Using 10-K text to gauge COVID-related corporate disclosure

Area of Science:

• Corporate Finance
• Accounting
• Natural Language Processing

Background:

• The COVID-19 pandemic created significant uncertainty for businesses.
• Effective corporate disclosure is crucial for market participants during crises.
• Existing measures for COVID-related disclosure are inadequate.

Purpose of the Study:

• To develop and validate systematic measures for COVID-19 disclosure intensity.
• To assess the impact of COVID-19 disclosures on corporate liquidity.
• To examine the relationship between corporate governance and COVID-19 disclosure.

Main Methods:

• Utilized natural language processing (NLP) on 10-K reports from U.S. listed firms.
• Developed an overall COVID dictionary and a contextual COVID dictionary.
• Validated dictionaries by analyzing associations with corporate liquidity events.

Main Results:

• The overall COVID dictionary effectively predicts corporate liquidity events.
• Contextual COVID dictionaries, especially those with negative sentiment, also predict liquidity.
• Firms with stronger governance exhibit higher COVID-19 disclosure levels.
• Increased COVID-19 disclosures negatively correlate with short-term market performance (Tobin's Q).

Conclusions:

• Developed reliable NLP-based tools for measuring corporate COVID-19 disclosures.
• Stronger corporate governance is linked to greater transparency, even at a short-term market cost.
• These findings offer insights into corporate communication strategies during global health crises.