Corporate Transparency: Unstructured Soft Information, Gossip, and Fake News
Van Lent studies the effects of soft information on corporate transparency. He uses machine-learning technology to explore how partly standardized and unstandardized qualitative information is disseminated by firms and organizations, and how market participants process and act on this information. The project examines how users perceive information and thereby influence market-wide transparency. A key objective of the project is to explore to what extent the market mechanism is able to process qualitative information so that rumors, biased narratives and resulting biased beliefs are revised.
-
Research Question
How does soft information affect transparency, and how does transparency of soft information affect price discovery in equity markets as well as product and factor markets?
-
Research Motivation
Corporate transparency is often discussed in the context of financial statement information released by companies in response to legal requirements or market demands. While information linked to financial statements is perhaps the form of transparency that gains the most attention in the market place, it is also clear that corporate financial reports are just one of many information sources emanating from firms that are used by investors and other stakeholders. Hence, the full picture of corporate transparency requires a careful examination of (1) the disclosure channels of soft information, (2) their unique characteristics compared to financial statement disclosures, (3) the market response to soft information, and (4) the ultimate consequences of soft information based transparency in financial, factor, and product markets.
-
Research Program
This project studies the effects of soft information on corporate transparency. We use machine-learning technology to explore how partly standardized and unstandardized qualitative information is disseminated by firms and organizations, and how market participants process and act on this information. We examine how users perceive information and thereby influence market-wide transparency. A key objective is to explore to what extent the market mechanism is able to process qualitative information so that rumors, biased narratives and resulting biased beliefs are revised. This project aims at refining machine learning techniques to collect data, categorize data in economically-meaningful classes, and analyze both text and non-text data to overcome the empirical challenges posed by working with soft information.
-
Research Contribution
Our project helps to build a better understanding of corporate transparency in a time when traditional forms of disclosures (through annual reports) appear to lose ground.
Result
The project has produced firm-level measures of risk and sentiment for US publicly listed firms. These measures are designed in particular to capture political risk, including specific topics such as health care, environment, and economic policy. The data are made publicly available on www.firmlevelrisk.com and the machine learning algorithm used to generate the measures is part of the replication package which is available on https://doi.org/10.7910/DVN/OBNRBP.