Innovation disclosure in times of uncertainty

• Author: Mario Daniele Amore
• Date: 01 October 2020
• Published date: 01 October 2020
• DOI: http://doi.org/10.1111/jems.12390

J Econ Manage Strat. 2020;29:792–815.wileyonlinelibrary.com/journal/jems792

|

© 2020 Wiley Periodicals LLC

Received: 21 December 2018

|

Revised: 5 June 2020

|

Accepted: 16 June 2020

DOI: 10.1111/jems.12390

ORIGINAL ARTICLE

Innovation disclosure in times of uncertainty

Mario Daniele Amore

Department of Management &

Technology, Bocconi University and

CEPR, Milan, Italy

Correspondence

Mario Daniele Amore, Department of

Management & Technology,

Bocconi University, Via Roentgen 1,

20126 Milan, Italy.

Email: mario.amore@unibocconi.it

Abstract

A recent literature shows that many firms feature missing R&D expenses in

their accounting statements. This study explores how economic policy

uncertainty affects the decision to disclose innovation‐related information.

Empirical analyses on a panel of U.S. listed companies show that policy

uncertainty increases the likelihood of missing R&D (as opposed to both po-

sitive and zero R&D). This result is more pronounced for firms that enjoy a

leadership position in their industry, firms in states subject to a weaker legal

protection of internal knowledge, and firms that rely more on government

demand. During uncertain times, firms also file patents that exhibit a greater

textual vagueness. Finally, the evidence suggests that missing R&D helps firms

alleviate the negative impact of policy uncertainty on market value.

1|INTRODUCTION

Macroeconomic events in recent years have fueled a research on the relationship between economic policy uncertainty

and corporate outcomes (Baker, Bloom, & Davis, 2016; Bloom, 2014). A useful perspective to understand this re-

lationship comes from real option theory, which suggests that uncertainty increases the value of delaying irreversible

investment (Bernanke, 1983). Moreover, uncertainty can affect companies by raising risk premia and increasing bor-

rowing costs (Pastor & Veronesi, 2013). While this literature has traditionally focused on investment and financing

policies (Bonaime, Gulen, & Ion, 2018; Colak, Durnev, & Qian, 2016; Gulen & Ion, 2016; Julio & Yook, 2012), recent

studies have begun to explore the effect of policy uncertainty on firms' disclosure activities.

Discretionary disclosure has been the cornerstone of a long‐running literature (e.g., Jovanovic, 1982; Verrecchia,

1983). Adverse selection arguments (Grossman, 1981; Grossman & Hart, 1980) suggest that if the market knows that a

firm's management possesses a certain private information, then the management has incentives to disclose this

information to avoid investors discounting the firm's shares for the lack of disclosure.

1

Recent works find that during

periods of high policy uncertainty firms provide more voluntary disclosure, as measured by management forecasts or

8K filings (Bird, Karolyi, & Ruchti, 2017; Boone, Kim, & White, 2017; Nagar, Schoenfeld, & Wellman, 2019).

Assessing innovative firms is notoriously difficult due to asymmetric information problems (e.g., Aboody &

Lev, 2000). As a result, stakeholders and analysts spend a great deal of effort to evaluate firms' R&D activities (Barth,

Kasznik, & McNichols, 2001). The transparency of innovation activities is key to ease this task and increase the

informativeness of stock prices. According to the U.S. Statement of Financial Accounting Standards, firms must provide

any relevant R&D expenditure in their accounting statements. Yet, firms often do not disclose the amount of R&D

expenditures. In particular, Koh and Reeb (2015) find that around half of the companies listed in the NYSE do not

report any R&D (i.e., neither positive nor zero).

2

Missing R&D may arise from several reasons, including the inability to separate R&D from general expenses, an

inattentive behavior due to a weak enforceability of R&D reporting requirements, the lack of materially relevant R&D

expenditures, and the way in which innovation activities are organized (Koh & Reeb, 2015). A recent literature further

suggests that missing R&D may be a deliberate choice of the firm aimed at obfuscating its innovation activities (Koh,

Reeb, & Zhao, 2017b, 2018). I contribute to this literature by studying how economic policy uncertainty shapes firms'

disclosure of innovation activities.

To measure policy uncertainty, I use the index developed in Baker et al. (2016), which builds on a comprehensive

text‐based analysis of major newspapers and is a widely used methodology to quantify policy‐related uncertainty

(Bonaime et al., 2018; Brogaard & Detzel, 2015; Gulen & Ion, 2016; Pastor & Veronesi, 2013; Wellman, 2017). Re-

gression analyses on a panel of U.S. listed firms from 1976 to 2012 indicate that economic policy uncertainty is

associated with a significantly higher likelihood of missing R&D. This result is robust to controlling for several variables

common to the R&D reporting literature as well as for other sources of uncertainty, macroeconomic conditions, and

constant heterogeneity. Results are also robust to measuring policy uncertainty in an alternative way, that is, by

exploiting the occurrence of political elections in an international sample (Julio & Yook, 2012).

The analysis then explores whether missing R&D occurs because firms reduce innovation expenditures to a ma-

terially irrelevant level during uncertain times.

3

To this end, I separately analyze the likelihood of missing R&D as

opposed to positive or zero R&D. Results show that policy uncertainty has a strong and significant impact on the

likelihood of missing R&D versus positive R&D but also the likelihood of missing R&D versus zero R&D. Moreover,

policy uncertainty increases the likelihood of missing R&D among the subsample of patenting firms (i.e., firms with

significant innovation activities). Finally, missing‐R&D firms do not experience a significant decline in patenting

activities during periods of policy uncertainty. Collectively, these tests suggest that missing R&D during uncertain times

may arise from a broader set of mechanisms than a mere decline in innovation activities. Seeking to elicit these

mechanisms, I explore the heterogeneity behind the average result.

First, I study whether industry leaders are more likely than other firms to withdraw R&D information during

uncertain times. The rationale here comes from the idea that when private signals are noisy and optimization is costly,

firms seek to reduce uncertainty and minimize the risk of unfavorable consequences by following the actions of

industry leaders (Chen, Huang, & Lin, 2017; Francis, Hasan, Sun, & Waisman, 2014; Leary & Roberts, 2014), which are

typically better‐positioned to figure the industry implications of policy uncertainty. In turn, to reduce the information

that rivals may exploit to orchestrate competitive actions, an industry leader may have more incentives to withhold

R&D information during uncertain times. Measuring industry leadership as in Cantrell and Dickinson (2020), I show

that being an industry leader amplifies the positive effect of policy uncertainty on missing R&D.

Second, I explore legal variations in the risk of external appropriation of a firm's disclosed knowledge. As argued,

disclosing R&D during uncertain times may provide useful clues on a firm's current or future strategies, which rivals

can use to engage in imitation or R&D races. A legal regime that restrains knowledge appropriation will weaken the

incentives to withhold R&D information during uncertain periods. I exploit the staggered adoption of the Inevitable

Disclosure Doctrine (IDD) across U.S. states during the 1970s and 1980s (Klasa, Ortiz‐Molina, Serfling, & Srinivasan,

2018; Koh et al., 2018; Li, Lin, & Zhang, 2018). The IDD enabled an employer to prevent current or former employees

from working for rival firms if the employer demonstrates that the employee will inevitably disclose trade secrets in the

new position. Results show that policy uncertainty has a weaker effect on missing R&D among firms subject to a better

protection of internal knowledge.

Finally, I explore how the reliance of a firm on government activities shapes the effect of policy uncertainty on

missing R&D. The intuition here is that government demand creates a transmission channel of policy uncertainty on

corporate activities: firms that rely more on government purchases would be affected more intensively by policy

uncertainty (Duong, Nguyen, Nguyen, & Rhee, 2020; Gulen & Ion, 2016). Using information on how much firms rely on

the government for their revenues, I find that a greater reliance on government expenditures amplifies the effect of

policy uncertainty on missing R&D.

Withholding R&D information is not the only way through which firms may obfuscate their innovation policy

during uncertain times. Another possibility comes from strategically drafting patents so as to limit the amount of

disclosed information (Hall & Harhoff, 2012). To investigate this issue, I conduct a text‐based analysis of all USPTO

patent documents filed by the sample firms (amounting to 500,000 patents) to construct a measure of textual vagueness

(Arinas, 2012; Kim, 2019). Results indicate that policy uncertainty increases patent vagueness, as proxied by a greater

number of vague expressions and a lower number of figures in the patent documents.

In conclusion, I study the value implications of withholding R&D information during uncertain times. In line with

the notion that innovation opacity encompasses strategic benefits, results show that when policy uncertainty surges

firms with missing R&D exhibit a milder decline in financial performance.

AMORE

|

793