard.

5.2 Implications for Practice

To shed light on the adoption effects of ASC 606, I read through and collected

information from the disclosures of the accounting changes and revenue recognition. The

disclosures on the revenue recognition policies and process from 10-Ks provide more

information about the nature and scope of changes in earnings and revenue due to the adoption of

ASC 606, which is useful for understanding the impact of this new standard. I focused on the

revenue recognition section under Part II, Item 8 - Financial Statement and Supplemental Data,

and summarized the findings as follows.

First, private firms that worked toward IPO offerings in 2017, 2018, and 2019 may

benefit from early adoption of the new revenue standard because proactive actions to update

systems, internal controls, and process help firms understand the impact of the new revenue

standard on financial statements sooner and reduce future information production costs. These

firms would save time on reviewing contracts for modification and avoid performing dual

tracking of revenues, especially for firms that opted for the full retrospective method.

Retroactively recalculating revenue would be time-consuming and resource intensive. Firms that

New Revenue Standard ASC606

expect major differences between revenue accounted for under ASC 606 vs ASC 605 may

benefit from the full retrospective method as it improves the comparability of financial reports

and helps firms mitigate negative adoption effects if ASC 606 accelerates revenue recognition.

Second, some emerging growth companies (EGCs) as defined in the Jumpstart Our

Business Startups Act of 2012 (JOBS Act) chose to be early adopters as EGCs have the option to

adopt ASC 606 early or at the standard adoption date for either a public company or a private

company (e.g., Docusign, Inc.). These firms may take advantage of specified reduced reporting

and disclosure requirements (e.g., present only two years of selected financial statements on the

initial registration form). Early adoption could reduce the negative impact of ASC 606 on EGCs

as the periods presented would be post-adoption periods on form S-1. This might be more

important for newly incorporated firms.

Third, several high-tech early adopters disclosed that the impact of the new standard

relates to their accounting for distributor and reseller revenue when they change from the more

conservative “sell-through method” (revenue is not recognized until goods are sold from

distributors to end-users) to the more aggressive “sell-in method” (revenue is recognized upon

transfer of control to customers including distributors). Under the sell-in method, firms need to

make estimates for variable consideration (e.g., sales returns and rebates) at the time of transfer

of control. The new standard results in a reduction in deferred distributor revenue and accounts

receivable for estimated returns (see Extreme Networks Inc., 2017 Annual Report). This finding

may bring additional information to the studies of sell-in versus sell-through revenue

recognition, where prior research shows that high-tech firms are less likely to use the “sell-in

method” when they have higher growth opportunities and strong corporate governance

(Rasmussen, 2009).

New Revenue Standard ASC606

Last, some early adopters (e.g., Commvault Systems, Inc.; Datadog, Inc.) disclosed that

they elected to early adopt ASC 606 using the full retrospective method and the adoption

increased retained earnings due to the deferral of commission costs, which are capitalized under

ASC 606 but expensed under ASC 605.

As indicated in the concurrent study (Choi, Kim & Wang, 2022), the effects of ASC 606

are likely to vary by industry because of the fundamental differences in their operating

environment, business models or practices, contract designs, and pre-ASC 606 revenue rules.

The evidence in this study could be of interest to investors, practitioners, and regulators.

5.3 Limitations and Future Studies

There are several limitations to this study. First, the final sample size of early adopters is

relatively small due to data limitations and the firms are mostly from eight industries, which

potentially limits the generalization of the findings to the other industries. Second, the size of the

control sample is limited due to the limitation of manual data collection. Manually collecting

adoption effects from Financial Statements and Supplemental Data requires extensive time and

accounting skills. I manually collected the cumulative adoption effects of early adopters and non-

early adopting peers matched by industry and size. Future studies can expand to include more

non-early adopters in the study to conduct an imbalanced sample test. Third, the study finds that

early adopters tend to choose the full retrospective method while non-early adopters tend to use

the modified retrospective method when adopting ASC 606. Future studies can investigate

whether significant differences exist between the firms in choosing the full retrospective method

and the modified retrospective method, and what the implications are to decision-makers. Last,

future studies may replicate this study using sample firms that chose to early adopt IFRS 15.

New Revenue Standard ASC606

5.4 General Conclusions

This study presented evidence of the direct impact of ASC 606 on firms’ financial

statements and examined the associations between the adoption timing choice and firm

characteristics. The results reveal that early adopters of ASC 606 are associated with more

favorable adoption effects and decreased earnings in the year before the adoption. The results

also show that most early adopters chose the full retrospective while most non-early adopters

chose the modified retrospective method. The findings may be interesting to investors to enhance

their understanding of the new revenue standard to make well-informed decisions. The findings

might also be interesting to managers who can use the increased discretion and judgment under

ASC606 to improve the quality of revenue reporting. Finally, the findings could be interesting to

standard setters who want to understand the impact and implications of adoption choices when

designing new accounting standards and reviewing existing standards.

84 
New Revenue Standard ASC606 
 
References 
Ahn, J., Hoitash, U., & Schmardebeck, R. (2021). The Informativeness of SAB 74  
Disclosures for Accounting Standard Adoption Outcomes: Evidence from ASC  
606 (December 07, 2021). Working Paper available at  
SSRN: https://ssrn.com/abstract=3835849   
Ali, A. & Kumar, K.R. (1994). The Magnitudes of Financial Statement Effects and  
Accounting Choice: The Case of the Adoption of SFAS 87. Journal of Accounting and 
Economics. Vol. 18, Issue 1, pp. 89-114. https://doi.org/10.1016/0165-4101(94)90020-5 
Ali, W. & Tseng, A. (2022). The Effects of ASC 606 on Revenue Recognition. Working  
Paper available at 
https://www8.gsb.columbia.edu/ceasa/sites/ceasa/files/Ali%20Tseng%2002-07-2022.pdf 
Arms, D. & Bercik, T. (2015). Preparing for ASC 606. Strategic Finance. October 2015.  
https://sfmagazine.com/post-entry/october-2015-preparing-for-asc-606/ 
Arora, S. (2019). Implications of ASC 606 on Airline Financial Statements. Management 
Accounting Quarterly. Vol. 21, Issue 1, 15-24 
AT&T Annual Report (2018). https://investors.att.com/financial-reports/annual-reports/2018 
Ayres, F.L. (1986). Characteristics of Firms Electing Early Adoption of SFAS 52. Journal of 
Accounting and Economics. Vol. 8. PP. 143-158 
Bellomy, A. & Driscoll, S. (2020). Licenses for Intellectual Property. RevenueHub. 
https://www.revenuehub.org/licenses-intellectual-property/ 
Beneish, M. & Press, E. (1993). Costs of Technical Violation of Accounting-Based Debt 
Covenants. The Accounting Review. Vo. 68. No. 2. PP. 233-257 

New Revenue Standard ASC606

Benston, G. J., Bromwich, M., & Wagenhofer, A. (2006). Principles‐versus rules‐based

accounting standards: the FASB's Standard-setting Strategy. Abacus, 42(2), 165-188.

Bjornsen, M. & Fornaro, J. (2019). Principles-Based Accounting Standards and

Corporate Governance Considerations. Journal of Accounting and Management. Vol. 9,

No. 1.

Carmichael, D. R. (2019). New Revenue Recognition Guidance and the Potential for Fraud and

Abuse: Are Companies and Auditors Ready? The CPA Journal, 89(3), 36-43.

Chung, K. H., & Chuwonganant, C. (2019). New Revenue Recognition Standard and Earnings

Informativeness. Working Paper. Available at SSRN 3508041.

Choi, D., Kim, S. & Wang, X. (2022). Heterogeneity in the Financial-Reporting Effects of ASC

606 Adoptions: Evidence from an Industry-Focused Approach. Working paper available

at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4111345.

Deloitte (2017). Accounting and Financial Reporting Update — Including Interpretive Guidance.

https://www2.deloitte.com/content/dam/Deloitte/us/Documents/audit/us-audit-life-

sciences-accounting-financial-reporting-update-2017.pdf

Delta Annual Report (2018). https://www.annualreports.com/Company/delta-air-lines-inc

Donelson, D.C., McInnis, J.M. & Mergenthaler, R.D. (2012). Rules-Based Accounting Standards

and Litigation. The Accounting Review, Vol.87, No.4.

Drew, J. (2021). FASB Issues 2 New Financial Accounting Concept Statements. Journal of

Accountancy. December 23, 2021

https://www.journalofaccountancy.com/news/2021/dec/fasb-2-inancial-accounting-

concept-statements.html

New Revenue Standard ASC606

Eakin, C.F. & Gramlich, J. D. (2000). Insider Trading and the Early Adoption of SFAS 96: A

test of the Signaling Hypothesis. Advances in Accounting. 17, p111-133

Fama, E.F. & French, K.R. (2001). Disappearing Dividends: Changing Firm Characteristics or

Lower Propensity to Pay? Journal of Financial Economics 60, p3-43

Ferreira, P. (2020). The Liquidity, Precision, and Comparability Effects of ASC 606:

Revenue from Contracts with Customers. Working Paper. Available at

https://ssrn.com/abstract=3522765

Fields, D. F., Lys, T.Z., & Vincent, L. (2001). Empirical Research on Accounting Choice.

Journal of Accounting and Economics 31, p. 255-307

Financial Accounting Standards Board (FASB). 2014-09. Revenue from Contracts with

Customers (Topic 606). An Amendment of the FASB Accounting Standards Codification.

No. 2014-09. May 2014.

Financial Accounting Standards Board (FASB). Accounting Standards Update (ASU) No. 2015-

14. Revenue from Contracts with Customers (Topic 606): Deferral of the Effective Date.

https://www.fasb.org/jsp/FASB/Document_C/DocumentPage?cid=1176166272502&acce

ptedDisclaimer=true

Financial Accounting Standards Board (FASB). Accounting Standards Update (ASU) No. 2020-

05. Revenue from Contracts with Customers (Topic 606) and Leases (Topic 842):

Effective Dates for Certain Entities.

https://www.fasb.org/jsp/FASB/Document_C/DocumentPage?cid=1176174696379&acce

ptedDisclaimer=true

Financial Accounting Standards Board (FASB). 2020-07-29. Revenue Post-

Implementation Review (PIR).

Glaze, J., Skinner, A. N., & Stephan, A. (2021). When are Concurrent Quarterly Reports

New Revenue Standard ASC606

Useful for Investors? Evidence from ASC 606. Working Paper. Available at

SSRN: https://ssrn.com/abstract=3639777

Gordon, E., Henry, E. & Hsu, H. (2018). Revenue Recognition: A Brave New World.

Working Paper. Available at https://ssrn.com/abstract=3088041

Guillaume, O., & Pierre, D. (2016). The Convergence of US GAAP with IFRS: A comparative

analysis of principles-based and rules-based accounting standards. Scholedge

International Journal of Business Policy & Governance, 3(5), 63-72.

Gujarathi, M.R. & Hoskin, R.E. (2003). Modeling the Early Adoption Decision: The Case of

SFAS 96. Review of Accounting & Finance. Vol. 2. November 4, 2003.

Han, L. & Hsiao. D.F. (2017). Examination of Firm Performance Following the Early Adoption

of SFAS 142. International Journal of Accounting & Information Management. Vol. 25.

PP. 138-176

Hu, H. (2021). Does the Converged Revenue Recognition Standard Improve Revenue

Recognition Comparability Between U.S. GAAP and IFRS? Working Paper.

https://hdl.handle.net/11244/330880

Hunsaker, M. (2018). ASC 606 Adoption Considerations. IPOhub. Oct. 6, 2018

https://www.ipohub.org/asc-606-adoption-considerations/

Hepp, J. (2018). ASC 606: Challenges in understanding and applying revenue recognition.

Journal of Accounting Education, 42, 49-51.

Kishi, J. (2020). New Revenue Recognition Landscape for Retailers. Article retrieved at

https://www.hcvt.com/article-ASC-606-revenue-recognition-retailers

Jones, J.P., & Pagach, D. (2013). The Next Step for Revenue Recognition. The CPA Journal,

83(10), 30

New Revenue Standard ASC606

Knachel, E. (2017). Revenue Recognition Disclosure Requirements: A Challenge that Can’t

Wait. CFO Publishing April 17, 2017

KPMG (2016). The Great Accounting Challenge – KPMG’s 2016 Accounting Change Survey.

https://assets.kpmg/content/dam/kpmg/pdf/2016/07/kpmg-accounting-change-survey.pdf

Kral, R., & Levy, H. B. (2019). Selecting Modified Retrospective Transition for Adopting ASC

606 and Related Standards: Minimum Disclosure Requirements and Management

Options. The CPA Journal, 89(9), 58-61.

Langer, R. & Lev, B. (1993). The FASB’s Policy of Extended Adoption for New Standards: An

Examination of FAS No. 87. The Accounting Review. Vol. 68. No. 3. Pp. 515-533

Lee, K., & Lee, S. W. (2020). Rules-based vs. Principles-based Accounting Standards: Earnings

Quality and the Role of Earnings in Contracting. Working Paper available at

http://dx.doi.org/10.2139/ssrn.3534039

Levitt, A. (1998). The “Numbers Game”. Remarks by Chairman of SEC at NYU Center for Law

and Business, September 28, 1998.

https://www.sec.gov/news/speech/speecharchive/1998/spch220.txt

Leone, M. (2009). New Revenue-Recognition Rules: The Apple of Apple’s Eye?

https://www.cfo.com/accounting-tax/2009/09/new-revenue-recognition-rules-the-apple-

of-apples-eye/

Lyons, B. & Tarasovich, B. (2018). The New Revenue Recognition Rules. Strategic Finance.

December 1, 2018. https://sfmagazine.com/post-entry/december-2018-the-new-revenue-

recognition-rules/

Marco, D. and Ovuka, S. (2017). Revenue Recognition: We’re in the Homestretch! Financial

Executives International. November 2017.

New Revenue Standard ASC606

McCarthy, M., & McCarthy, R. (2014). Financial statement preparers' revenue decisions:

Accuracy in applying rules-based standards and the IASB-FASB revenue recognition

model. Journal of Accounting and Finance, 14(6), 21.

McClane, J. (2019). Boilerplate and the Impact of Disclosure in Securities Dealmaking.

Vanderbilt Law Review. Vol. 72, Issue 1, p191-295.

McHugh, M. (2009). The Odds Ratio: Calculation, Usage, and Interpretation. Biochemia Media.

Vol. 19, Issue 2, p120-126

McKenna, F. (2017). Boeing and Other Big Defense Companies Face Special Challenge in New

Revenue Rules. MarketWatch. https://www.marketwatch.com/story/boeing-and-other-

big-defense-companies-face-special-challenge-in-new-revenue-rules-2017-08-23

Meyers, L., Gamst, G., & Guarino, A. (2017). Applied Multivariate Research: Design and

Interpretation. Third Edition. Thousand Oaks: Sage Publications.

Mortensen, J, & Schwendiman, A. (2020). Lost Revenue. RevenueHub, February 11, 2020.

https://www.revenuehub.org/lost-revenue/

Nehoray, M., Rose, B., Das, J., & Sidher, S. (2018). ASC 606 and Transfer Pricing: The Devil is

in the Details. https://www2.deloitte.com/content/dam/Deloitte/us/Documents/Tax/us-

tax-asc-606-transfer-pricing.pdf

Olsen, L., & Weirich, T. R. (2010). New Revenue‐Recognition Model. Journal of Corporate

Accounting & Finance, 22(1), 55-61.

Papa, V. (2016). Watching The “Top Line”: Area for Investor Scrutiny on Revenue Recognition

Changes. CFA Institute. https://www.cfainstitute.org/en/advocacy/policy-

positions/watching-the-top-line

New Revenue Standard ASC606

Peters, Mo. (2016). The New IFRS 15 Standard: Implementation Challenges for Belgian

Companies.

Peters, R. (2018). Impacts of New Revenue Recognition Standards on Public Companies

https://blog.auditanalytics.com/early-adoption-of-asc-606-first-glance/

PWC (2019). Revenue from Contract with Customers.

https://www.pwc.com/gx/en/services/audit-assurance/assets/revenue-from-contracts-

with-customers.pdf

Rasmussen, S.J.B. (2009). Sell-in Versus Sell-through Revenue Recognition: An Examination of

Firm Characteristics and Financial Information Quality. Texas A&M University.

Regan, G. & Regan, T. (2007). Software Revenue Recognition on the Rise. Journal of

Accountancy. https://www.journalofaccountancy.com/issues/2007/dec/

softwarerevenuerecognitionontherise.html

Reddy, J. (2018). Welcome to year one of ASC 606. Accounting Today.

https://www.accountingtoday.com/opinion/welcome-to-year-one-of-asc-606

Rutledge, R. W., Karim, K. E., & Kim, T. (2016). The FASB's and IASB's New Revenue

Recognition Standard: What Will Be the Effects on Earnings Quality, Deferred Taxes,

Management Compensation, and on Industry‐Specific Reporting? Journal of Corporate

Accounting & Finance, 27(6), 43-48.

Sami, H. & Welsh, M.J. (1992). Characteristics of Early and Late Adopters of Pension

Accounting Standard SFAS No. 87. Contemporary Accounting Research.

Schipper, K. (2003). Principles-based accounting standards. Accounting Horizons, 17(1), 61-72.

Securities and Exchange Commission (1999). SEC Staff Accounting Bulletin: No. 101 –

New Revenue Standard ASC606

Revenue Recognition in Financial Statements

https://www.sec.gov/interps/account/sab101.htm

Securities and Exchange Commission (2003). Study Pursuant to Section 108(d) of the

Sarbanes-Oxley Act of 2002 on the Adoption by the United States Financial Reporting

System of a Principles-Based Accounting System. July 2003.

https://www.sec.gov/news/studies/principlesbasedstand.htm

Sherman, H.D. & Young, S. D. (2016). Where Financial Reporting Still Falls Short.

Harvard Business Review. Jul/Aug 2016, Vol. 94 Issue 7/8, p76-84.

Shumsky, T. (2019). Revenue Recognition Compliance Costs Are Higher Than Expected,

Companies Say. The Wall Street Journal. https://www.wsj.com/articles/revenue-

recognition-compliance-costs-are-higher-than-expected-companies-say-11549406561

Simon. D.T. & Costigan, M. L. (1996). Additional Evidence on the Determinants of

Accounting Policy Choice: The Case of Positive Early Adopters of SFAS 96.

Quarterly Journal of Business and Economics. Vol. 35. No. 4, pp.49-63

Statement of Financial Accounting Concepts No. 5 (1984). Recognition and

Measurement in Financial Statements of Business Enterprises. December 1984.

https://www.fasb.org/pdf/con5.pdf

Storey, R. & Storey, S. (1998). The Framework of Financial Accounting Concepts and

Standards. Financial Accounting Series. No. 181-C. January 1998

Streaser, S., Sun, K. J., Zaldivar, I. P., & Zhang, R. (2014). Summary of the new FASB and

IASB revenue recognition standards. Review of Business, 35(1), 7.

Sundvik, D. (2019). The Impact of Principles-based vs Rules-based Accounting Standards on

Reporting Quality and Earnings Management. Journal of Applied Accounting Research.

New Revenue Standard ASC606

Sultanoglu, B. 2017. The Consensus Among US GAAP and IFRS Having Stemmed from

Pressure of the U.S. Software Industry: “Apple Rules”. Muhasebe Bilim Dünyası Dergisi,

19 (1), 34-57.

Trombley, M.A. (1989). Accounting Method Choice in the Software Industry:

Characteristics of Firms Electing Early Adoption of SFAS No. 86. The

Accounting Review. Vol. 64. No. 3. pp. 529-538.

Thorps, D., Dohrer, B., Kushmerick, K., & Lee, T. (2020). Challenges to Consider in Audting

Revenue Recognition. Journal of Accountancy

Tysiac, K. (2014). FASB, IASB Release Historic Revenue Recognition Standard. Journal

of Accountancy. May 28, 2014

https://www.journalofaccountancy.com/news/2014/may/201410215.html

Tysiac, K. (2014). Revenue Recognition: NO Time to Wait. Journal of Accountancy. July

1, 2014. https://www.journalofaccountancy.com/issues/2014/jul/revenue-recognition-

20137809.html

Tzuo, T. (2017). Yes, It's Accounting, But ASC 606 Could Have a Huge Impact On your

Company. https://www.forbes.com/sites/groupthink/2017/08/10/asc-606-is-the-biggest-

business-story-youve-never-heard-how-it-will-affect-your-company/?sh=50420d171d52

Yeaton, K. (2015). A new world of revenue recognition: revenue from contracts with customers.

The CPA Journal, 85(7), 50.

Whitehouse, T. (2019). Earliest Revenue Restatements Fix Adoption Problems. Compliance

Week for the Well-Informed Chief Compliance Officer and Audit Executive.

https://www.complianceweek.com/earliest-revenue-restatements-fix-adoption-

problems/2101.article

New Revenue Standard ASC606

Wilks, J. & Ellsworth, C. (2018). Common ASC 606 Issues: Aerospace & Defense.

RevenueHub. https://www.revenuehub.org/common-asc-606-issues-aerospace-defense-

entities/

Watts, R.L. & Zimmerman, J.L. (1978). Towards a Positive Theory of the Determination of

Accounting Standards. The Accounting Review. Vol. 53, pp.529-538

Watts, R.L. & Zimmerman, J.L. (1990). Positive Accounting Theory: A Ten Year Perspective.

The Accounting Review. Vol. 65, pp.131-156

Zmijewski, M. & Hagerman, R (1981). An Income Strategy Approach to the Positive Theory of

Accounting Standard Setting/Choice. Journal of Accounting And Economics 3. PP. 129-

149

Appendix A. Industry Membership of ASC 606 Adopters**

SIC*

Industry

# of Firms

Frequencies

Agricultural Production Crops

0.22%

Agriculture Production Livestock and Animal

Specialties

0.06%

Agriculture Services

0.06%

Metal Mining

0.33%

Coal Mining

0.28%

Oil and Gas Extraction

3.65%

Mining and quarrying of Nonmetallic Minerals

0.17%

Building Construction General Contractors

0.55%

Heavy construction Other than Building Construction

0.72%

Construction Special Trade Contractors

0.33%

Food and Kindred Products

1.60%

Tobacco Products

0.06%

Textile Mill Products

0.17%

Apparel and Other Finished Products Made from

Fabrics

0.89%

Lumber and Wood Products, except Furniture

0.17%

Furniture and Fixtures

0.44%

Paper and Allied Products

0.44%

Printing, Publishing, and Allied Industries

0.89%

Chemicals and Allied Products

248

13.72%

Petroleum Refining and Related Industries

0.33%

Rubber and Misc. Plastics Products

0.55%

Leasure and Leather Products

0.17%

Stone, Clay, Glass, and Concrete Products

0.17%

Primary Metal Industries

0.77%

Fabricated Metal Products, except Machinery and

Transportation Equip

1.44%

Industrial and Commercial Machinery and Equip

4.81%

Electronic and Other Electrical Equipment, except

Computer Equip

131

7.25%

Transportation Equipment

2.77%

Measuring, Analyzing, and Controlling Instruments

108

5.98%

Misc. Manufacturing Industries

0.44%

Railroad Transportation

0.06%

Local and Suburban Transit

0.06%

Motor Freight Transportation and Warehousing

0.77%

Water Transportation

0.22%

Transportation by Air

0.89%

Pipelines, except Natural Gas

0.55%

Transportation Services

0.33%

New Revenue Standard ASC606

Communications

2.77%

Elecric, Gas, and Sanitary Services

113

6.25%

Wholesale Trade-durable Goods

1.77%

Whoesale Trade-non-durable Goods

1.00%

Building Materials, Hardware, Garden Supply

0.11%

General Merchandise Stores

0.55%

Food Stores

0.22%

Automotive Dealers and Gasoline Service Stations

0.77%

Apparel and Accessory Stores

1.38%

Home furniture, Furnishings, and Equipment Stores

0.66%

Eating and Drinking Places

2.10%

Misc. Retail

2.21%

Hotels, Rooming Houses, Camps, and Other Lodging

Places

0.66%

Personal Services

0.28%

Computer Programming, Data Processing,

Prepackaged Software

363

20.09%

Automotive Repair, services, and Parking

0.17%

Motion Pictures

0.50%

Amusement and Recreation Services

1.55%

Health Services

2.10%

Legal Services

0.06%

Educational Services

0.50%

Engineering, Accounting, Research, Management

Services

1.72%

Nonclassifiable Establishements

0.28%

Total

1807

100.00%

* Standard Industry Classification (SIC): https://www.osha.gov/data/sic-manual

** Identified by "ASU2014-09" in Compustat