Text

theses

My dissertation centers on the study of qualitative corporate disclosures. I integrate the relevant theory from the linguistics and incorporate new tools and methodologies from the text mining literature. My research goal is to advance our understanding of how market participants analyze and interpret the qualitative aspect of firm disclosures. My dissertation takes a step towards this goal by studying voluntary disclosures in several areas of corporate communication. These areas are: preliminary earnings announcements (Form 8-K), conference call transcripts, and corporate governance disclosures. The first essay studies qualitative order backlog (OB) disclosures in the preliminary earnings announcements. Despite the obvious potential for OB disclosures to predict future sales and stock returns, and the strong interest in OB data among market participants, these disclosures are not a part of the required audited financial statements, and OB disclosure is not required on a quarterly basis. Companies can choose to disclose OB in the earnings press release (Form 8-K) and they can also choose the format of these disclosures: quantitative or qualitative. Prior literature on the implications of order backlog for stock returns are both sparse and inconclusive (Rajgopal et al. (2003), Lev and Thiagarajan (1993). In my dissertation I address several questions related to OB disclosures, starting with the question of whether OB disclosures are useful in predicting next-period sales. I then focus on managers’ OB disclosure decisions in the annual as well as quarterly corporate filings. In particular, I examine the determinants of earlier and more frequent OB disclosures than the mandatory annual disclosure in Form 10-K, as well as whether these disclosures are presented in either quantitative or qualitative form. Finally, I study the incremental information content of OB disclosures (over and above other information released at the same time such as earnings, accruals, and other financial statements data) for future returns. I find that the backlog growth rate is positively and significantly associated with the next period sales. I show that the main determinants of voluntary OB disclosures are OB magnitude, prior OB disclosure habits, changes in OB, changes in the inventory levels, and the level of uncertainty faced by a firm. I provide evidence that the annual OB change signal is significantly associated with abnormal returns incrementally to earnings and accrual surprises both in the short-window and drift returns around the 4th-quarter preliminary earnings announcements, and around the Form 10-K filings. Finally, I find that there are significant market reactions to the quantitative and qualitative OB disclosures during quarterly earnings announcements beyond the market reactions to contemporaneous earnings surprises. This suggests that OB disclosures serve a useful role in interpreting the implications of the current earnings surprises on future returns. The second essay examines the value relevance of euphemisms in the conference call transcripts. Euphemisms are “indirect words or phrase that people often use to refer to something embarrassing or unpleasant, sometimes to make it seem more acceptable than what it really is” (Hornby 2004). These are one of the linguistic tools used by managers to soften their explanation of poor company performance during conference calls. Prior studies in accounting and finance find evidence that managers use conference calls with investors not only to communicate financial results, but also to manage investor perception of firm performance. Managers tend to avoid taking responsibility for negative results by blaming external factors (for example, industry or weather) or by talking about it in an evasive and confusing manner. I extend these empirical studies by exploring how euphemisms are used during earnings conference calls to manage investor perception of company performance. This is the first study to document the use of euphemisms in corporate communication. I have built the first dictionary of euphemisms used in business discourse and shown that euphemisms are indeed used in conference calls (on average more than 70% of calls will have at least one euphemism) across various sectors and time periods. In my study I focus on the determinants of euphemism usage as well as investors’ reaction to the use of euphemisms during conference calls. I predict and find that higher use of euphemisms is negatively associated with firm’s operating performance. I also find that the use of euphemisms is perceived as a negative signal by investors and results in the immediate negative market reaction. However, due to the impression management aspect of euphemisms, investors underreact to the signal as they underestimate the severity of the problems faced by the company, which results in a delayed negative reaction to the content of a conference call. In the third essay, I study the relationship between the length of director tenure and two main functions of the board: monitoring and advising. I examine whether corporate boards consisting of longer-serving directors are better able to fulfill these functions due to the firm-specific knowledge accumulation, or whether director performance suffers due to the deterioration of their technical knowledge and due to the decreasing independence of the board from managers. Using a sample of up to 3,000 firms over an 18-year period, the evidence suggests that board tenure is positively related to forward-looking measures of market value, with the relationship reversing after about 9 years on average. The detrimental effect of longer board tenure on market value is stronger for high growth firms, which is consistent with the deterioration of the board members’ ability to provide useful advice on technical matters relating to the operations of a firm. I also find that board tenure is reflected in stock returns in a similar manner to market values, and that the declining effect of long board tenure is similarly more pronounced for dynamic, growing firms.

ETD_8695

Ph.D.

Includes bibliographical references

by Katsiaryna Suslava

doi:10.7282/T3N301CC

ETD doctoral

The author owns the copyright to this work.