Better or worse? Revealing the impact of common institutional ownership on annual report readability

In today's complex capital markets, corporate annual reports are crucial for investors to assess corporate performance and risk. Annual reports contain both numerical data and narrative text, the latter offering valuable supplementary information. However, the lack of robust integrity mechanisms and incomplete regulatory frameworks can leave external investors struggling to accurately evaluate companies. This presents opportunities for manipulation, especially in narrative text, where firms may obfuscate information to mask poor performance by reducing annual report readability (ARR). Regulators are increasingly focused on ARR as a measure of transparency and investor protection, highlighting the need to understand the factors influencing it. Common institutional ownership (CIO), where an institutional investor holds shares in multiple companies within the same industry, is a significant aspect of modern corporate governance. CIO can potentially exert both positive and negative influences. While it can enhance governance through information synergy and supervisory oversight, it might also lead to collusive behavior among firms within the investor's portfolio. This study aims to investigate the impact of CIO on ARR in the context of China's developing capital market, exploring whether CIO promotes greater transparency or leads to reduced readability for private gain and examining the underlying mechanisms.

Existing literature on the economic consequences of CIO is divided between the synergistic governance perspective and the collusive manipulation perspective. The synergistic view posits that CIO enhances governance by improving information sharing and enabling more effective monitoring. Studies supporting this perspective highlight CIO's positive impact on mitigating earnings management, constraining managerial rent extraction, improving investment efficiency, and promoting innovation. In contrast, the collusive manipulation perspective argues that CIO can foster collusive behavior among firms within an investor's portfolio, leading to negative consequences such as reduced corporate social responsibility and hampered digital transformation. Regarding ARR, research suggests that poor readability can conceal negative information, leading to adverse outcomes such as increased equity mispricing and hindered access to trade credit. However, prior studies on ARR have primarily focused on internal firm-level factors, with limited research on the impact of shareholder characteristics, particularly CIO.

This study uses data from Chinese listed companies between 2007 and 2021 from the CSMAR and WinGo databases. After data cleaning (excluding financial firms, ST and *ST companies, and those with missing data), the final sample consisted of 35,916 observations. Annual report readability (ARR) was measured using the sentence production probability method developed by Shin et al. (2020). CIO was measured in three ways: (1) a dummy variable indicating the presence of CIO; (2) the degree of CIO linkage; and (3) the proportion of CIO shareholding. Control variables included firm characteristics such as firm size, board size, proportion of independent directors, and financial ratios. Ordinary Least Squares (OLS) regression with year and industry fixed effects was used as the baseline model. Robustness checks were conducted using propensity score matching (PSM), two-stage least squares (2SLS), a difference-in-differences model, a model using the t+1 period dependent variable, and a model with a shortened sample period. Moderating effects of analyst attention (measured by the number of analysts following the firm), industry concentration (measured by the Herfindahl-Hirschman Index), and media coverage (measured by the number of media reports) were examined. The mediating effect of operational risk (measured using the cumulative distribution probability of the standard deviation of the rolling values of the pretax depreciation and amortization profit margin for the previous four years) was also analyzed.

The baseline regression analysis and subsequent robustness tests consistently showed a negative relationship between CIO and ARR. Specifically, the presence of CIO, the degree of CIO linkage, and the proportion of CIO shareholding were all significantly and negatively associated with ARR. Further analysis revealed that the negative impact of CIO on ARR was stronger in firms with higher analyst attention, higher industry concentration, and higher media coverage, suggesting that these factors exacerbate the collusive manipulation effect of CIO. The study also found that operational risk partially mediated the relationship between CIO and ARR, indicating that increased CIO leads to higher operational risk, which in turn reduces ARR.

The findings support the collusive manipulation hypothesis, suggesting that CIO can lead to less transparent reporting practices. Common institutional investors, seeking to maximize their portfolio returns, may collude with management to reduce ARR and obscure information that could harm their interests. This behavior is amplified when there is increased scrutiny from analysts, higher industry concentration, and greater media attention. The mediating role of operational risk suggests that the collusive behavior increases firms' operational risk, which management then seeks to mask through less readable reporting. These results have significant implications for understanding the complexities of corporate governance and the potential downsides of CIO.

This study provides valuable insights into the relationship between CIO and ARR. It demonstrates the negative impact of CIO on transparency in corporate reporting and suggests that regulatory authorities should be aware of this collusive manipulation effect of CIO, particularly in industries with high concentration and under increased scrutiny. Future research might explore the economic consequences of reduced ARR and investigate potential mitigation strategies for stakeholders. More sophisticated measures of ARR tailored to the linguistic nuances of the Chinese context are also needed.

The study acknowledges limitations. The measurement of ARR using the sentence production probability method might not fully capture all aspects of readability in the Chinese context. Furthermore, while the study examines the impact of CIO on ARR, it does not directly measure the economic consequences of the reduced readability. Future research could address these limitations by employing more comprehensive measures of readability and examining the economic impact of less transparent reporting.