Constructing economic taxonomy reflecting firm relationships based on news reports

Abstract

Purpose

This paper aims to propose a novel approach to constructing an economic taxonomy that demonstrates the complex relationships between firms, which are not fully revealed by traditional industry classification systems such as the NAICS or ICB.

Design/methodology/approach

Based on narrative economic theory, data from CNBC news reports between 01/01/2019 and 03/27/2019 regarding four selected firms, namely, Walmart, Amazon, Netflix and Boeing, were analyzed and coded as the basis to guide the construction of a firm-to-firm relationship taxonomy.

Findings

The relationships between firms are more complex than the simple relationships defined by the traditional classification systems with yes or no in terms of production process (NAICS) or major profit resource (ICB). Based on the sample firms, the authors proposed a four-layer hierarchical taxonomy framework that quantitatively reveals the inherent contradictory relationships between firms, which the authors defined as competition vs consistency. The proposed taxonomy framework is sufficiently flexible to accommodate complex relationships between firms, and it is also adaptable to new information. Under both the competition and consistency categories in the taxonomy model, more detailed subcategories are further coded into two more layers quantitatively to represent the firms' nuanced relationships.

Originality/value

This study provides a novel atheoretical approach to reveal complex firm relationships utilizing narrative text data gathered from news media. The framework of the firm relationship taxonomy constructed in this study provides an alternative and supplementary approach to the classical industry classification systems that can quantitatively specify comprehensive and dynamic connections between firms.

Citation

Zhou, Z., Mu, X. and Lin, X. (2022), "Constructing economic taxonomy reflecting firm relationships based on news reports", Data Technologies and Applications, Vol. 56 No. 1, pp. 132-145. https://doi.org/10.1108/DTA-11-2020-0287