Discussion: How XBRL Affects the Cost of Equity Capital? Evidence from Emerging Market S. Chen, W. Li, and D. Wu Beijing Institute of Technology

Transcription

1 Discussion: How XBRL Affects the Cost of Equity Capital? Evidence from Emerging Market S. Chen, W. Li, and D. Wu Beijing Institute of Technology By Samir Trabelsi, Ph.D., CGA

2 Summary of the paper How XBRL affects directly (information asymmetry) and indirectly (financial reporting quality) the cost of equity? The mechanism through which earnings quality affects the cost of capital is still controversial. Bhattacharya, Ecker, Olsson, and Schipper (TAR-2011) do not identify one independent path linking measures of earnings quality and the cost of equity. Previous literature documents that XBRL adoption results in a significant reduction in firms cost of equity (Li, Lin, and Ni, WP-2012) and that the adoption of XBRL leads to the reduction of information asymmetry (Yoon, Zo, and Ciganek, JBR-2011).

3 Summary of the paper (Cont d) H1: XBRL adoption reduces CEC. XBRL provides investors with an effective way to prepare, publish, exchange, and analyze financial information (Hodge et al., TAR-2004) Li et al. (2012) finds that XBRL adoption results in a significant reduction in CEC by reducing investors information processing costs. H2: High CG leads to greater reduction in CEC after XBRL adoption. XBRL creates a natural linkage between FRQ, external investors, and CEC. XBRL is an engine connecting good governance with CEC H3: State ownership has a nonlinear relation with CEC and there is a significant difference in CEC reduction between SOEs and NSOEs after XBRL implementation. SOEs could benefit more from XBRL use and, to some extent, decrease their CEC more, than NSOEs. Under this invisible government guarantee, outside investors may not absolutely require a higher risk premium and thus a higher CEC for SOEs.

4 Major findings Results provide support for H1, that XBRL adoption decreases CEC. From an economic perspective, the estimated coefficients 18 translate into a range of basis point reductions, from 172 to 600 basis points in CEC, after firms initiated XBRL filings. Firms with better CG, XBRL adoption resulted in a 68 basis-point reduction in CEC. XBRL adoption results in a significantly greater basis point reduction (121 basis points) in CEC for SOEs than for NSOEs (108 basis point reduction), suggesting that the effect of XBRL on CEC is greater for SOEs than for NSOEs. No evidence that XBRL increases IA in China in the short-term period as it does in the US market. - Hong Kong, June 5-7, 2014

5 Very interesting paper Some promising findings Comments But this is a challenging area, and some fine-tuning to the paper would improve its impact in this currently uncrowded research area! Having said the above, I preferred the earlier version of the paper where the authors aimed to extend Bhattacharya, Ecker, Olsson, and Schipper (TAR-2011) path analysis by including XBRL.

6 Comments (Cont d) Motivation of the paper Hoogervorst, IASB Chairman, noted that Despite the potentials of the technology, certain aspect of financial reporting appear to be stuck in the offline, pre-pc world...welcomed to the 21 st century. XBRL has the potential to supplement the one-sizefits-all approach of today s financial statements with an a la carte menu of financial information XBRL is a computer language that defines data with tags using predefined standard definitions. These tags classify and define financial information in accordance with U.S. GAAP and IFRS. The tagged financial statements would then be used to create an interactive data file that can be viewed using computer software. According to the SEC, XBRL will facilitate comparison of financial and business performance across companies, reporting periods and across industries. Like 10K and 10Q filings preceded by previous earnings announcements, XBRL filings do not provide new information about firm performance. Rather, the potential usefulness of XBRL filings arise from the potential to enhance investors ability to analyze and process previously disclosed accounting information.

7 Comments (Cont d) Motivation of the paper (Cont d) Recently, there has been complaints by several Fortune 500 companies that XBRL has not proven useful and have advocated for the SEC to scale implementation of XBRL tagging requirements. Complaints by corporate controllers from General Motors, Best Buy, Johnson & Johnson and McCormick and Co. that XBRL is not something their companies finds useful. These controllers claim they are not seeing any benefits from the disclosure requirement and doubt whether investors find it useful as well Wall Street Journal November 13, 2012, 3:19 PM ET. Harris and Morsfield (2012) concluded that "XBRL has succeeded in so far as the objective of providing users with free, interactively-available numerical data from portions of published financial statements and footnotes, as soon as they are filed with the SEC." Such doubts about the usefulness of XBRL filings may provide the motivation for this study to investigate how XBRL filings map into CofC.

8 Comments (Cont d) Hypothesis development H1: XBRL adoption reduces CEC. H2: High CG leads to greater reduction in CEC after XBRL adoption. H3: State ownership has a nonlinear relation with CEC and there is a significant difference in CEC reduction between SOEs and NSOEs after XBRL implementation. It s not clear from a theoretical perspective how XBRL adoption lower CofC: Direct link: via the average precision of investors assessment of firms future cash flows (Lambert et al. 2008). Indirect link: via information composition, information dissemination, and the precision of private and public information (Botosan and Plumlee, 2008).

9 Comments (Cont d) Sample selection Timing of XBRL adoptions. I am not sure that the end of 2008 is the effective date of adoption for all firms. The discussion in page 6 implies that many firms adopted XBRL starting from 2004 (either as a requirement from stock exchanges or voluntarily). Huge variation in the total number of observation depending on which analysis is performed. Research design The first CEC model was not developed by Dhaliwal et al. (2005). This is the model developed by Ohlson and Juettner-Nauroth (2005) and implemented by Gode and Mohanram (2003). Rf-3%. The 3% is intended to capture the inflation rate. I am not sure 3% is appropriate for China. They can use the other two models of CEC (Gebhardt et al., 2001; Easton, 2004) JIAR Conference - Hong Kong, June 5-7, 2014

10 Comments (Cont d) Research design (cont d) The use of analyst forecasts to estimate the CEC creates many complications which must be addressed, e.g., systematic optimism, sluggishness, etc. The authors also need to test the robustness of their finding with respect to the growth assumptions in the CEC models. XBRL is one aspect of earnings quality. Other aspects include accrual quality, persistence, smoothness, etc. (i.e., all the measures used in the FRQ analysis). All these proxies could be substitutes or complements. So one needs to include all FRQ variables as controls in the CEC regressions. XBRL needs to have an incremental impact, over and above these proxies (see Francis et al., 2004). In addition to the FRQ aspects examined in the paper, Feng Li has a few papers (Lehavy, LI, and Merkley, TAR-2011) that measure disclosure quality using annual report readability. If we agree that XBRL is about facilitating readability, right, so controlling for annual report readability may also be important JIAR Conference - Hong Kong, June 5-7, 2014

11 Comments (Cont d) Econometric implementation Econometric models estimation do not control for firms fixed effect? Statistical inference should be based on standard errors clustered by firm and year, as derived by Cameron et al. (2006) et Thomson (2009) and applied by Petersen (2009) and Gow et al. (2010). As the number of firms vary between 2005 to 2011, it might be useful to run the models year by year JIAR Conference - Hong Kong, June 5-7, 2014

12 Comments (Cont d) Results The reason for the difference between our results and Liu et al. s (2014) and their similarity to Chen and Li s (2013), all using data from Chinese firms, may be the use of different starting points for XBRL implementation. (page 18). How Korea is similar to China? Specification problems? Sensitivity of the results to alternative proxies of earnings quality, and cost of equity? 2014 JIAR Conference - Hong Kong, June 5-7, 2014

13 How XBRL Affects Cost of equity capital? To the best of my understanding, Cheng, Lu, and Wu answered the question when rather than the question how. How XBRL maps into the cost of equity remains an unanswered empirical research question: Once can use Bhattacharya, Ecker, Olsson, and Schipper (TAR-2011) and investigate whether XBRL adoption strengthen the indirect path linking earnings quality to cost of equity via information asymmetry. One also may build upon Lehavy, Li, and Merkley (TAR-2011), who document that analysts following, the amount of effort incurred to generate their report, and the informativeness of their report are greater for firms with less readable 10K, by examining how XBRL filings are beneficial for firms with complex annual report. Once can also look to whether XBRL filings improves the information environments.

14 Conclusion Great paper I did really enjoy working on it. Honestly, I liked the first version. It made me read many papers on path analysis in particular Bhattacharya, Ecker, Olsson, and Schipper (TAR-2011).

15 Remember that The Most Important Is: HAVE FUN WITH BIG BOY YASIN IN HONG KONG