Big Data and Machine Learning in Quantitative Investment by Tony Guida

Author:Tony Guida
Language: eng
Format: epub, pdf
ISBN: 9781119522218
Publisher: Wiley
Published: 2018-11-30T00:00:00+00:00

CHAPTER 8

A Social Media Analysis of Corporate Culture

Andy Moniz

8.1 INTRODUCTION

In today's globalized, service‐based economy, many firms derive substantial value from their intangible assets. Examples include corporate reputation, brand value, innovative efficiency (Chan et al. 2001), human capital (Edmans 2011) and organizational capital. The lack of physical substance associated with intangible assets, their opaque ownership rights and non‐existent market prices limit firms from valuing and recording most types of intangible assets in their financial statements. Until accounting standards change, investors seeking to resolve this ‘value paradox’ and integrate intangible asset valuations into their decision‐making processes must seek alternative sources of information beyond a firm's own financial statements. In our view, one alternative source of information is publicly available text published on the web, and in particular, social media.

The term ‘social media’ describes a variety of ‘new and emerging sources of online information that are created, initiated, circulated and used by consumers intent on educating each other about products, brands, services, personalities and issues’ (Blackshaw and Nazzaro, 2006; Gaines‐Ross 2010). Social media enables individuals to share their opinions, criticisms and suggestions in public. To the best of our knowledge, prior textual analysis studies of social media datasets have mostly captured the perspective of consumers (for example, Amazon product reviews). By contrast, this study seeks to examine a potentially overlooked stakeholder group, namely, a firm's employees. The goal of this study is to describe how mining social media datasets may help investors learn about a firm's corporate culture. This multidimensional concept is typically defined as ‘a set of values, beliefs, and norms of behavior shared by members of a firm that influences individual employee preferences and behaviors’.

For the purposes of this study, we retrieve 417 645 posts for 2237 US companies from the career community website Glassdoor.com and employ computational linguistic techniques to analyze employees' discussions about their firms. The website acts as a forum for employees to provide commentary on the ‘pros’ and ‘cons’ of their firms' cultures for the benefit of potential job seekers. Employee discussions cover a diverse set of topics ranging from perceptions of canteen food, work/life balance, salaries and benefits to views on company strategy and management.

We offer two important contributions to the academic literature. First, we provide a methodology to infer corporate culture from social media. The intangible nature of corporate culture has generated much controversy regarding the creation of a valid construct (Cooper et al. 2001; Pinder 1998; Ambrose and Kulik 1999; O'Reilly et al. 1991). Prior organizational literature either relies upon measures that lack sufficient depth or contain substantial measurement errors (Waddock and Graves 1997; Daines et al. 2010). In recent years, the development of computational linguistics techniques has enabled researchers to automatically organize, summarize and condense unstructured text data and extract key themes from vast amounts of data. Our approach provides a means to infer employee perceptions at a higher frequency and for a broader cross‐section of companies than is possible using traditional survey‐based measures. Second, we contribute to the literature on investors' underreaction to intangible information.

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.