A contrast of the popularity and the interpretation of non‐GAAP earning disclosures in different industries

• Published date: 01 January 2024
• Author: Kang Cheng,Mohammad Tavakolifar,Barkat Ullah
• Date: 01 January 2024
• DOI: http://doi.org/10.1002/jcaf.22652

Received:  April Revised:  June Accepted:  July 

DOI: ./jcaf.

RESEARCH ARTICLE

A contrast of the popularity and the interpretation of

non-GAAP earning disclosures in different industries

Kang Cheng1Mohammad Tavakolifar2Barkat Ullah1

Department of Accounting and Finance,

Earl. G. Graves School of Business and

Management, Morgan State University,

Baltimore, Maryland, USA

Department of Accounting, Business

Law and Finance, College of Business and

Technology, NortheasternIllinois

University, Chicago, Illinois, USA

Correspondence

Kang Cheng, Department of Accounting

and Finance, Earl. G. Graves School of

Business and Management, Morgan State

University, Baltimore,MD , USA.

Email: kang.cheng@morgan.edu

Abstract

This study examines non-GAAP disclosures in two selected industries, the

consumer nondurable goods industry and the business services industry, to

address the question of whether non-GAAP measures are applied and inter-

preted uniformly across industries, and more importantly, if the market reacts

to non-GAAP disclosures similarly across different industries. Industry mem-

bership potentially has an impact on the usage and interpretation of non-GAAP

disclosures; some industries issue their own policy trying to standardize the

use of non-GAAP measures in their industry. However, industry effects on non-

GAAP disclosures have not been thoroughly studied. This study fills in the gap.

Using hand-collected non-GAAP measures disclosed in the -K reports from 

firms in the selected two industries, this study finds that: () the use of non-GAAP

measures is more popular than reported in previous studies, and non-GAAP

measures are not limited to performance measures; () there is weak evidence of

industry preference as to which non-GAAP measures are more popular in that

industry; and () the market reacts differently to non-GAAP disclosures in differ-

ent industries. Combining the empirical findings, this study documents industry

effects and market reactions in the interpretationof non- GAAP disclosures. Con-

sidering that these effects have not been formally academically documented

in previous studies, this study carries practical implications for investors and

financial analysts.

KEYWORDS

business services, consumer nondurable goods, earning response coefficients (ERC), industry

effects, non-GAAP disclosures

1 INTRODUCTION

The use of non-GAAP financial measures in corporate

performance reporting is a widely studied accounting phe-

nomenon in the extant literature (Baik et al., ; Bentley

et al., ; Bhattacharya et al., ; Curtis et al., ;

Doyle et al., ; Leung & Veenman, ). Non-GAAP

financial measures are important to investorsand financial

analysts as such measures provide additional information

beyond what is mandated by standard accounting rules.

Investors use such measures to intrinsically assess the

operational performance of a company, value a company,

and assess company-specific potential risks and uncer-

tainties that may not be revealed from standard GAAP

financial measures alone. The variability of non-GAAP

metrics across industries poses a significant challenge

when attempting to directly compare companies operating

in different sectors. However,conducting industry-specific

non-GAAP research empowers investors and analysts to

gain a profound understanding of the distinct nuances

92 ©  Wiley Periodicals LLC.J Corp Account Finance. ;:–.wileyonlinelibrary.com/journal/jcaf

CHENG  .93

associated with each industry’s non-GAAP measures. This

understanding enables them to make informed decisions

and precise comparisons between companies. Addition-

ally, the complexity and diversity of non-GAAP measures

across industries present a challenge for regulatory over-

sight as it demands a nuanced and deep understanding of

each industry’s specific practices. Non-GAAP measures

often vary greatly due to industry-specific operational

characteristics, regulations, and financial idiosyncrasies,

and therefore standardizing the use and presentation of

such measures across all industries is complex, partic-

ularly for regulators tasked with ensuring consistency

and compliance across diverse sectors. Studies related to

non-GAAP financial measures contribute to accounting

research and carry significant implications for regulators

and practitioners including both investors and financial

analysts.

The major challenge of non-GAAP financial research

is the compilation of data in constructing the non-GAAP

financial measures. Non-GAAP disclosures reported by

managers are not available in any databases and therefore,

researchers have to extract non-GAAP financial mea-

sures from reporting entities’ press releases. Compiling

such data often requires a costly and time-consuming

hand-collection process.Non-GAAP disclosures on press

releases are not standardized or even required. It is totally

the reporting entity’s discretion as to whether to report,

what to report, and how much to report. To compile a

dataset with adequate variables to carry out empirical stud-

ies, researchers often conduct programmatic keyword text

searches to identify initially their sample firm-quarters

with non-GAAP financial measure disclosures (Barth

et al., ; Bentley et al., ; Bhattacharya et al.,

, ; Brown et al., ; Leung & Veenman, ;

Lougee & Marquardt, ; Zhang & Zheng, ). Often,

a keyword-search computer algorithm is applied to the

entire population of available -K filings from SEC’sEdgar

database, cross-sectional and over a selected time period

(Bentley et al., ; Leung & Veenman, ). The result

is a sample randomly scattered across industries. One clear

limitation of this approach is that the sample selection is

highly sensitive to the researcher’s choice of keywordsrel-

ative to the reporting entity’s choice of keywords in their

reporting. If the reporting entity uses the term “GAAP

adjusted” instead of non-GAAP, the press release will not

be picked up as having non-GAAP disclosures when “non-

GAAP” is used by the researcher as the search keyword.

Also, the resulting sample will be a scattered random sam-

ple. While basic information such as industry membership

and time periods are often analyzed within the sample

(Bhattacharya et al., ; Leung & Veenman, ), the

popularity of non-GAAP disclosures regarding the overall

population can only be inferenced (Marques, ).

An alternative to the widely-used keyword-search

approach is to use researcher-selected criteria to define

a sample and then conduct a comprehensive analysis

within the defined sample, regardless of the presence of

non-GAAP disclosure. The underlying assumption is that

there is information content even from firms that choose

not to disclose non-GAAP information. Such a defined-

sample approach is used by Marques () who studies

S&P  firms, and by Curtis et al. () who study

firm-quarters with identified transitory gain. Unlike the

keyword-search approach, the defined-sample approach

analyzes every observation within the defined sample.

Basic statistics within the sample can be more precisely

analyzed, instead of making inferences. Also, by blan-

ket studying observations within the defined sample, this

approach is not subject to the sensitivity of keyword

selection.

This study extends this line of defined-sample approach

to address non-GAAP disclosure issues; the criterion

selected to define the research sample is industry mem-

bership. To contrast the popularity and the interpretation

of non-GAAP disclosures, two industries are selected and

studied: consumer nondurables (SIC code  to )

and business services (SIC code  to ). Industry

membership is defined by the Fama-French -industry

classification. Previous studies have documented that the

business service industry (SIC codes  to ) has the

highest tendency in using non-GAAP financial measures,

while the consumer goods industry (SIC codes  to

) has the lower tendency (Bhattacharya et al., ;

Leung & Veenman, ). However, given the keyword-

search approach employed in these studies, the conclu-

sions are reached through inferences,that is, by looking at

the number of sample observations in each industry rela-

tive to the industry membership distribution of the entire

population in Compustat or I/B/E/S firms (Bhattacharya

et al., ). A more precise statement can only be made

after a blanket study of a specific industry.Using a defined-

sample approach to provide a more comprehensive and

direct comparison, this study selects these two industries

for analysis.

Industry membership potentially has an impact on the

usage and interpretation of non-GAAP disclosures based

on each industry’s attitude toward non-GAAP measures

and the group dynamic in each industry. For example,

the National Association of Real Estate Investment Trusts

(NAREIT) has issued its own policy in an attempt to

standardize the use of non-GAAP financial measures in

their industry (Baik et al., ; Bentley et al., ).

See Appendix Afor NAREIT’s self-regulated definition

of some non-GAAP measures applied in a public firm’s

press release (Spirit Realty Inc., June ). So far, how-

ever, industry effects on non-GAAP disclosures have not