![Sample our Economics, Finance,Business & Industry journals, sign in here to start your access, latest two full volumes FREE to you for 14 days]({"type":"image" , "src" : "/sda/5931/TOC-Subject-Sample-2021-Economics-Finance-Business-Industry.jpg"})
Abstract
We examine whether China’s industry-information disclosure guidelines (IIDGs) improve firms’ stock price informativeness and the quality of their textual disclosure. Employing a stacked difference-in-differences design, we find that applicable firms respond to IIDG implementation by including additional IIDG-related and firm-specific information in the Management Discussion and Analysis (MD&A). Our results also show that the stock market reacts to the enhanced disclosure and incorporates more firm-specific (rather than market-wide or industry-level) information into the stock price, as evidenced by a decrease in stock return synchronicity. Further analyses show that firms with fewer increases in textual disclosure following IIDG implementation are more likely to receive disclosure-related comment letters. Additionally, the extent to which firms change their textual disclosure varies with the industry concentration and firms’ profit margins due to the proprietary costs of disclosure.
Acknowledgments
We thank Jeffrey Ng (editor) and the anonymous reviewers for their constructive comments and suggestions. Additionally, our gratitude extends to Hongqi Yuan, Joseph Zhang, and Haoqiang Wei, as well as seminar participants at Wuhan University and Shanghai University of International Business and Economics. All authors contributed equally to this work.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Supplemental Materials
Supplemental materials for this article can be accessed online at https://doi.org/10.1080/09638180.2024.2370860.
Notes
1 See http://www.sse.com.cn/disclosure/credibility/dynamic/c/c_20160302_4052889.shtml (in Chinese), http://www.sse.com.cn/disclosure/credibility/supervision/dynamic/c/c_20160524_4118135.shtml (in Chinese), http://www.sse.com.cn/aboutus/mediacenter/hotandd/c/c_20161202_4209904.shtml (in Chinese), http://www.szse.cn/aboutus/trends/news/t20150226_518573.html (in Chinese), and http://www.szse.cn/aboutus/trends/news/t20191027_571575.html (in Chinese).
2 Online Appendix Table I presents a list of all IIDGs and their effective implementation dates. Online Appendix Table II lists the keywords for our four sample IIDGs and the corresponding firms’ sample MD&A disclosure changes.
3 For example, firms with high proprietary costs may be cautious in their disclosure practices and more reluctant to fully comply with the IIDGs.
4 We thank the anonymous reviewers for pointing this out.
5 In the early 1990s, capital markets were introduced in China with the establishment of the Shanghai Stock Exchange in December 1990 and the Shenzhen Stock Exchange in May 1991. Subsequently, the Shenzhen Stock Exchange established the small and medium enterprise (SME) board in 2004 and the ChiNext board in 2009, which respectively support financing channels for small and middle-sized firms and firms in emerging industries. In 2021, the SME Board was merged into the Main Board of the Shenzhen Stock Exchange.
6 Of the 28 IIDGs issued by the Shanghai Stock Exchange, 27 are industry-specific IIDGs and one is titled ‘General Principles,’ which specifies the common principles that apply to all firms that must comply with at least one industry-specific IIDG. Given our focus on the information disclosure associated with specific industries, we do not consider the General Principles in the empirical tests.
7 The disclosure requirement is intended to enhance the usefulness of the information that listed firms provide. However, exceptional circumstances or legitimate reasons may prevent a firm from fully complying with its disclosure obligations.
8 Cho and Muslu (Citation2021) state that MD&As represent a rich public depository of corporate narratives, as regulators require that MD&As provide managerial commentary about trends and events that are expected to materially affect the capital market. However, we note that some IIDG-related information may also be disclosed in other portions of firms’ annual reports that are narrative in nature, such as financial footnotes. By comparing the changes to MD&As and annual reports, as well as their similarity with IIDGs, we find that MD&As are more likely than annual reports to change around IIDG implementation. This finding highlights the importance of the MD&A section as the primary avenue through which firms can comply with industry-specific disclosure requirements. Moreover, when the stock exchanges issue comment letters to firms that do not sufficiently comply with the relevant IIDG, they are more likely to require that the firm provide such information in the MD&A section. See https://www.sohu.com/a/383627722_733484 (in Chinese).
9 In theory, firms should disclose all material events. However, many events may be deemed insignificant. Consequently, businesses may choose to not disclose clear information, to disclose only specific types of events, or to manipulate the timing of the disclosure (Li, Citation2013).
10 As policymakers state, some firms provide lengthy but uninformative or strategically timed disclosures. Firms can also use too many technical terms or hard-to-understand jargon. Therefore, policymakers’ standards should regulate the additional valuable information that listed firms should reveal. See http://www.sse.com.cn/disclosure/credibility/dynamic/c/c_20150918_3990860.shtml (in Chinese).
11 The securities law that came into effect in 2020 significantly increases listed firms’ disclosure responsibilities and may change how firms disclose information in their 2020 annual reports. In addition, the first IIDGs to be implemented were for the radio, film and television, and pharmaceuticals and biological products industries. These guidelines became effective on January 7, 2013, making 2012 the adoption year. Accordingly, applicable firms’ 2012 annual reports should disclose the required information. We also note that we do not use some of the initial IIDGs in the regressions due to stringent data requirements, such as having at least one observation per treated or control firm both before IIDG implementation and after it.
12 For example, when the earlier treatment firms serve as the controls for the later treatment firms, the larger change in outcomes for the earlier firms relative to that for the later ones may yield reverse effects.
13 For the control firms, IIDG_related_words is the natural logarithm of 1 plus the number of IIDG-related words that appear in these firms’ MD&As, where the IIDG applies to the treated firms in the same cohort. If more than one IIDG applies to a single cohort, we calculate the mean value.
14 Another example is the IIDG for the agricultural, forestry, livestock, and fishery industry. In response to the IIDG, Shandong Homey Aquatic Development Co. Ltd. (stock code: 600467; a firm that has to comply with the IIDG) added in its MD&A: ‘To promote research and development, the firm has set up the College of Marine Life Sciences (Chinese name: 海洋生命科学院) and the Marine Food Testing Center (Chinese name: 海洋食品检测中心), and cooperates with the Yellow Sea Fisheries Research Institute of Chinese Academy of Fisheries Sciences (Chinese name: 中国水产科学研究院黄海水产研究所), the Ocean Research Institute of Chinese Academy of Sciences (Chinese name: 中科院海洋研究所), the Shandong Academy of Sciences (Chinese name: 山东省科学院), and the Ocean University of China (Chinese name: 中国海洋大学).’ These names can also be identified by the Stanford NER technique as organization names.
15 Our analysis does not include the categories of money and percentage that Hope et al. (Citation2016) propose because identifying these two categories is less precise than identifying other categories such as organization, person, and location (https://nlp.stanford.edu/software/CRF-NER.html). Moreover, we only focus on the MD&A’s textual content; money and percentage are more likely to appear in tabular format.
16 We use the Jieba package to split MD&As. Before splitting an MD&A, we add accounting keywords from the Beijing Sogou Technology Development Co., Ltd., which developed one of the mainstream Chinese input tools (see accounting keywords from https://pinyin.sogou.com/dict/detail/index/20659 (in Chinese)). When splitting each MD&A into words, we remove stop words because they do not have a semantic meaning. We also remove quantitative data (e.g., numbers and percentages) from MD&As to calculate the textual similarity.
17 By using TF-IDF, the weight of a term (word) increases with the frequency with which it appears in a specific MD&A, but it is penalized when it is common across all MD&As in a year.
18 The signifiers SH, SZ, and SZ-C respectively indicate that IIDGs apply to firms listed on the main board of the Shanghai Stock Exchange, the main board of the Shenzhen Stock Exchange, and the ChiNext board of the Shenzhen Stock Exchange.
19 The mean of IIDG_related_words is low because control firms, which dominate the sample due to the stacked DiD research design, do not need to comply with IIDGs. We thank an anonymous reviewer for pointing this out.
20 The result does not change when we calculate AbsCAR using alternative event windows. Online Appendix Table V offers further evidence of the information usefulness associated with IIDG implementation by examining its effect on the verifiability of a firm’s financial and operational information and its comparability with firms in the same IIDG industry. Verifiability is captured by the firm’s financial quality, which is the signed discretionary accruals based on the modified Jones model. We determine comparability by the mean textual similarity of a firm’s MD&As to those of other firms that must comply with the same IIDG. We find that after IIDG implementation, both the verifiability and comparability of these firms’ information increases relative to those of the control firms.
21 The two textual disclosure variables may be endogenous and exposed to IIDG implementation events. As Hudson et al. (Citation2017) suggest, the instrument variable estimate is a properly weighted average of the causal effects of textual disclosure in response to IIDGs for firms whose textual disclosure status is affected by the instrument.
22 In the new models, the textual disclosure and price informativeness in the three years before adoption serve as the benchmark; the sample period is the [−3, +2] year window.
23 The coefficients on X and Treat*X (X denotes HHI) are not shown in the regression in Columns (1) and (2). Since X is an industry-level indicator, both X and its interaction with Treat are absorbed by cohort×firm fixed effects.