Print
E-MailAnastasia Zakolyukina is Associate Professor of Accounting and Neubauer Family Faculty Fellow at University of Chicago Booth School of Business; Ian D. Gow is Professor at the University of Melbourne; and David F. Larcker is the James Irvin Miller Professor of Accounting at Stanford Graduate School of Business. This post is based on their recent paper.
“Sunlight is the best disinfectant.”
— Justice Louis D. Brandeis
Disclosure of information has long been a key element of corporate governance. While much disclosure is governed by laws, regulations, standards and the like, much of the information investors rely on is provided voluntarily by firms. Since Regulation Fair Disclosure was introduced by the United States Securities and Exchange Commission (SEC), corporate conference calls have emerged as an important channel for firms to disclose information to capital markets.
A challenge for research into the determinants and consequences of firms’ disclosure choices is that measuring these choices is difficult. Two measures have been particularly popular in academic research. For a number of years in the 1990s, the AIMR (a predecessor organization to the CFA Institute) provided a measure of disclosure quality based on analyst ratings. But that rating has not been produced since 1997. A number of studies have measured disclosure according to whether a firm issued management forecasts of earnings (“earnings guidance”). However, many have decried the practice of issuing such guidance, especially of quarterly earnings, arguing that it promotes short-termism, and many firms have elected to stop providing such guidance. As such, the notion that guidance represents “good” disclosure practices has become dubious.
Our paper addresses the need for alternative measures of disclosure choice by constructing a new measure. A feature of a typical conference call is that a portion of it is devoted to the firm’s managers providing responses to questions asked by participants, who are primarily sell-side equity analysts. However, many of these questions are met by explicit “non-answers”, such as “we do not disclose those numbers” or “I can’t give you any specifics” or, simply, “I don’t know”. Our measure uses linguistic analysis of a manager’s response to determine if it involves an explicit “non-answer” to the analyst’s question. Using our measure, about 11% of questions elicit non-answers, a rate that is stable over time and similar across industries.
To demonstrate the usefulness of our measure, we apply it to test two longstanding hypotheses from academic research.
The first hypothesis holds that firms will be less willing to disclose when product market competition is more intense, as they will be reluctant to disclose information that would be useful to competitors. Although several papers have tested the prediction that greater competition leads to less disclosure due to proprietary costs, empirical support for this prediction has been mixed. Unlike many prior studies, we find a negative association between competition and disclosure that holds for the four different measures of competition we consider.
The granular nature of our disclosure measure enables us to provide a stronger test of the relation between product-market competition and disclosure choice. We find that questions about products are less likely to be answered than other questions, and less likely to be answered when competition is stronger.
The second hypothesis posits that firms will be more willing to disclose information when they prior to raising capital. While firms are more forthcoming prior to raising capital, we find that firms are less likely to answer future-performance-related questions shortly before equity or debt offerings, consistent with concerns about legal liability being an important determinant of disclosure choices notwithstanding safe-harbor provisions of the 1995 Private Securities Litigation Reform Act.
We examine disclosure behavior around the bankruptcy of Lehman Brothers in 2008, when capital-market concerns were clearly heightened. We find the non-answer rate is significantly lower for distressed firms in the wake of the Lehman bankruptcy, consistent with accentuated capital market incentives causing these firms to be more forthcoming with information.
Our paper adds to prior literature in a number of ways. Our primary contribution is a novel measure of disclosure choice based on non-answers by managers during conference calls. Our classification algorithm for non-answers can be easily applied by researchers the increasingly large sample of conference calls. Additionally, our measure can be constructed at the level of individual question-answer pairs, allowing much finer analysis than could be done with historically popular measures of disclosure.
We also provide additional evidence of firms’ disclosure choices being driven by both product-market and capital-market concerns. Although several papers have examined either competition or capital market incentives, the evidence in support of standard hypotheses has often been weak. Our paper examines both incentives and finds robust evidence in support of their association with disclosure choice.
The complete paper is available here.
