Textual Complexity in Nonperiodic Disclosures and Investors’ Use of Social Media as an Alternative Channel for Corporate Information

Abstract

Textual complexity within corporate disclosures may reduce investors' comprehension of those disclosures. However, investors can rely on alternative sources of information to process textually complex disclosures. Unlike periodic disclosures, where various alternative sources (e.g., analysts and news media reports) are available to investors, there are limited alternative sources for unanticipated nonperiodic disclosures. Social media is one such source and has been shown to facilitate retail investors' access to information for investment decision-making. This leads to the question of how effective social media are for processing textually complex nonperiodic disclosures.

This thesis examines the connection between the causes and consequences of nonperiodic disclosures' textual complexity and whether social media can assist investors in processing such disclosures. The research consists of three empirical studies using corporate announcements from the Australian Securities Exchange and Hotcopper data, Australia's largest stock message board. The first study compares the textual complexity of nonperiodic and periodic disclosures and identifies whether they have different determinants. The second study examines the consequences of managers' strategic reporting on nonperiodic disclosures' textual complexity and their market outcomes. The third study examines the investors' use of alternative information from a regulated stock message board for complex disclosures and whether that improves stock price informativeness.

In the first study, evidence shows that nonperiodic disclosures are textually less complex than periodic disclosures, though both disclosures become less readable, lengthy and boilerplate over the sample period. Further, firms from the resource, finance, and regulated industries tend to produce more textually complex periodic disclosures than nonperiodic due to their complex structure and business practices. In terms of determinants, compared to periodic disclosures, the textual complexity of nonperiodic disclosures is more influenced by time trends, market-to-book ratio, special items, and gearing, but not by poor performance. Besides firm-specific factors, the textual complexity of nonperiodic disclosure is also influenced by managers' strategic reporting, mainly when bad news is released after trading hours or on the week's last trading day.

The second study finds that after the GN8 revision of Listing Rule 3.1 by the Australian Securities Exchange, there is a significant reduction in the readability of nonperiodic disclosures among firms using boilerplate language in their disclosures. The boilerplate language, I argue, is used for reducing the litigation risks arising from the GN8 revision requirements for complete disclosures within strict time frames. The study also finds that the market returns of firms are negatively associated with the use of boilerplate language in nonperiodic disclosures, particularly for firms with more retail investors. These findings suggest that the GN8 revision has had unintended consequences. It has increased the use of boilerplate language in nonperiodic disclosures, which can adversely affect the readability of the disclosures and reduce their informativeness.

For the third study, I utilise the Australian Regulatory Guidance 162 within the GN8 revision environment. This guidance restricts CEOs' and experts' engagement on stock message boards. The results of my study reveal that less readable periodic and nonperiodic disclosures receive more views and discussions on the stock message board. Upon partitioning periodic and nonperiodic disclosures into subsamples, I find that longer and less readable nonperiodic disclosures attract greater stock message board views and discussions. Further, higher stock message board discussions are associated with lower stock price synchronicity for firms with textually complex disclosures, particularly for those with lower institutional ownership. This implies that retail investors view the regulated stock message board as a credible and timelier alternative channel to process textually complex nonperiodic disclosures.

This thesis first contributes to the extant literature on textual complexity in corporate disclosures by adding an examination of the determinants and effects of textual complexity of nonperiodic disclosures. Next, it extends the social media literature by examining investors' use of regulated stock message boards to deal with textually complex nonperiodic disclosures.

This thesis first contributes to the extant literature on textual complexity in corporate disclosures by adding an examination of the determinants and effects of textual complexity of nonperiodic disclosures. Next, it extends the social media literature by examining investors' use of regulated stock message boards to deal with textually complex nonperiodic disclosures.