Governance, sentiment analysis, and initial public offering underpricing

• Author: Alexander Guzmán,Maximiliano González,María Andrea Trujillo,Diego F. Tellez‐Falla
• DOI: http://doi.org/10.1111/corg.12272
• Published date: 01 May 2019
• Date: 01 May 2019

ORIGINAL ARTICLE

Governance, sentiment analysis, and initial public offering

underpricing

Maximiliano González

1

|Alexander Guzmán

2

|Diego F. Tellez‐Falla

3

|

María Andrea Trujillo

2

1

School of Management, Universidad de los

Andes, Bogotá, Colombia

2

School of Business, CESA, Bogotá, Colombia

3

Department of Finance, Universidad EAFIT,

Medellín, Colombia

Correspondence

Diego F. Tellez‐Falla, Department of Finance,

Universidad EAFIT, Carrera 49 7 sur 50,

Medellín 050022, Colombia.

Email: dftellezf@eafit.edu.co

Abstract

Research Question/Issue: What is the relationship between governance, tone in

language, and underpricing of initial public offerings (IPOs) in Latin America?

Research Findings/Insights: We find a positive (negative) and statistically signifi-

cant relationship between board size (board independence) and IPO underpricing at

the time firms go public. But more importantly, when interacting with corporate gov-

ernance variables at the firm level, uncertainty in tone still shows statistical signifi-

cance, which suggests that tone in communications matters for underpricing

independently of governance provisions at the firm level.

Theoretical/Academic Implications: Bigger boards seem to lead to greater

underpricing, whereas more independent boards mitigate it. This finding stresses

the relevance of functional convergence in regions with weak institutions such as

Latin America. Our results also suggest that tone matters for underpricing even in

firms with good governance practices. We argue that tone in corporate communica-

tions is a strong signaling mechanism for market participants.

Practitioner/Policy Implications: Our results suggest that tone in firms' communi-

cations is relevant for market valuation. In the context of family firms in an

underresearched context such as Latin America, we show that reputation effect is

priced by the market valuation of the IPO.

KEYWORDS

corporate governance, family firms, IPO underpricing, Latin America, sentiment analysis

1|INTRODUCTION

Going public can be a unique undertaking in a firm's life. The reasons

to go public differ. Some are explicit, such as obtaining capital to

strengthen an investment policy, mainly for capital expenditures

(CAPEX) and acquisitions (Celikyurt, Sevilir, & Shivdasani, 2010; Kim

& Weisbach, 2008) or to change the capital structure. Other reasons

could include gaining market liquidity (Brau & Fawcett, 2006), grab-

bing market share from private competitors (Chemmanur & He,

2011), or shareholder diversification (Bodnaruk, Kandel, Massa, &

Simonov, 2008) to permit idiosyncratic risk by tolerating higher profit

variability and hence allowing riskier and more aggressive output mar-

ket strategies (Chod & Lyandres, 2011).

Regardless of the reason to go public, the level of information

asymmetry between the issuing firm and potential investors is high.

This fact clearly implies an adverse selection problem affecting share

liquidity. A direct effect of lower liquidity is a higher cost of capital

for new firms in the market. Therefore, the decision to go public

implies that issuing firms must leave money on the table because

investors will undervalue their stocks as compensation for the risks

of information asymmetry.

1

Issuing firms know this ex ante and will

try to reduce the information asymmetry by sending credible signals

Received: 23 January 2018 Revised: 24 November 2018 Accepted: 4 January 2019

DOI: 10.1111/corg.12272

226 © 2019 John Wiley & Sons Ltd Corp Govern Int Rev. 2019;27:226–244.wileyonlinelibrary.com/journal/corg

to the market (Healy & Palepu, 2001). This process is supported by a

legal environment that facilitates information production, particularly

important in developing economies.

2

In this paper, we approach the relationship between governance,

sentiment analysis, and initial public offering (IPO) performance using

a database of Latin American IPOs. In a region like Latin America,

characterized by low investor protection, higher cost of capital, and

poor information disclosure, studying the IPO process is a major

concern. De la Torre and Schmukler (2007) analyzed the evolution

of capital markets in Latin America after reforms. The authors found

the region running behind other emerging markets, mainly Asian, in

terms of capital market development. Despite intense reforms under-

taken during the nineties (including privatizations and pension

reforms), and efforts to stabilize the macroeconomic environment,

capital markets in Latin America are still illiquid, small, and highly

concentrated.

In our setting, a potential investor is collecting information to

decide on investing in an IPO. Most of the available information is in

the IPO prospectus that the issuing firm has prepared with the help

of the underwriter. The prospectus describes the firm's current situa-

tion and future expectations. We argue that the tone used impacts

investors' assessment of the offer, and the market reads the prospec-

tus as a credible signal regarding a firm's ex ante uncertainty, a signal

relevant regardless of the highly tested corporate governance arrange-

ments at the firm level. A higher level of ex ante uncertainty reduces

the value of the issue and increases the level of underpricing (Beatty

& Ritter, 1986; Loughran & McDonald, 2013).

Using a database of Latin American IPOs spanning the period

2000–2014, we observe several results that arise from the regression

analysis. First, as shown in previous studies (e.g., Baker & Gompers,

2003; Li & Naughton, 2007), we find a positive (negative) and statisti-

cally significant relationship between board size (board independence)

and IPO underpricing at the time of going public. But more impor-

tantly, when interacting with corporate governance variables at the

firm level, uncertainty in tone still shows statistical significance, which

suggests that tone in communications matters for underpricing inde-

pendently of governance provisions at the firm level. We argue that

tone in corporate communications is a strong signaling mechanism

for market participants.

We also find empirical evidence for a link between uncertainty in

tone in the IPO prospectus and the level of IPO underpricing, which

supports the Loughran and McDonald (2013) findings regarding tone

as a proxy for ex ante uncertainty. In addition, we do not find empir-

ical evidence that governance provisions at the country level explain

IPO underpricing. Finally, we explore the role of family firms in Latin

American IPOs. We find that the average underpricing for family

firms is lower than for nonfamily firms. This supports the idea that

entrepreneurial families in Latin America care about their reputation

as controlling shareholders and treat minority shareholders fairly

(Gomes, 2000).

This study differs from previous ones regarding IPO underpricing

(e.g., Chahine & Filatotchev, 2008; Ferris, Hao, & Liao, 2013; Hanley

& Hoberg, 2010; Loughran & McDonald, 2013) by analyzing the

interaction of governance mechanisms at the firm level with tone

uncertainty in the IPO prospectus. In addition, to the best of our knowl-

edge, our study is the first to measure meaning in information disclo-

sures in a language other than English, using the financial word lists

proposed by Loughran and McDonald (2011). Finally, our study uses a

sample in a specific and underresearched context, that of Latin America.

Large information asymmetries represent a challenge for Latin

American countries, and market participants value good practices

leading to transparency at the firm level. Garay, González, Guzmán,

and Trujillo (2013) presented empirical evidence regarding the effect

that information disclosure has on firm valuation in Latin America.

They built an Internet‐based corporate disclosure index and found a

positive relation between the information disclosed, Tobin's Q, and

firms' Return On Assets (ROA). In addition, previous research in Latin

America shows a positive effect of corporate governance practices

on firm value (Leal & Carvalhal‐da‐Silva, 2007, for Brazil and Chile;

Chong & López‐de‐Silanes, 2007, for Mexico; Lefort & Walker,

2007, for Chile; Garay & González, 2008, for Venezuela).

The rest of this paper is organized as follows. In the next section,

we briefly review the literature on information asymmetry and IPO

underpricing, focusing on the role of language in mitigating informa-

tion asymmetries and developing our hypotheses. Next, we present

the data and research methodology. We then develop the results sec-

tion, which is divided into three subsections. The first explores the role

of language in IPO underpricing. This is followed by an empirical anal-

ysis of governance and IPO underpricing in Latin America and then a

section dedicated to information disclosure and family firms. Finally,

we conclude.

2|THEORETICAL FRAMEWORK AND

HYPOTHESIS DEVELOPMENT

In an IPO process, different investors analyze the prospectus prepared

by the firm to decide whether or not to participate. As a description of

the firm's current and future situation, as seen by the firm's manage-

ment and advisors, the tone used could affect investors' assessment

of the offer. We argue that the market reads the prospectus as a cred-

ible signal about firms' ex ante uncertainty. Moreover, we posit that

this signal is relevant regardless of the corporate governance arrange-

ments at the firm level.

To measure the tone of the prospectuses in our sample, we use

the dictionaries proposed by Loughran and McDonald (2011). These

have been used widely to study the relation between media reports

and firm performance. Some of these studies present evidence about

the role of news sentiment in predicting stock and market returns

(Chen & De, 2014; Dougal & Engelberg, 2012; Garcia, 2013). Mayew

and Venkatachalam (2012) used the dictionaries to look at conference

calls and future performance, whereas Solomon (2012) and Bajo and

Raimondo (2017) considered media coverage and stock prices. Liu

and McConnell (2013) studied media attention to proposed acquisi-

tions and the probability of abandoning a deal. Ahern and Sosyura

(2014) analyzed press releases and merger stock prices. The

GONZÁLEZ ET AL.227