Does XBRL disclosure management solution influence earnings release efficiency and earnings management?

Tien-Shih Hsieh (Department of Accounting Dartmouth Charlton College of Business, University of Massachusetts, Dartmouth, Massachusetts, USA)
Zhihong Wang (Graduate School of Management, Clark University, Worcester, Massachusetts, USA)
Mohammad Abdolmohammadi (Department of Accounting, Bentley University, Waltham, Massachusetts, USA)

International Journal of Accounting & Information Management

ISSN: 1834-7649

Article publication date: 4 March 2019

Abstract

Purpose

This study aims to investigate whether eXtensible Business Reporting Language (XBRL) disclosure management solution improves public companies’ earnings release efficiency and mitigates earnings management.

Design/methodology/approach

This study adopts a unique survey data set from the Financial Executives Research Foundation 2013 to identify companies’ XBRL implementation strategies. Earnings release efficiency is measured by earnings announcement time lag. Multiple indicators of both accruals- and real activities-based earnings management are adopted to examine the research hypotheses.

Findings

The authors find that the disclosure management solution (DMS) XBRL implementation is positively associated with earnings release efficiency for companies with good news. The authors also find that DMS implementation strategy is negatively related to accruals-based earnings management, but positively related to real activities-based earnings management measured by abnormal cash flows.

Research limitations/implications

The results of this study can inform regulators, investors and corporate management on how XBRL adoption is associated with corporate financial reporting.

Originality/value

The study contributes to the XBRL literature by providing empirical evidence on how the strategies adopted by companies to implement XBRL may affect the results of XBRL mandatory adoption.

Keywords

Citation

Hsieh, T.-S., Wang, Z. and Abdolmohammadi, M. (2019), "Does XBRL disclosure management solution influence earnings release efficiency and earnings management?", International Journal of Accounting & Information Management, Vol. 27 No. 1, pp. 74-95. https://doi.org/10.1108/IJAIM-06-2017-0079

Publisher

:

Emerald Publishing Limited

Copyright © 2019, Emerald Publishing Limited


1. Introduction

In 2009, the Securities and Exchange Commission (SEC) in the USA issued rule No. 33-9002 (SEC, 2009) on using interactive data to improve financial reporting. This rule mandated public companies to adopt the eXtensible Business Reporting Language (XBRL) to disclose their financial statements to the SEC, and also on their corporate websites. As one of the top ten technologies in the field of accounting (Liu, 2013), XBRL has attracted much attention in accounting research to understand its influences for accounting and auditing professionals. Prior conceptual studies argue that XBRL reporting generates reliable and comparable financial reporting and improves reporting efficiency by generating data in a timely manner (Gunn, 2007; Pinsker, 2003; Pinsker et al., 2005; Roohani et al., 2009; Wu and Vasarhelyi, 2004). However, the complexity of XBRL implementation and limited assurance for XBRL-based financial reports may generate concerns about financial reporting quality (Guragai et al., 2015; Harris and Morsfield, 2012; Janvrin and No, 2012; Müller-Wickop et al., 2013).

Early empirical evidence has documented mixed results regarding the effectiveness of XBRL mandatory filings. Some studies suggest that mandatory XBRL filings may reduce information risk and improve information efficiency (Kim et al., 2012). It may also increase market reactions to earnings surprises (Yen and Wang, 2015), improve analyst forecast accuracy (Liu et al., 2014) and increase market liquidity (Liu et al., 2017). Other studies, however, suggest that XBRL mandate may actually reduce financial statement comparability (Dhole et al., 2015) and enlarge information asymmetry between larger investors and smaller investors (Blankespoor et al., 2014). The effects of XBRL adoption may also vary globally due to the differences in accounting values across different nations (Liu and O’Farrell, 2013).

Liu (2013) reviews the XBRL literature and suggest that XBRL implementation strategies play an important role on how corporate reporting activities could benefit from their XBRL adoption, which is consistent with the theoretical arguments of Garbellotto (2009a, 2009b, 2009c). However, few studies have addressed the effects of XBRL implementation strategies. The purpose of this study is to investigate the effects of XBRL implementation strategies, specifically, disclosure management solution (DMS) versus stand alone solution (SAS), on public companies’ earnings release efficiency and earnings management.

According to Garbellotto (2009a, 2009b, 2009c), DMS and SAS are two dominant strategies for XBRL implementation. Adopting the SAS strategy, companies first generate their traditional financial statements, and then convert them into XBRL-based filings. Theoretically, this strategy does not change companies’ financial reporting system, but may require additional conversion time. DMS enables companies to produce XBRL-based filings in a more integrated manner. It requires companies to apply standardized XBRL taxonomies to the financial data sources (e.g. trial balance or general ledger), thus making XBRL reporting more automatic than SAS. As the Financial Executives Research Foundation’s 2013 (FERF2013)[1] survey suggests, approximately 71 per cent of the financial executives who responded to FERF2013 indicated that their companies were using DMS for XBRL implementation at the time the survey was completed (Sinnett, 2013).

PricewaterhouseCoopers (2012) states that DMS for XBRL implementation could facilitate a more automated report assembly and validation, and provides a better environment for contextual and collaborative report review. These enhancements in the business reporting process reduce manual data entry errors and shorten data entry and the review time, thus, are able to improve the transparency and efficiency of financial reporting. However, companies’ reporting efficiency is subject to managers’ discretion. Companies with bad earnings news may deliberately postpone their earnings announcement to mitigate their negative impacts in the financial markets (Begley and Fischer, 1998; Chai and Tung, 2002). Companies with good earnings news may have a stronger motivation to announce their good news by disclosing their financial information more efficiently. Thus, we expect that public companies with good earnings news are likely to make more efficient earnings announcements when they use DMS for XBRL implementation.

With the improved transparency of financial reporting and control environment generated by DMS for XBRL implementation, we posit that the cost and constraints of accruals-based earnings management activities will increase. Thus, companies using DMS for XBRL implementation are less likely to engage in accruals-based earnings management in the post-DMS stage. Zang (2012) suggests a trade-off between accruals- and real activities-based earnings management when companies present different constraints for earnings management activities. We expect that DMS for XBRL implementation is associated with an increased level of real activities-based earnings management.

We use FERF2013 data set to identify specific XBRL implementation strategies (DMS versus SAS). We then extract financial data from COMPUSTAT for the companies identified in FERF2013. Using earnings announcement time lag as a proxy for earnings release efficiency, we find that DMS for XBRL implementation is negatively associated with earnings release time lag for companies with good earnings news. We also adopt multiple indicators of accruals-based and real activities-based earnings management as proxies for earnings management. Our results show that DMS for XBRL implementation is negatively associated with accruals-based earnings management, but positively associated with real activities-based earnings management.

Our study contributes to XBRL literature in multiple ways. First, our study provides empirical support for a recent SEC announcement of the inline XBRL program (iXBRL)[2], which requires embedding XBRL tags in companies’ traditional reporting documents to further enhance corporate disclosure. This regulation emphasizes the importance of using DMS to ensure the accuracy of XBRL reporting. Second, we note that the US House of Representatives has recently issued H.R.1965, “Small Company Disclosure Simplification Act.” If passed[3], this bill will exempt approximately 60 per cent of public companies with revenues below $250m per year from XBRL reporting (US Congress, 2016). Our results could help these small companies in their decision on whether to voluntarily adopt XBRL reporting, and if so, which implementation strategies to use. Finally, the SEC introduced the accounting quality model (AQM) in 2013 trying to use XBRL tags to detect accounting fraud and improper disclosures (Boyle et al., 2014). The AQM model emphasizes the importance of understanding the effects of XBRL on financial reporting. The results of our study can help managers use XBRL implementation as a tool to mitigate earnings management activities, which could be a response to SEC’s introduction of AQM to discover accounting anomalies.

Section 2 provides a discussion of the literature, leading to our research hypotheses. The research method and results are presented in Sections 3 and 4. Section 5 provides a summary and conclusions from the study.

2. Literature review and hypotheses

2.1 eXtensible business reporting language benefits and implementation strategies

Since its introduction in 1999, XBRL-based financial reporting has been implemented by many public companies in the USA and other countries (Gettler, 2014). Liu (2013) reviews prior XBRL studies and suggest that XBRL may benefit its adopters and other stakeholders in multiple ways. For example, XBRL could improve the quality of financial reports by increasing the reporting efficiency and providing an interactive data format for internal business reporting users (e.g. board of directorsand internal auditors) to analyze data and monitor business activities (Alles and Piechocki, 2010; Cohen, 2009; Roohani et al., 2009). It can also integrate raw financial data into XBRL-enabled applications to process financial data for management purposes, such as preparing managerial reports for strategic decision-making and identifying abnormal transactions (Wu and Vasarhelyi, 2004).

However, empirical results are not fully consistent with the theoretical benefits of XBRL implementation. Some studies suggest positive impacts of XBRL on financial reporting practices. For example, Lai et al. (2015) find that XBRL adoption reduces information asymmetry and this effect is stronger for larger and non-high-technology companies. XBRL may also assist a more efficient dissemination of financial information and increase information efficiency (Kim et al., 2012; Yoon et al., 2011). The adoption of XBRL could increase analyst following and improve analyst forecast accuracy (Liu et al., 2014). This line of research also attempts to distinguish the impact of XBRL adoption on financial markets for firms with different characteristics. For example, Efendi et al. (2014) find that the financial market responds more efficiently to positive earnings news issued by large accelerated filers. Yen and Wang (2015) examine the effects of XBRL for both 10-Q and 10-K filings and find that market reactions are stronger for earnings surprises of smaller companies than larger companies. Hao et al. (2014) find that XBRL adoption is negatively associated with the cost of equity capital, contingent on corporate governance mechanisms. Similarly, Lai et al. (2015) find a reduction of cost of capital after companies’ XBRL adoption, and this effect is stronger for voluntary XBRL adopters.

There are also other empirical studies suggesting the limitations of XBRL implementation. For example, Blankespoor et al. (2014) find that XBRL mandate actually increases information asymmetry for the first years of mandatory adoption because the learning curve encountered by smaller investor’s increases their information extraction costs. Dhole et al. (2015) suggest a decrease in financial statement comparability during the initial years of XBRL mandatory adoption. The authors attribute this finding to the use of company-specific XBRL extension tags. Rao et al. (2013) find that firms with more independent directors, combined CEO/chair of the board position and larger firms are more likely to use more extensions in their XBRL filings. Harris and Morsfield (2012) conclude that XBRL presents the limitations of providing reliable data to help investors and analysts improve their decision-making.

XBRL implementation strategies may play a role in explaining the inconsistent empirical results noted above. Garbellotto (2009a, 2009b, 2009c) suggests that the extent to which XBRL could benefit external and internal users might vary depending on the strategies used by companies to implement XBRL. FERF2013 classifies XBRL implementation strategies into DMS and SAS. The SAS strategy[4] fulfills the basic financial reporting requirements by using a “bolt-on” software or services provided by third parties to prepare XBRL filings without changing their traditional financial reporting processes. This strategy requires additional conversion and review time for companies to file XBRL financial reports. The DMS strategy integrates XBRL into the business reporting processes to facilitate more accurate and efficient financial reports. According to Garbellotto (2009a, 2009b)[5], the DMS integration can “built-in” XBRL at the trial balance level or at the general ledger level. Both levels can generate certain “built-in” benefits and provide more efficient XBRL reporting and a better control environment for DMS adopters, with more benefits being expected at the general ledger level. Thus, adopting DMS for XBRL implementation may generate greater benefits for companies with regard to their reporting practices relative to SAS.

2.2 Disclosure management solution and financial reporting

Wu and Vasarhelyi (2004) suggest that once XBRL is integrated into the reporting process, companies can eliminate manual input errors and automatically update financial reports from the source data through XBRL-enabled applications. Thus, companies could use DMS for XBRL implementation to generate financial reports in a timely manner to facilitate efficient decision-making for stakeholders such as management teams, board members and external users. Contextual review of all relevant topical disclosures within financial reports is also feasible because of the availability of multidimensional data generated by XBRL-embedded reporting systems (Li et al., 2007; PricewaterhouseCoopers, 2012). This practice may enhance internal controls and improve the quality of financial information (Liu, 2013). These statements are consistent with Gray and Miller’s (2009) documentation that advanced XBRL technology (e.g. DMS) could improve the control environment and internal control quality.

DMS services are available from various IT service providers. For example, IBM[6] defines DMS as an automated and collaborative reporting solution that provides companies with merged enterprise data in a controlled and auditable environment. DisclosureNet[7] states that its DMS can help adopters improve the efficiency and accuracy of their corporate financial reporting and reduces risks in both internal and external reporting cycles. SAP[8] states that its DMS can provide customers with a more efficient, accurate, flexible and collaborative disclosure process across various data sources. PricewaterhouseCoopers (2012) introduces DMS using SAP’s XBRL implementation technology and suggests that the use of DMS for XBRL implementation could automate financial reporting assembly and provide automated validation of financial information. Thus, companies that use DMS for XBRL implementation are more likely to have improved financial reporting quality and efficiency than those that use SAS.

2.2.1 Earnings release efficiency.

While improving financial reporting efficiency is one of the primary goals of XBRL implementation, public companies may delay their earnings announcement for various reasons (Sengupta, 2004). For example, companies that have diversified business operations may require excess time to process their financial information. Companies that incur mergers and acquisitions transactions involve time-consuming integration of various accounting systems into one financial reporting system, thus prolonging their reporting process. DMS for XBRL implementation streamlines the reporting process for companies that have diversified business operations, resulting in a more efficient financial reporting process than SAS.

However, prior studies suggest that managers exercise discretion in the timing of earnings disclosure (Begley and Fischer, 1998; Chai and Tung, 2002). For example, companies that report “bad news” may postpone announcing their earnings information to mitigate negative effects on their stock prices. Companies with “good news” may expedite releasing their earnings information to generate a positive response in capital markets (Begley and Fischer, 1998; Chai and Tung, 2002; Haw et al., 2000; Kothari et al., 2009). Although concealing “bad news” may harm managers’ reputation once revealed by stakeholders, Hermalin and Weisbach (2007) theoretically analyze the cost and benefit tradeoffs and argue that managers tend to conceal “bad news”, especially when they have career concerns. Baginski et al. (2016) document empirical evidence that managers, when facing career concerns, tend to withhold bad news with the hope of subsequent good news to bury bad news. To summarize, managers exercise their discretion on the timing of disclosure to control the market response of their earnings announcement. With the assistance of DMS for XBRL implementation, companies with “good news” may accelerate the release of their good news to the public. However, DMS for XBRL implementation may not generate faster “bad news” earnings announcement because of managers’ concerns with the stock market responses and their management career. Based on this discussion, we state H1 as follows:

H1.

Use of DMS for XBRL implementation is positively associated with earnings release efficiency for companies with good news.

2.2.2 Accruals-based earnings management.

The accruals-based accounting system is subject to significant accounting estimations. Previous studies indicate that managers exercise their discretion to engage in earnings management by altering accounting estimates and/or methods to satisfy their self-interests (Degeorge et al., 1999; Merchant, 1990). With DMS for XBRL implementation, the automated reporting assembly, validation and control processes may provide more effective internal controls (PricewaterhouseCoopers, 2012), which may help detect abnormal discretionary accruals during the monitoring process. Using Fujitsu Limited as an example, Roohani et al. (2009) suggest that once XBRL is integrated into companies’ financial reporting system, both trial balance information and internal control information can be effectively evaluated for risk and control purposes. Roohani et al. (2009) argue that this is an effective control mechanism for financial information transparency and financial activities monitoring. As Zang (2012) indicates, accruals-based earnings management may be constrained when the quality of control systems is improved. Thus, the use of DMS for XBRL implementation is likely associated with an enhanced control environment, which in turn reduce abnormal discretionary accruals. This discussion leads to H2:

H2.

Use of DMS for XBRL implementation is negatively associated with accruals-based earnings management.

2.2.3 Real activities-based earnings management.

Real activities-based earnings management refers to the use of specific real activities to manage earnings. Examples include accelerating the timing of sales to increase revenue, expanding production to reduce the cost of inventory and lowering discretionary expenditure to increase earnings (Dechow and Skinner, 2000; Healy and Wahlen, 1999). Cohen et al. (2008) suggest that companies engage in less accruals-based earnings management and more real activities-based earnings management in the post-SOX stage when the control environment is improved. Zang (2012) also document this trade-off relationship and suggest that when constraints and costs of accruals-based earnings management are high, managers could opt for real activities-based earnings management to achieve their earnings goals. She also suggests that real activities-based earnings management might be constrained by firms’ financial health, competition and tax consequences. These factors, however, are less likely to be affected by XBRL implementation in the short time horizon because XBRL implementation is not likely to influence companies’ abilities to sell goods and services to customers or purchasing materials from suppliers. As a trade-off, companies may engage in more real activities-based earnings management when the improved control environment constraints accruals-based earnings management. In addition, DMS for XBRL implementation also helps management gather financial data more efficiently, providing more information on real activities-based earnings management. This argument leads to H3 as follow:

H3.

Use of DMS for XBRL implementation is positively associated with real activities-based earnings management.

3. Research method

3.1 Model specification

3.1.1 Disclosure management solution and earnings release efficiency.

We adopt Model 1 as specified below to examine (H1). We create an indicator variable DMS that equals 1 if a company uses DMS for XBRL implementation in a given year; 0 otherwise. We calculate earnings announcement time lag (LAG) using the natural logarithm of the number of days between fiscal year-end and earnings announcement date to proxy for earnings release efficiency (Brazel and Dang, 2008; Haw et al., 2000). Shorter LAG indicates more efficient earnings release. Following Brazel and Dang (2008), we generate earnings surprise (ESURP), which is the change of earnings per share from the previous year to the current year[9]. A positive (negative) ESURP indicates good (bad) earnings news. We then include an interaction term DMS*ESURP and expect a negative association between DMS*ESURP and LAG.

(1) LAG=α+β1DMS+β2ESURP+β3DMS*ESURP+β4SIZE+β5IBE+β6CFO+β7LEV+β8BIG4+β9LIT+β10XBRL_MFP+β11CG_INDEX+YEAREFFECT+INDUSTRYEFFECT+ε

Appendix summarizes the definitions of the variables included in the model.

Previous studies suggest that larger companies and companies with lower financial performance tend to have longer earnings announcement time lag (Brazel and Dang, 2008; Haw et al., 2000). We include the natural logarithm of the market value of equity (SIZE), income before extraordinary items scaled by total assets (IBE) and cash flows from operating activities (CFO) to control for firm size and financial performance. Following DeFond et al. (2007), we control the leverage ratio (LEV) and expect a negative sign. Auditor size (BIG4) is also included in the model because Big 4 audit firms tend to perform faster audits than other audit firms, which may contribute to more efficient earnings release (Abidin and Ahmad-Zaluki, 2012).

Disclosure efficiency is also affected by firms’ litigation risk (Lee et al., 2008). Therefore, we control firms’ litigation risk (LIT) and expect a negative sign. Kim et al. (2012) find that mandatory XBRL reporting improves the financial information environment and information efficiency. However, Garbellotto (2009a, 2009b) argues that whether adopting XBRL improves financial reporting efficiency may be contingent on the strategies that companies use to implement XBRL. Thus, we include a binary indicator variable XBRL_MFP to control for the effects of mandatory XBRL adoption on reporting efficiency without a predicted sign.

Prior studies imply that companies with strong corporate governance tend to have greater and more efficient disclosure to reduce information asymmetry (Eng and Mak, 2003; Kanagaretnam et al., 2007). Therefore, we control the corporate governance using a composite index (CG_INDEX), which is a linear addition of board size, independent directors and CEO duality. Board size equals 1 when the size of a board is smaller than the industry average; 0 otherwise. Independent directors equals 1 when the percentage of independent directors on a board is larger than the industry average; 0 otherwise. CEO duality equals 1 if a CEO is not the chair of the board; 0 otherwise. We expect that CG_INDEX is negatively associated with LAG. A year and industry fixed effects are also included as control variables in the model, with robust standard errors clustered by firm.

3.1.2 Disclosure management solution strategy and accruals-based earnings management.

We adopt Model 2 as specified below to examine (H2). We adopt the modified Jones model to calculate the absolute value of discretionary accruals (ABDACC) as a proxy for accruals-based earnings management (Dechow et al., 1995). High ABDACC indicates high levels of accruals-based earnings management. We expect a negative association between DMS and ABDACC for H2.

(2) ABDACC=α+β1DMS+β2SIZE+β3LEV+β4GROWTH+β5BMR+β6IBE+β7BIG4+β8LIT+β9XBRL_MFP+β10CG_INDEX+β11NOA+β12CFO+YEAREFFECT+INDUSTRYEFFECT+ε

Variable definitions are summarized in Appendix.

Prior studies suggest that large firms are more likely to attract attention from capital markets and to encounter greater scrutiny than small firms (Lobo and Zhou, 2001; Xie et al., 2003), thus are less likely to engage in accrual-based earnings management activities. We control SIZE and expect to observe a negative association between SIZE and ABDACC. DeFond and Jiambalvo (1994) suggest that companies with high leverage ratios (LEV) are likely to manage their earnings upward to avoid violating debt covenants. However, high LEV may also restrict managers’ opportunistic behavior (Christie and Zimmerman, 1994). Therefore, we include LEV as a control variable without a directional sign. We also include sales growth (GROWTH) and book-to-market ratio (BMR) in Model 2 to control for the impact of companies’ growth on ABDACC. Companies with high growth rate are more likely to engage in earnings management than low growth companies because it is more difficult for financial information users to comprehend high growth companies’ business activities (Park and Shin, 2004). We expect to observe a positive sign for GROWTH, and a negative sign for BMR. Moreover, we include IBE to control for companies’ financial performance and expect a positive sign (Degeorge et al., 1999).

Companies audited by big four firms are more likely to have better financial reporting quality than non-big four firms because of the high audit quality provided by big four firms (Becker et al., 1998; DeAngelo, 1981). Thus, we control BIG4 and expect a negative coefficient. In addition, some studies suggest that companies with high litigation risk are likely to engage in earnings management behavior to avoid lawsuits from shareholders (Scoffer et al., 2000; Matsumoto, 2002). However, other studies find that companies with high litigation risk are less likely to conduct earnings management to avoid earnings surprises (Matsumoto, 2002). Therefore, we include LIT to control for companies’ litigation risk without a predicted sign. We also include XBRL_MFP to control for the effect of mandatory XBRL adoption and expect a negative sign (Kim et al., 2012).

Prior studies also suggest that companies with strong corporate governance are more likely to adopt effective controls to improve their financial reporting quality (Cohen et al., 2004; Klein, 2002). Therefore, we control CG_INDEX and expect to observe a negative sign. We also control for none operating assets (NOA) and expect a positive sign (Barton and Simko, 2002; Gunny, 2005). CFO is also included in the model because literature has shown that cash flows are negatively associated with earnings management activities (Givoly and Hayn, 2000). A year and industry fixed effects are also included in the model, with robust standard errors clustered by firm.

3.1.3 Disclosure management solution strategy and real activities-based earnings management.

Following Roychowdhury (2006) and Cohen et al. (2008), we calculate abnormal cash flow from operations (ABCFO), abnormal production costs (ABPROD) and abnormal discretionary expenses (ABDISEXP) as proxies for companies’ real activities-based earnings management. Below we provide the rationales for these measures.

First, companies can use more price discounts or more generous credit policy to accelerate the timing of sales temporarily. However, these activities will generate abnormally low cash flows for the accounting period. Thus, companies adopt this method for earnings management are likely to have abnormally low cash flows from operations. To capture abnormal cash flows generated from sales, we first use equation REM1 to estimate the normal levels of cash flows. All variables are scaled by total assets at the beginning of the year.

(REM1) CFOit=β1+β2Salesit+β3ΔSalesit+εit

Where

CFOit = Cash flow from operations for firm i in year t;

Salesit = Total sales for firm i in year t; and

ΔSalesit = Change in sales for firm i from year t − 1 to t.

As shown in equation REM2, we calculate abnormal CFO (ABCFO) using actual CFO minus normal CFO calculated from REM1. All items are scaled by total assets at the beginning of the year. A lower level of abnormal CFO (ABCFO) indicates higher level of real activities-based earnings management.

(REM2) ABCFOit=CFOit-(b1+b2Salesit+b3ΔSalesit)

Second, fixed production costs are to be allocated to the total number of units that are produced in the accounting period. Companies can increase the level of production to decrease the cost of sales per unit, resulting in a higher level of gross profit. Thus, companies that increase the levels of production tend to have higher levels of production costs, which is defined as the total cost of goods sold plus the change in inventory during a specific year.

As shown in equation REM3, we express production costs as a linear function of the cost of goods sold, which is a linear function of total sales, and change in total inventory. Change in total inventory is also a linear function of the change in the change of sales. All variables including the intercept are scaled by total assets at the beginning of the year.

(REM3) Prodit=β1+β2Salesit+β3ΔSalesit+β4ΔSalesit-1+εit

Where:

Prodit = Total production costs for firm i at year t;

Salesit = Total sales for firm i at year t;

ΔSalesit = Change in sales for firm i from year t − 1 to t; and

ΔSalesit-1 = Change in sales for firm i from year t−2 to t − 1.

Next, we adopt equation REM4 to calculate abnormal production cost (ABPROD) using actual production cost minus normal production cost calculated from the estimated coefficients in equation REM3. A higher level of abnormal production cost (ABPROD) indicates a higher level of real activities-based earnings management. All items are scaled by total assets at the beginning of the year.

(REM4) ABPRODit=Prod-(b1+b2Salesit+b3ΔSalesit+b4ΔSalesit-1)

Third, we note that companies can decrease discretionary expenses to generate higher earnings for the current accounting period. We express discretionary expenses as a linear function of total sales from past year as in equation REM5.

(REM5) DISEXPit=β1+β2Salesit-1

Where:

DISEXPit = Total discretionary expense from operations for firm i in year t; and

Salesit-1 = Total sales for firm i in year t − 1.

Then we adopt equation REM6 to calculate abnormal discretionary expense (ABDISEXP) using actual discretionary expense minus normal discretionary expense calculated from REM5. A lower level of abnormal discretionary expense (ABDISEXP) indicates higher level of real activities-based earnings management. All items are scaled by total assets at the beginning of the year:

(REM6) ABDISEXPit=DISEXP-(b1+b2Salesit-1)+εit

After we calculate the three indicators, we use them as the dependent variables in Model 3 to examine the effect of DMS for XBRL implementation and real activities-based earnings management. We include the same set of control variables, except litigation risk because real activities-based earnings management is consistent with accounting standards and therefore present limited litigation risks to control for their effects on companies’ earnings management activities.

(3) REAL_EM=α+β1DMS+β2SIZE+β3LEV+β4GROWTH+β5BMR+β6IBE+β7BIG4+β8XBRL_MFP+β9CG_INDEX+β10NOA+β11CFO+YEAREFFECT+INDUSTRYEFFECT+ε

Definitions of the variables in the model are summarized in Appendix.

As H3 suggests, we expect to observe a negative association between DMS and two real activities-based earnings management measures (ABCFO and ABDISEXP) and a positive association between DMS and ABPROD. The year and industry fixed effects are also included in the model, with robust standard errors clustered by firm.

3.2 Sample selection

Our data is extracted from two sources. The first source is FERF2013, which surveys members of Financial Executives International, who are corporate executives and SEC reporting professionals from 442 US companies, about their XBRL implementation experiences. We identify companies that have adopted DMS and the specific adoption year from two survey questions. The first question asks participants whether their companies use DMS for the most recent filings. If the answer is YES, participants then answer the second question, which asks how many years their companies have used DMS. We identify the specific year that a company started to use DMS for XBRL implementation using information provided by both questions. For example, if a company uses DMS for the most recent XBRL-based 10-K filing (i.e. 2012 annual report)[10], and the company has used DMS for two years, we determine the company started to use DMS for XBRL-based 10-K filing in the year 2011[11].

To distinguish the effect of XBRL adoption in general and the specific effect of DMS for XBRL implementation, we select our sample period from 2007 to 2014 to include the two years of XBRL voluntary reporting period. From the 442 companies that participated in FERF2013, we generate 3,094 firm-year observations of XBRL implementation activities. Then we match FERF2013 with the financial data extracted from COMPUSTAT. We exclude firms from regulated industries[12] and firms without a proper identification code for the matching process. This process reduces our sample to 1,350 observations for 191 firms over seven years. We then constitute our final sample by further reducing the missing values of variables in each empirical model.

Table I presents the sample selection process. Specifically, the reporting efficiency model as presented in Model 1 has a final sample of 970 firm-year observations. The accruals-based earnings management model as presented in Model 2 has a final sample of 1,000 firm-year observations. The real activities-based earnings management model as presented in Model 3 has a final sample of 794 when using abnormal cash flows as the dependent variable, 649 when using abnormal production cost as the dependent variable, and 584 when using abnormal discretionary expense as the dependent variable.

4. Empirical results

4.1 Descriptive statistics

Table II presents descriptive statistics of the variables in each empirical model. Panel A of Table II shows that the mean value of LAG is 3.57 for the full sample. Companies have a lower LAG in the post-DMS stage (mean LAG = 3.54) than in the pre-DMS stage (mean LAG = 3.59). As reported in Panel C of Table II, companies have more real activities-based earnings management as measured by ABCFO (mean ABCFO = 0.21) than in the pre-DMS stage (mean ABCFO = 0.06). The descriptive statistics also suggest that large companies (SIZE), companies with better financial performance (IBE) and growing companies (GROWTH and BMR) are more likely to use DMS for XBRL implementation. Also, in the XBRL mandatory filing period (XBRL_MFP), there are more companies use DMS for XBRL implementation, relative to the pre-XBRL mandatory filing period.

4.2 Multivariate analysis for hypothesis testing

4.2.1 Earnings release efficiency.

Table III presents the results for H1. Panel A reports the results of the main effect. We observed a marginally significant and negative effect of DMS (p = 0.087) on LAG. Panel B reports the results of the full model with the interaction term DMS*ESURP. Consistent with the prediction of H1, results show a negative interaction effect of DMS*ESURP (p = 0.005) on LAG, suggesting that the negative association between DMS and LAG is stronger for companies with good news than companies with bad news. Consistent with prior studies, the results also suggest that large companies (SIZE: p < 0.001), companies with high cash flows from operations (CFO: p = 0.022), companies with higher leverage ratios (LEV: p = 0.006) and higher litigation risk (LIT: p < 0.001) are more likely to have shorter reporting lags compared with other companies.

We further partition our sample into two groups based on the value of ESURP, namely: one group includes companies with positive ESURP and another group includes companies with negative ESURP. We perform the same empirical analysis for both groups by adopting Model 1. Panels C and D of Table III present the results for the two groups, respectively. The results suggest that DMS (p = 0.073) is negatively associated with LAG for companies with good earnings news, however, DMS (p = 0.211) is not significantly related to LAG for companies report bad earnings news. This sub-sample analysis suggests that the significant main effect of DMS on LAG for the full sample is driven by the companies with good news, thus provides further support for H1[13].

4.2.2 Accruals-based earnings management.

Table IV presents the results for H2. Consistent with H2, DMS (p = 0.022) is inversely related to the ABDACC, suggesting that the DMS for XBRL implementation is associated with a decrease in accruals-based earnings management. The results also suggest that LIT (p = 0.030) is positively associated with ABDACC. Corporate governance (CG_INDEX: p = 0.046) is negatively associated with ABDACC, indicating that good corporate governance may serve as an effective tool to reduce accruals-based earnings management.

4.2.3 Real activities-based earnings management.

Panels A, B and C of Table V present the empirical results for H3. In Panel A, ABCFO is used as the dependent variable. Consistent with H3, the results suggest that DMS (p = 0.040) is inversely related to ABCFO, indicating that companies tend to have more real activities-based earnings management through accelerating the timing of sales in the post-DMS stage. Panel A also suggests that companies with high growth rate (GROWTH: p = 0.050), high book-to-market ratio (BMR: p = 0.044) and high cash flow from operations (CFO: p = 0.010) are likely to have a high level of abnormal cash flows[14].

Table V Panel B reports the results of H3 where ABPROD is used as the dependent variable. The results do not show a significant association between DMS and ABPROD. Operating cash flow (CFO: p = 0.065) is marginally and negatively associated with ABPROD and book-to-market ratio (BMR: p = 0.053) is marginally and positively associated with ABPROD.

Panel C of Table V reports the results for H3 using ABDISEXP as the dependent variable. The results show that DMS is not significantly associated with ABDISEXP, thus do not provide support for H3. Panel C suggests that BMR (p = 0.043), IBE (p = 0.013) and NOA (p = 0.075) are negatively associated with abnormal discretionary expenditure. Also, companies that are audited by the big four audit firms (BIG4: p = 0.048) and companies report high operating cash flows (CFO: p = 0.017) tend to have higher abnormal discretionary expenditure.

4.3 Robustness tests

To examine the robustness of our results, we restrict our sample to the firms that have used DMS for XBRL implementation and perform a pre- and post-analysis. More specifically, we removed companies that did not use DMS for XBRL implementation from the sample and only kept the observations of the DMS companies up to three years before and three years after their use of DMS for XBRL implementation. We adopt Models 1, 2 and 3 to retest our research hypotheses. The results (untabulated) are consistent with our main findings, suggesting that using DMS for XBRL implementation is associated with decreased earnings release time lag and accruals-based earnings management as measured by the absolute value of discretionary accruals, but increased real activities-based earnings management as measured by abnormal cash flows.

5. Conclusion

Using a proprietary survey data FERF2013, we investigate how DMS for XBRL implementation influences companies’ earnings release efficiency and earnings management. We find that DMS for XBRL implementation is positively associated with reporting efficiency for companies with good news. We also find that DMS for XBRL implementation is inversely associated with accruals-based earnings management activities because of the enhanced control environment in the post-DMS adoption stage. In addition, we find evidence that managers are likely to engage in real activities-based earnings management by accelerating the timing of sales as a trade-off for decreased accrual-based earnings management in the post-DMS stage. However, we do not find support for real activities-based earnings management by expanding production to reduce the cost of goods sold and decreasing discretionary expenses to increase earnings. A possible explanation for this finding could be that it is more difficult to adjust production plans and expenditure spending for earnings management purposes at the end of a fiscal year, making it less likely for companies to adopt these two methods.

Although XBRL implementation strategy could play an important role in realizing the expected benefits of XBRL, empirical evidence about how XBRL implementation strategies actually affect users of internal business reports or companies’ business reporting activities are rare. Prior studies suggest, a theoretical link between XBRL adoption and improved financial reporting quality contingent on how XBRL is implemented into business reporting processes. Our study provides empirical evidence for managers, auditors, investors and other financial statement users to better understand the impact of XBRL implementation strategies on corporate reporting efficiency and quality. We contribute to the accounting information management literature by demonstrating the role of XBRL implementation strategies on reducing different types of earnings management activities and thus help companies improve their reporting efficiency. Our results should also be of interest to regulators who are seeking evidence on the effects of XBRL implementation by public companies to evaluate the effectiveness of the mandatory XBRL adoption and to move forward on promoting certain strategies to implement XBRL reporting. For example, our results may be viewed as empirical support for the SEC’s requirement of mandatory XBRL adoption.

Sample selection process

Reporting
efficiency
(H1)
Accruals-based earnings
management
(H2)
Real activities-based
earnings management
(H3)
DV = LAG ABDACC ABCFO ABPROD ABDISEXP
Firm year observations (# of firms) participating in FERF2013 survey 3,094 (442)
Less: firm year observations (# of firms) in regulated industries or without identification code to match with data from COMPUSTAT 1,350 (191)
Less: Firm year observations (# of firms) with missing value in variables necessary for hypothesis testing 774 (58) 987 (107) 950 (102) 1095 (113) 1160 (141)
Firm year observation (# of firms) for hypothesis testing 970 (193) 757 (144) 794 (149) 649 (138) 584 (110)

Descriptive statistics

Variable name Mean Median STD Mean Median STD Mean Median STD Sig.
Panel A: DV = LAG
Full sample (N = 970) Pre-DMS (N = 489) Post-DMS (N = 481)  
LAG 3.57 3.56 0.34 3.59 3.58 0.33 3.54 3.53 0.34 **
ESURP 0.18 0.02 10.41 0.37 −0.03 14.35 −0.02 0.05 3.07  
SIZE 8.36 8.23 1.60 8.06 7.98 1.50 8.66 8.63 1.63 ***
IBE 0.05 0.05 0.08 0.05 0.05 0.09 0.06 0.06 0.07 ***
CFO 0.11 0.11 0.07 0.11 0.11 0.07 0.11 0.11 0.07  
LEV 0.54 0.56 0.20 0.55 0.57 0.20 0.54 0.55 0.20  
BIG4 0.94 1.00 0.24 0.93 1.00 0.25 0.95 1.00 0.23  
LIT 0.29 0.00 0.45 0.30 0.00 0.46 0.28 0.00 0.45  
XBRL_MFP 0.73 1.00 0.44 0.51 1.00 0.50 0.96 1.00 0.18 ***
CG_INDEX 2.54 3.00 1.01 2.52 3.00 1.04 2.56 3.00 0.98  
Panel B: DV = ABDACC
Full sample (N = 757) Pre-DMS (N = 390) Post-DMS (N = 367)  
ABDACC 0.86 0.10 2.49 0.92 0.10 2.66 0.81 0.10 2.29  
SIZE 8.38 8.26 1.59 8.05 7.91 1.53 8.72 8.74 1.59 ***
LEV 0.52 0.51 0.20 0.52 0.52 0.20 0.51 0.50 0.19  
GROWTH 1.06 1.06 0.14 1.05 1.05 0.16 1.07 1.06 0.13 **
BMR 0.48 0.40 0.42 0.53 0.50 0.03 0.43 0.29 0.02 ***
IBE 0.06 0.06 0.07 0.05 0.06 0.08 0.07 0.07 0.06 ***
BIG4 0.93 1.00 0.26 0.92 1.00 0.27 0.94 1.00 0.24  
LIT 0.35 0.00 0.48 0.36 0.00 0.48 0.35 0.00 0.48  
XBRL_MFP 0.75 1.00 0.44 0.53 1.00 0.50 0.98 1.00 0.15 ***
CG_INDEX 2.56 3.00 0.94 2.52 3.00 0.96 2.60 3.00 0.91  
NOA 17.12 11.37 16.94 16.41 11.55 15.72 17.86 11.35 18.15  
CFO 0.12 0.12 0.07 0.12 0.07 0.00 0.13 0.06 0.00  
Panel C: DV = ABCFO
Full sample (N = 794) Pre-DMS (N = 408) Post-DMS (N = 386)  
ABCFO 0.17 0.06 0.49 0.14 0.06 0.42 0.21 0.07 0.56 *
SIZE 8.41 8.32 1.61 8.08 7.97 1.55 8.76 8.78 1.61 ***
LEV 0.52 0.52 0.20 0.53 0.53 0.20 0.52 0.51 0.19  
GROWTH 1.06 1.06 0.15 1.05 1.05 0.16 1.07 1.06 0.13 ***
BMR 0.47 0.40 0.32 0.51 0.43 0.34 0.42 0.36 0.29 ***
IBE 0.06 0.06 0.07 0.05 0.06 0.08 0.07 0.07 0.06 ***
BIG4 0.93 1.00 0.25 0.92 1.00 0.27 0.94 1.00 0.23  
LIT 0.36 0.00 0.48 0.36 0.00 0.48 0.35 0.00 0.48  
XBRL_MFP 0.75 1.00 0.44 0.53 1.00 0.50 0.98 1.00 0.15 ***
CG_INDEX 2.55 3.00 0.98 2.53 3.00 1.01 2.58 3.00 0.94  
NOA 17.52 11.82 17.25 16.62 11.83 15.61 18.47 11.75 18.80  
CFO 0.12 0.12 0.06 0.12 0.12 0.07 0.13 0.12 0.06  
Panel D: DV = ABPROD
Full sample (N = 649) Pre-DMS (N = 322) Post-DMS (N = 327)  
ABPROD −0.13 −0.10 0.27 −0.12 −0.08 0.28 −0.14 −0.12 0.26  
SIZE 8.44 8.34 1.61 8.07 7.98 1.54 8.80 8.80 1.60 ***
LEV 0.51 0.50 0.20 0.52 0.51 0.20 0.50 0.50 0.20  
GROWTH 1.06 1.06 0.15 1.05 1.05 0.16 1.08 1.06 0.13 ***
BMR 0.45 0.38 0.31 0.50 0.43 0.33 0.40 0.34 0.27 ***
IBE 0.06 0.07 0.07 0.05 0.06 0.08 0.08 0.07 0.06 ***
BIG4 0.92 1.00 0.27 0.91 1.00 0.29 0.94 1.00 0.25  
LIT 0.37 0.00 0.48 0.38 0.00 0.49 0.35 0.00 0.48  
XBRL_MFP 0.76 1.00 0.43 0.54 1.00 0.50 0.98 1.00 0.13 ***
CG_INDEX 2.56 3.00 0.97 2.53 3.00 1.01 2.58 3.00 0.94  
NOA 16.87 11.83 16.09 16.17 12.21 14.69 17.57 11.37 17.35  
CFO 0.12 0.12 0.07 0.12 0.12 0.07 0.13 0.13 0.07 **
Panel E: DV = ABDISEXP
Full sample (N = 584) Pre-DMS (N = 306) Post-DMS (N = 278)  
ABDISEXP 0.05 0.02 0.41 0.03 0.02 0.37 0.08 0.02 0.45  
SIZE 8.54 8.42 1.69 8.16 8.08 1.66 8.95 9.01 1.64 ***
LEV 0.51 0.49 0.21 0.51 0.49 0.21 0.50 0.49 0.20  
GROWTH 1.07 1.06 0.14 1.06 1.05 0.15 1.08 1.06 0.12 *
BMR 0.41 0.35 0.26 0.45 0.39 0.29 0.36 0.31 0.23 ***
IBE 0.07 0.07 0.07 0.06 0.06 0.08 0.08 0.08 0.06 ***
BIG4 0.93 1.00 0.25 0.92 1.00 0.27 0.95 1.00 0.21 *
LIT 0.46 0.00 0.50 0.46 0.00 0.50 0.45 0.00 0.50  
XBRL_MFP 0.75 1.00 0.44 0.53 1.00 0.50 0.98 1.00 0.13 ***
CG_INDEX 2.58 3.00 0.97 2.54 3.00 0.99 2.62 3.00 0.95  
NOA 12.83 10.39 9.62 12.03 10.02 9.07 13.72 10.50 10.15 **
CFO 0.13 0.12 0.07 0.12 0.12 0.07 0.13 0.13 0.06 *

Effects of DMS on earnings release efficiency

DV = LAG   Panel A: main effect Panel B: interaction effect Panel C: good news co. Panel D: bad news co.
Variable Expected sign Coefficient p-value* Coefficient p-value* Coefficient p-value* Coefficient p-value*
DMS −0.046 0.087 −0.048 0.075 −0.052 0.073 −0.035 0.211
ESURP −0.003 0.304 0.01 0.162 0.012 0.363 −0.013 0.064
DMS*ESURP       −0.028 0.005        
SIZE −0.062 <0.001 −0.062 <0.001 −0.065 <0.001 −0.059 0
IBE −0.269 0.106 −0.273 0.1 −0.204 0.209 −0.187 0.248
CFO −0.505 0.024 −0.508 0.022 −0.429 0.072 −0.470 0.065
LEV −0.258 0.006 −0.257 0.006 −0.256 0.008 −0.249 0.019
BIG4 0.048 0.528 0.049 0.259 0.136 0.065 −0.061 0.254
LIT −0.175 <0.001 −0.179 <0.001 −0.164 0.001 −0.188 0.001
XBRL_MFP ? −0.015 0.626 −0.017 0.573 −0.016 0.738 −0.009 0.873
CG_INDEX 0.005 0.784 0.004 0.818 −0.012 0.251 0.019 0.162
Intercept   4.042 <0.001 4.052 <0.001 3.975 <0.001 4.149 <0.001
N 970 970 505 463
Adjusted R2 35.55% 35.93% 38.29% 31.28%
Note:

*The reported p-values are adjusted based on the predicted sign

Effects of DMS on accruals-based earnings management

DV = ABDACC
Variable Expected sign Coefficient p-value*
DMS −0.548 0.022
SIZE 0.064 0.503
LEV ? −0.254 0.610
GROWTH + −0.421 0.672
BMR 0.478 0.173
IBE + 1.571 0.117
BIG4 −0.576 0.173
LIT ? 0.633 0.030
XBRL_MFP 0.212 0.454
CG_INDEX −0.180 0.046
NOA + 0.001 0.423
CFO −1.430 0.246
Intercept 1.237 0.022
N 757
Adjusted R2 7.46%
Note:

*The reported p-values are adjusted based on the predicted sign

Effects of DMS on real activities-based earnings management

Variable Expected sign Coefficient p-value*
Panel A: DV = ABCFO
DMS −0.080 0.040
SIZE 0.011 0.415
LEV ? 0.021 0.801
GROWTH + 0.313 0.050
BMR 0.166 0.044
IBE + 0.223 0.177
BIG4 −0.153 0.108
XBRL_MFP −0.024 0.341
CG_INDEX 0.006 0.723
NOA + 0.000 0.481
CFO 1.084 0.010
Intercept 0.010 0.483
N 794
Adjusted R2 7.86%
Panel B: DV=ABPROD
DMS + −0.012 0.726
SIZE 0.010 0.454
LEV ? 0.116 0.192
GROWTH + 0.079 0.187
BMR 0.141 0.053
IBE + −0.281 0.225
BIG4 −0.064 0.190
XBRL_MFP −0.014 0.361
CG_INDEX −0.014 0.187
NOA + 0.001 0.356
CFO + −0.451 0.065
Intercept −0.440 0.028
N 649
Adjusted R2 22.14%
Panel C: DV = ABDISEXP
DMS −0.044 0.183
SIZE −0.008 0.327
LEV ? −0.108 0.335
GROWTH + 0.147 0.144
BMR −0.142 0.043
IBE + −0.917 0.013
BIG4 0.268 0.048
XBRL_MFP 0.039 0.498
CG_INDEX 0.010 0.654
NOA + −0.004 0.075
CFO + 0.834 0.017
Intercept −0.335 0.305
N 584
Adjusted R2 13.64%
Note:

*The reported p-values are adjusted based on the predicted sign

Notes

1.

The purpose of FERF2013 is to understand companies XBRL implementation experiences, including XBRL strategies, concerns and difficulties, future plans, etc., from their financial executives’ perspectives.

2.

In June 13, 2016, the SEC announced a voluntary adoption program that allows public companies to use iXBRLfor their financial disclosure. A copy of this SEC press release can be found at: www.sec.gov/rules/exorders/2016/34-78041.pdf

3.

The status of this bill, as of January 28, 2016, is being reported to house without amendment and on the union calendar. Bill information can be found at: www.congress.gov/bill/114th-congress/house-bill/1965/all-info

4.

Garbellotto (2009a) terms SAS as the bolt-on strategy.

5.

Garbellotto (2009b) states that many companies have “built-in” XBRL at the general ledger level. Two main cases are Wacoal and Fujitsu. Due to the restriction of our data source, we do not distinguish DMS at these two levels and only discuss the general features of DMS.

9.

To avoid the confusions of “good” versus “bad” earnings news created by negative EPS, we also remove observations with negative EPS from our regression analysis. The results are consistent with the full sample analysis.

10.

As FERF2013 survey was completed in 2013, we determine the most recent filing for the companies in the survey is 2012.

11.

FERF2013 provides four categories of the number of years that a company uses DMS for XBRL implementation. The four categories are (1) Less than one year: we define the starting year of use DMS as 2012; (2) one-two years: we define the starting year of use DMS as 2011; (3) two-three years: we define the starting year of use DMS as 2010; and (4) more than three years: we define the starting year of use DMS as 2009.

12.

Following Matsumoto (2002), we defined regulated industries as financial institutions, utility companies and other quasi-regulated industries. Matsumoto (2002) provides the 4-digit SIC code for the regulated industries.

13.

Unexpected result of BIG4 could be related to the data partition because we split our sample into 1) good news companies and 2) bad news companies, two groups. The sample could differ from prior studies that use random sample collected from commercial databases such as COMPUSTAT and CRSP.

14.

Unexpected results in some control variables could be related to the data selection processes. Our study relies on the FERF2013 survey database, which limit our sample to the survey participants. Thus, the sample could differ from prior studies that use random sample collected from commercial databases.

Appendix. Variable definitions

Dependent variables:

LAG = The natural logarithm of the number of days between fiscal year end and earnings announcement date;

ABDACC = Absolute value of discretionary accruals;

REAL_EM = Real activities-based earnings management measured by;

ABCFO = abnormal levels of cash flow from operations;

ABPROD = abnormal levels of production costs; and

ABDISEXP = abnormal levels of discretionary expenses.

Independent variables:

DMS = 1 if the company implement XBRL using DMS strategy in the year; 0 otherwise;

ESURP = Difference between earnings per share (EPS) in year t and year t−1 scaled by EPS in year t−1;

SIZE = the natural logarithm of the market value of equity;

IBE = income before extraordinary items divided by total asset;

CFO = cash flows from operating activities scaled by total assets;

LEV = total liabilities scaled by total assets;

BIG4 = 1 if the company is audited by one of big four accounting firms; else BIG4 = 0;

LIT = 1 if the company is in a high litigation risk industry (SIC codes 2833-2836, 3570-3577, 7370-7374, 3600-3674, 5200-5961); 0 otherwise;

XBRL_MFP = 1 if the firm-year observation is in XBRL mandatory filing period; 0 otherwise;

CG_INDEX = linear addition of board size, independent directors and CEO duality;

GROWTH = Sales in period t divided by sales in period t−1;

BMR = book-to-market ratio; and

NOA= net operating assets calculated by PPE + total current assets − total current liabilities scaled by common shares outstanding multiplied by adjustment factor.

References

Abidin, S. and Ahmad-Zaluki, N.A. (2012), “Auditor industry specialism and reporting timeliness”, Procedia-Social and Behavioral Sciences, Vol. 65, pp. 873-878.

Alles, M. and Piechocki, M. (2010), “Will XBRL improve corporate governance? A framework for enhancing governance decision making using interactive data”, International Journal of Accounting Information Systems, Vol. 13 No. 2, pp. 91-108.

Baginski, S.P., Campbell, J.L., Hinson, L.A. and Koo, D.S. (2016), “Do career concerns affect the delay of bad news disclosure?”, The Accounting Review, Forthcoming.

Barton, J. and Simko, P.J. (2002), “The balance sheet as an earnings management constraint”, The Accounting Review, Vol. 77 No. s-1, pp. 1-27.

Becker, C.L., DeFond, M.L., Jiambalvo, J. and Subramanyam, K.R. (1998), “The effect of audit quality on earnings management”, Contemporary Accounting Research, Vol. 15 No. 1, pp. 1-24.

Begley, J. and Fischer, P.E. (1998), “Is there information in an earnings announcement delay?”, Review of Accounting Studies, Vol. 3 No. 4, pp. 347-363.

Blankespoor, E., Miller, B.P. and White, H.D. (2014), “Initial evidence on the market impact of the XBRL mandate”, Review of Accounting Studies, Vol. 19 No. 4, pp. 1468-1503.

Boyle, D.M., Boyle, J.F. and Carpenter, B.W. (2014), “The SEC’s renewed focus on accounting fraud”, The CPA Journal, Vol. 84 No. 2, pp. 68-72.

Brazel, J.F. and Dang, L. (2008), “The effect of ERP system implementations on the management of earnings and earnings release dates”, Journal of Information Systems, Vol. 22 No. 2, pp. 1-21.

Chai, M.L. and Tung, S. (2002), “The effect of earnings–announcement timing on earnings management”, Journal of Business Finance and Accounting, Vol. 29 Nos 9/10, pp. 1337-1354.

Christie, A.A. and Zimmerman, J.L. (1994), “Efficient and opportunistic choices of accounting procedures: Corporate control”, The Accounting Review, Vol. 69 No. 4, pp. 53-566.

Cohen, E.E. (2009), “XBRL’s global ledger framework: exploring the standardized missing link to ERP integration”, International Journal of Disclosure and Governance, Vol. 6 No. 3, pp. 188-206.

Cohen, D.A., Dey, A. and Lys, T.Z. (2008), “Real and accrual-based earnings management in the pre-and post-Sarbanes-Oxley periods”, The Accounting Review, Vol. 83 No. 3, pp. 757-787.

Cohen, J., Krishnamoorthy, G. and Arnie, W. (2004), “The corporate governance mosaic and financial reporting quality”, Journal of Accounting Literature, Vol. 23, pp. 87-152.

DeAngelo, L. (1981), “Auditor size and audit quality”, Journal of Accounting and Economics, Vol. 3, Vol. 3 No. 3, pp. 183-199.

Dechow, P.M. and Skinner, D.J. (2000), “Earnings management: Reconciling the views of accounting academics, practitioners, and regulators”, Accounting Horizons, Vol. 14 No. 2, pp. 235-250.

Dechow, P.M., Sloan, R.G. and Sweeney, A.P. (1995), “Detecting earnings management”, The Accounting Review, Vol. 70 No. 2, pp. 193-225.

DeFond, Μ.L. and Jiambalvo, J. (1994), “Debt covenant violations and manipulation of accruals”, Journal of Accounting and Economics, Vol. 17 Nos 1/2, pp. 145-176.

DeFond, M., Hung, M. and Trezevant, R. (2007), “Investor protection and the information content of annual earnings announcements: International evidence”, Journal of Accounting and Economics, Vol. 43 No. 1, pp. 37-67.

Degeorge, F., Patel, J. and Zeckhauser, R. (1999), “Earnings management to exceed thresholds”, The Journal of Business, Vol. 72 No. 1, pp. 1-33.

Dhole, S., Lobo, G.J., Mishra, S. and Pal, A.M. (2015), “Effects of the SEC’s XBRL mandate on financial reporting comparability”, International Journal of Accounting Information Systems, Vol. 19, pp. 29-44.

Efendi, J., Park, J.D. and Smith, L.M. (2014), “Do XBRL filings enhance informational efficiency? Early evidence from post-earnings announcement drift”, Journal of Business Research, Vol. 67 No. 6, pp. 1099-1105.

Eng, L.L. and Mak, Y.T. (2003), “Corporate governance and voluntary disclosure”, Journal of Accounting and Public Policy, Vol. 22 No. 4, pp. 325-345.

Garbellotto, G. (2009a), “XBRL implementation strategies: the bolt-on approach”, Strategic. Finance, Vol. May.

Garbellotto, G. (2009b), “XBRL implementation strategies: the built-in approach”, Strategic Finance, Vol. August.

Garbellotto, G. (2009c), “XBRL implementation strategies: the deeply embedded approach”, Strategic Finance, Vol. November.

Gettler, L. (2014), “XBRL, making the world turn faster, institute of chartered accountants of Australia”, available at: www.charteredaccountants.com.au/News-Media/Charter/Charter-articles/Reporting/2011-11-XBRL-Making-the-World-Turn-Faster.aspx

Givoly, D. and Hayn, C. (2000), “The changing time-series properties of earnings, cash flows and accruals: Has financial reporting become more conservative?”, Journal of Accounting and Economics, Vol. 29 No. 3, pp. 287-320.

Gray, G.L. and Miller, D.W. (2009), “XBRL: Solving real-world problems”, International Journal of Disclosure and Governance, Vol. 6 No. 3, pp. 207-223.

Gunn, J. (2007), “XBRL: Opportunities and challenges in enhancing financial reporting and assurance processes”, Current Issues in Auditing, Vol. 1 No. 1, pp. 36-43.

Gunny, K.A. (2005), “What are the consequences of real earnings management?”, Ph.D. dissertation. University of California, Berkeley. available at: http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.463.841&rep=rep1&type=pdf

Guragai, B., Hunt, N. and Neri, M.P. (2015), “Accounting information systems and ethics research: Review, syntheses, and the future”, Journal of Information Systems, Vol. 31 No. 2, pp. 65-81.

Hao, L., Zhang, J.H. and Fang, J. (2014), “Does voluntary adoption of XBRL reduce cost of equity Capital?”, International Journal of Accounting and Information Management, Vol. 22 No. 2, pp. 86-102.

Harris, T.S. and Morsfield, S.G. (2012), “An evaluation of the current state and future of XBRL and interactive data for investors and analysts”, available at: http://hdl.handle.net/10022/AC:P:20283

Haw, I.M., Qi, D. and Wu, W. (2000), “Timeliness of annual report releases and market reaction to earnings announcements in an emerging Capital market: The case of China”, Journal of International Financial Management and Accounting, Vol. 11 No. 2, pp. 108-131.

Healy, P.M. and Wahlen, J.M. (1999), “A review of the earnings management literature and its implications for standard setting”, Accounting Horizons, Vol. 13 No. 4, pp. 365-383.

Hermalin, B.E. and Weisbach, M.S. (2007), “Transparency and corporate governance”, Working Paper, University of CA at Berkeley.

Janvrin, D.J. and No, W.G. (2012), “XBRL implementation: A field investigation to identify research opportunities”, Journal of Information Systems, Vol. 26 No. 1, pp. 169-197.

Kanagaretnam, K., Lobo, G.J. and Whalen, D.J. (2007), “Does good corporate governance reduce information asymmetry around quarterly earnings announcements?”, Journal of Accounting and Public Policy, Vol. 26 No. 4, pp. 497-522.

Kim, J.W., Lim, J.H. and No, W.G. (2012), “The effect of first wave mandatory XBRL reporting across the financial information environment”, Journal of Information Systems, Vol. 26 No. 1, pp. 127-153.

Klein, A. (2002), “Audit committee, board of director characteristics, and earnings management”, Journal of Accounting and Economics, Vol. 33 No. 3, pp. 375-400.

Kothari, S.P., Shu, S. and Wysocki, P.D. (2009), “Do managers withhold bad news?”, Journal of Accounting Research, Vol. 47 No. 1, pp. 241-276.

Lai, S.C., Lin, Y.S., Lin, Y.H. and Huang, H.W. (2015), “XBRL adoption and cost of debt”, International Journal of Accounting and Information Management, Vol. 23 No. 2, pp. 199-216.

Lee, H.Y., Mande, V. and Son, M. (2008), “A comparison of reporting lags of multinational and domestic firms”, Journal of International Financial Management and Accounting, Vol. 19 No. 1, pp. 28-56.

Li, Y., Roge, J.N., Rydl, L. and Hughes, J. (2007), “Achieving Sarbanes-Oxley compliance with XBRL-based ERP and continuous auditing”, Issues in Information Systems, Vol. 8 No. 2, pp. 430-436.

Liu, C. (2013), “XBRL: a new”, Journal of Global Information Management, Vol. 21 No. 3, pp. 60-80.

Liu, C., Luo, X.R. and Wang, F.L. (2017), “An empirical investigation on the impact of XBRL adoption on information asymmetry: Evidence from Europe”, Decision Support Systems, Vol. 93, pp. 42-50.

Liu, C. and O’Farrell, G. (2013), “The role of accounting values in the relation between XBRL and forecast accuracy”, International Journal of Accounting and Information Management, Vol. 21 No. 4, pp. 297-313.

Liu, C., Wang, T. and Yao, L.J. (2014), “XBRL’s impact on analyst forecast behavior: An empirical study”, Journal of Accounting and Public Policy, Vol. 33 No. 1, pp. 69-82.

Lobo, G.J. and Zhou, J. (2001), “Disclosure quality and earnings management”, Asia-Pacific Journal of Accounting and Economics, Vol. 8 No. 1, pp. 1-20.

Matsumoto, D.A. (2002), “Management’s incentives to avoid negative earnings surprises”, The Accounting Review, Vol. 77 No. 3, pp. 483-514.

Merchant, K.A. (1990), “The effects of financial controls on data manipulation and management myopia”, Accounting, Organizations and Society, Vol. 15 No. 4, pp. 297-313.

Müller-Wickop, N., Schultz, M. and Nüttgens, M. (2013), “XBRL: impacts, issues and future research directions”, In Enterprise Applications and Services in the Finance Industry, Springer Berlin Heidelberg, pp. 112-130.

Park, Y.W. and Shin, H.H. (2004), “Board composition and earnings management in Canada”, Journal of Corporate Finance, Vol. 10 No. 3, pp. 431-457.

Pinsker, R. (2003), “XBRL awareness in auditing: a sleeping giant?”, Managerial Auditing Journal, Vol. 18 No. 9, pp. 732-736.

Pinsker, R., Gara, S. and Karim, K. (2005), “XBRL usage: A socio-economic perspective”, Review of Business Information Systems, Vol. 9 No. 4, pp. 59-72.

PricewaterhouseCoopers (2012), “Disclosure management: Streamlining the last mile”, available at: www.pwc.com/gx/en/xbrl/pdf/pwc-streamlining-last-mile-report.pdf

Rao, Y., Guo, K. and Hou, J. (2013), “Who extends the extensible? The effects of corporate governance on XBRL taxonomy extensions in China”, International Journal of Accounting and Information Management, Vol. 21 No. 2, pp. 133-147.

Roohani, S., Furusho, Y. and Koizumi, M. (2009), “XBRL: Improving transparency and monitoring functions of corporate governance”, International Journal of Disclosure and Governance, Vol. 6 No. 4, pp. 355-369.

Roychowdhury, S. (2006), “Earnings management through real activities manipulation”, Journal of Accounting and Economics, Vol. 42 No. 3, pp. 335-370.

Scoffer, L.C., Thiagarajan, S.R. and Walther, B.R. (2000), “Earnings preannouncement strategies”, Review of Accounting Studies, Vol. 5 No. 1, pp. 5-26.

Securities and Exchange Commission (2009), “Interactive data to improve financial reporting”, available at: www.sec.gov/rules/final/2009/33-9002.pdf

Sengupta, P. (2004), “Disclosure timing: determinants of quarterly earnings release dates”, Journal of Accounting and Public Policy, Vol. 23 No. 6, pp. 457-482.

Sinnett, W. (2013), “SEC reporting and the impact of XBRL: 2013 survey”, Financial Executives Research Foundation, Morristown, NJ, available at: www.financialexecutives.org/ferf/download/2013%20Final/2013-022.pdf

Congress, U.S. (2016), “H.R.1912 - small company disclosure simplification act”, available at: www.congress.gov/bill/114th-congress/house-bill/1965/all-info

Wu, J. and Vasarhelyi, M. (2004), “XBRL: a new tool for electronic financial reporting”, in Business Intelligence Techniques, Springer Berlin Heidelberg, pp. 73-92.

Xie, B., Davidson, W.N. and DaDalt, P.J. (2003), “Earnings management and corporate governance: The role of the board and the audit committee”, Journal of Corporate Finance, Vol. 9 No. 3, pp. 295-316.

Yen, J.C. and Wang, T. (2015), “The association between XBRL adoption and market reactions to earnings surprises”, Journal of Information Systems, Vol. 29 No. 3, pp. 51-71.

Yoon, H., Zo, H. and Ciganek, A.P. (2011), “Does XBRL adoption reduce information asymmetry?”, Journal of Business Research, Vol. 64 No. 2, pp. 157-163.

Zang, A. (2012), “Evidence on the trade-Off between real activities manipulation and accrual-based earnings management”, The Accounting Review, Vol. 87 No. 2, pp. 675-703.

Acknowledgements

The Financial Executive Research Foundation (FERF) generously granted access to its 2013 survey database for this study, for which we are grateful. We thank the anonymous referees and the Editor in Chief Maggie Liu for their valuable comments and suggestions. The paper has also benefitted from comments during presentations at the 2016 Canadian Academic Accounting Association Annual Conference and 2016 American Accounting Association Annual Meeting.

Corresponding author

Zhihong Wang can be contacted at: zhihwang@clarku.edu