Annual report readability and stock liquidity

• Author: Sabri Boubaker,Hatem Rjiba,Dimitrios Gounopoulos
• Published date: 01 May 2019
• DOI: http://doi.org/10.1111/fmii.12110
• Date: 01 May 2019

DOI: 10.1111/fmii.12110

ORIGINAL ARTICLE

Annual report readability and stock liquidity

Sabri Boubaker1,2 Dimitrios Gounopoulos3Hatem Rjiba4

1South Champagne Business School, Troyes,

France

2IRG,University of Paris Est, France

3University of Bath, School of Management,

Bath, UK

4ESSCA School of Management, France

Correspondence

DimitriosGounopoulos, University of Bath,

Schoolof Management, Bath, UK.

Email:D.Gounopoulos@bath.ac.uk

Abstract

We examine the effect of annual report textualcomplexity on firms’

stock liquidity. Using techniques from computational linguistics, we

predict and find that less readable filings are associated with lower

stock liquidity. Our study provides evidence that difficult-to-read

annual reports hinder investors’ ability to process and analyze infor-

mation contained in corporate annual reports, reducing thereby

their willingness to trade which decreases stock liquidity. Our find-

ings are robust to a battery of sensitivity tests, including endogene-

ity, use of alternative estimation techniques, and use of alternative

liquidity and readability proxies.

KEYWORDS

annual report readability, information asymmetry, stock liquidity,

textual complexity

JEL CLASSIFICATION

G11, G12, M48

“The finest words in the world are only vain sounds if you can't understand them. The best sentence? The

shortest.”

–Anatole France

1INTRODUCTION

Corporatedisclosure plays a crucial role in the functioning of capital markets. Existing theoretical (e.g., Diamond, 1985;

Diamond & Verrecchia, 1991) and empirical (e.g., Bhattacharya, Desai, & Venkataraman, 2013; Brown & Hillegeist,

2007; Heflin, Shaw, & Wild, 2005; Ng, 2011) literature demonstrates unambiguously that greater and better public

disclosure reduces information asymmetry and enhances the liquidity of firm shares. Nevertheless, most prior studies

focus on quantitative information in their appraisalof disclosure quality, thus overlooking a key feature that character-

izes corporate reporting, namely,qualitative information in the form of text. This striking paucity of research might be

Thisarticle is published with the permission of the Controller of HMSO and the Queen 's Printer for Scotland.

c

2019 Crown copyright. Financial Markets, Institutions, and Instruments c

2019 New YorkUniversity Salomon Center and Wiley

Periodicals, Inc.

Financial Markets,Inst. & Inst. 2019;28:159–186. wileyonlinelibrary.com/journal/fmii 159

160 BOUBAKER ET AL.

partly attributable to the lack of evidence as to whether corporate narrativesconvey value-relevant information over

and above that contained in quantitative disclosure.

Toelucidate the above concern, several studies examine the incremental explanatory power of “soft” information

(i.e.,verbal disclosure) over firms’ reported “hard” information (i.e., quantitative disclosure). Relying on a machine learn-

ing naïve Bayesianapproach, Li (2010a) discovers that the tone of forward-looking statements in a firm's management

discussion and analysis (MD&A) is positively correlated with its future earnings and liquidity.It also has an explanatory

power incremental to other firm characteristics in predicting future performance.

Using a larger set of disclosure sources, Kothari, Li, and Short (2009) classify information from firms’ MD&A sec-

tions, analysts’ reports and news stories published in the business press into favorable and unfavorable categories.

They document that firms with more favorable disclosure havelower cost of equity, lower stock return volatility, and

more accurate analyst forecasts. Larcker and Zakolyukina (2012) document that models based on linguistic features

appeared in CEO and CFO narratives during earnings conference calls are as informative as accounting-based mod-

els in detecting financial statement misrepresentation.1Other studies focus on CEOs speech to measure verbal and

nonverbal vocal cues in detecting financial misreporting (e.g., Capalbo, Frino, Lim, Mollica, & Palumbo 2017; Craig,

Mortensen, & Iyer 2013). For example,Craig et al. (2013) examine the linguistic features of letters signed by Satyam's

chairman Ramalinga Raju, who was involvedin the biggest corporate scandal in India. The authors find that Raju's word

choices changed noticeably in his five annual-report letters prior to the collapse of his firm, as the scale of his financial

misstatements increased. Capalbo et al. (2017) find that narcissistic CEOs, that is, CEOs who overly use first person

singular pronouns in the question-and-answer session of conference calls, are more likely to engage in opportunistic

earnings management.

Recent years have also witnessed the emergence of a great deal of research that employstechniques from compu-

tational linguistics to gauge the quality of firm financial disclosure. The seminal study by Li (2008) introduced readabil-

ity measures in the accounting literature by examiningthe association between 10-K textual complexity and earnings

quality, as measured by earnings persistence. Subsequent research investigatesthe effects of the readability of vari-

ous corporate disclosures on market pricing (e.g.,Lawrence, 2013; Lee, 2012), analyst behavior (e.g., De Franco, Hope,

Vyas, & Zhou, 2014; Lehavy,Li, & Merkley, 2011), and litigation risk (e.g.,Nelson & Pritchard, 2016), among others.

Research related to finance and psychology demonstrates that investorsare susceptible to biases in their cognitive

information processing related to the way information is conveyedto them (e.g., Merkl-Davies & Brennan, 2007; Mul-

lainathan & Shleifer,2005). In particular, reading and interpreting a large body of text requires considerable cognitive

efforts on the part of a reader to analyze the information contained in firms’ reports. As a consequence, more textually

complex disclosures are expected to add additional barriers to investor willingness and ability to process the filings,

and thus constitute an important impediment to trading.

This study bridgesthe literature from finance and psychology by examining the link between the readability of firms’

annual reports and the liquidity of their stocks. Using a sample of 3899 annual corporate filings overthe period 2002–

2013, our findings suggest that less readable disclosure undermines investors’ability to process information, resulting

in a decline in the liquidity of firm shares. These results are in line with our theoretical predictionand constitute a novel

contribution to the literature. Moreover,they appear to be robust to a number of sensitivity checks, including the use

of several estimation techniques, alternative measures of main dependent and independent variables, and endogene-

ity issues. Additional analyses reveal that complex annual reports as measured by abnormal liquidity around report

filing dates affect stock liquidity in the short run, and document that mandating the adoption of International Financial

Reporting Standards (IFRS) reduces the effect of less readable disclosure on stock liquidity.

This study makes severalcontributions to the literature. First, it indicates that textual features of corporate filings

do matter in explaining cross-sectional differences in firms’ stock liquidity. We further supplement the growing

literature by investigatingthe capital market outcomes of narrative disclosure (e.g., Lee, 2012; Loughran & McDonald,

2009; Miller,2010). Second, our work supplements the studies that flesh out the effect of impaired cognitive capacity

on stock prices and investor trading behavior, including stock buying and selling decisions (e.g., Barber & Odean,

2008), category learning and price dynamics (e.g., Peng & Xiong, 2006), and under- and over-reaction to news and

accounting information (e.g., Hirshleifer, Lim, & Teoh, 2011; Li & Yu, 2012). Finally,by focusing on the readability of