Ranking corporate sustainability: a flexible multidimensional approach based on linguistic variables

• Date: 01 May 2018
• DOI: http://doi.org/10.1111/itor.12469
• Author: V. Liern,B. Pérez‐Gladish
• Published date: 01 May 2018

Intl. Trans. in Op. Res. 25 (2018) 1081–1100

DOI: 10.1111/itor.12469

INTERNATIONAL

TRANSACTIONS

IN OPERATIONAL

RESEARCH

Ranking corporate sustainability: a flexible multidimensional

approach based on linguistic variables

V. Liernaand B. P´

erez-Gladishb

aDpto. Matem´

aticas para la Econom´

ıa y la Empresa, Universitat de Val`

encia, Spain

bDpto. Econom´

ıa Cuantitativa, Universidad de Oviedo, Spain

E-mail: vicente.liern@uv.es [Liern]; bperez@uniovi.es [P´

erez-Gladish]

Received 15 October 2016; receivedin revised form 24 August 2017; accepted 6 September 2017

Abstract

Corporate sustainability implies a compromise between the present environmental, social, and economic

needs of a firm’s stakeholders and their futureneeds. Corporate sustainability is therefore a multidimensional

concept. Nowadays, several independent rating agencies rate firms in terms of environmental, social, and

governance (ESG) criteria. These ratings are usually used by main sustainability indices such as the Dow

Jones Sustainability Index, FTSE4 Good, Stoxx Sustainability Index, or Euronext Vigeo Family to select

companies to invest in. Only those firms performing better than the average of their sector are selected.

However, although providing linguistic ratings about the performance of the firms in individual ESG criteria

with respect to their sector, rating agencies do not usually provide overall ESG rates describing the global

performance of the firms in terms of ESG. In this paper, we propose a flexible operator, linguistic ordered

weighted geometric aggregating operator(LOWGA), which will allowus to define the fuzzy ESG performance

of the firms based on the linguistic labels provided by the rating agencies. Once overall ESG scores have been

obtained, we will use them together with financial criteria to rankthe fir ms in terms of their sustainabilityusing

a suitable multiple criteria decision aid (MCDA) approach, namely, TOPSIS (technique for order preference

by similarity to ideal solution).

Keywords: corporate sustainability; ESG ratings; ordered weighted aggregating operators; linguistic variables; TOPSIS;

ranking

1. Introduction

Corporate sustainability implies a compromise between the present environmental, social, and

economic needs of a firm’s stakeholders (shareholders, employees, clients, pressure groups, commu-

nities, etc.) and their future needs. This supposes maintenance and growth of the firms’ economic,

social, and environmental capital base and the integration of these dimensions in a triple-bottom

line as “economic sustainability alone is not sufficient condition for the overall sustainability of a

C

2017 The Authors.

International Transactionsin Operational Research C

2017 International Federation of OperationalResearch Societies

Published by John Wiley & Sons Ltd, 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main St, Malden, MA02148,

USA.

1082 V.Liern and B. P ´

erez-Gladish / Intl. Trans.in Op. Res. 25 (2018) 1081–1100

corporation” (Dyllick and Hockerts, 2002). Sustainability is therefore, by definition, a multidimen-

sional concept and any attempt to evaluate the firms’ performance in terms of their sustainability

should take several economic and noneconomic (environmental, social, and governance [ESG])

criteria into account.

Nowadays, several independent ratingagencies try to supply transparent and credible information

about the ESG performanceof companies throughout the world. Some examples are the MSCI ESG

STATS (previously known under the name of KLD Research & Analytics, Inc.), Ethibel, Vigeo,

Oekom Research, SAM (Sustainable Asset Management), or EIRIS, which recently merged with

Vigeo. The social rating agencies score firms based on several ESG dimensions and they provide

both, individual scores for each dimension and an overall score. The aggregation process followed

by most of the rating agencies has traditionally relied on arithmetic averages. The use of the geo-

metric mean has been relatively rare in social sciences. However, due to its interesting properties, its

use has increased in the recent years (e.g., in 2010 the United Nations Human Development Index

changed its aggregating methodology with the use of the geometric mean based on its better ability

to reflect the nonsubstitutable nature of the statistics being compiled and compared, see FAQ in

http://hdr.undp.org/en/faq-page/human-development-index-hdi#t292n50). The use of the geo-

metric mean is highly recommended when a low achievement in one considered dimension is not

desired to be linearly compensated by a high achievement in another dimension. This is the case of

social rating agencies that aim at comparing the achievements of companies in terms of different

ESG dimensions to provide overall ESG scores or ratings. The use of the geometric average di-

rectly reflects the poor performance in one dimension reducing the level of substitutability between

dimensions. The main sustainability indices rely on these firms’ overall ESG scores to select com-

panies to invest in (e.g., Dow Jones Sustainability Index, FTSE4 Good, Stoxx Sustainability Index,

or Euronext Vigeo Family). However, most sustainability indices base their inclusion decisions on

the performance of the firms relative to the performance of their sectors, and there is a lack of

information from rating agencies regarding this question. Some rating agencies rate firms based

on their relative performance with respect to their sectors in linguistic terms. Vigeo, for example,

provides a complementary rating of the firms based on sector peers’ comparison. Companies are

classified as leaders, advanced, average, below average, or unconcerned. This classification of the

firms is based on the knowledge of the experts from the rating agency. Nevertheless, firms are rated

in linguistic terms with respect to their relative performance in each individual ESG dimension

and the rating agency does not facilitate any overall relative ESG performance rating. However,

corporate sustainability indices require a global or overall rating of the performance of the firms

compared to the performance of their sectors.

The aim of this paper is to complete the information provided by the ESG rating agencies. In

particular, we propose a method to obtain overall linguistic valuations of the ESG performance

of the firms taking into account the average performance of the sector of the firms. The aggre-

gation method is based on the ordered weighted geometric operator, widely used by academics

and practitioners (Gil-Lafuente and Merig´

o 2012) accommodated to the situation in which the

input arguments are linguistic labels. More specifically, we propose the use of the linguistic ordered

weighted geometric aggregating (LOWGA) operator (Xu, 2004) showing how, given the character-

istics of the aggregation problem addressed in this paper, this method overcomes several important

issues acknowledged by the practitioners and, in particular, by the ESG rating agencies (e.g., to

which extent compensation among very bad and very good performances of a firm in different ESG

C

2017 The Authors.

International Transactionsin Operational Research C

2017 International Federation of OperationalResearch Societies