Page 252 - Contributed Paper Session (CPS) - Volume 3

CPS2011 Dominique H. et al.

Using SOM-based visualization to analyse the
financial performance of consumer discretionary
firms
1
1
1
Dominique Haughton 1,2,3 , Zefeng Bai , Nitin Jain , Ying Wang
1 Bentley University
2 Université Paris 1 (SAMM)
3 Université Toulouse 1 (TSE-R)

Abstract
This paper analyzes financial ratios of 27 consumer discretionary firms listed
on the S&P 500 over an eleven-year period from 2006-2016. It adopts a two-
step approach wherein first a confirmatory factor analysis (CFA) on the
financial time-series is conducted and the resulting constructs’ scores are then
used to perform a cluster analysis using self-organizing maps (SOMs). The
consumer discretionary sector is considered an economic and stock market
predictor. It consists of non-essential goods and services which in an economic
slump are more likely to be foregone. The suggested approach is expected to
be a useful reference guide to help understand the past performance of inter-
and intra-sector companies. It also enriches the body of literature on the
application of machine learning techniques to the analysis of firm- and
sectoral-level performance.

Keywords
Consumer Discretionary Sector; Clustering; Financial Ratios; Self-Organizing
Maps; Time series

1. Introduction
The advent of machine learning has lent a new dimension to the analysis
of financial and accounting ratios. A growing body of research integrates
machine learning techniques – both supervised and unsupervised – to bring
out useful insights from these ratios, beyond the traditional approach to
analysing financial ratios. This paper contributes to this research by proposing
a two-step dynamic process to facilitate understanding firms from the
consumer discretionary sector in the US.
This paper considers the consumer discretionary firms due to the inherent
nature of this sector. It consists of goods and services that are not essential
but are desirable if income is sufficient to purchase them. Therefore, lower
stock values in this sector, which includes durable goods, apparel,
entertainment and leisure, etc., can be considered as a signal to an economic
slump. Such stock tend to outperform other sectors’ stock during strong
economic times and underperform them during an economic slump. It is

241 | I S I W S C 2 0 1 9