Date of Award
2025
Degree Type
Dissertation
Degree Name
Doctor of Business Administration (DBA)
Department
Business Administration
First Advisor
Devendra Kale
Abstract
This study investigates the predictability of goodwill impairment occurrence and magnitude with financial data and textual analysis. Goodwill, an unidentifiable intangible asset, plays a pivotal role in financial reporting but remains subjective in valuation. To improve impairment forecasting, the study examines the predictive capabilities of goodwill financial and textual metrics. It also addresses the limitations of the impairment-only accounting model under SFAS 142, which depends on reactive and subjective managerial judgments.
Using longitudinal panel data (2001-2022), the study develops predictive models for goodwill impairments by introducing direct goodwill-factor indicators, including goodwill ratios, keywords, and amortization. A two-dimensional measurement framework integrates cumulative occurrence variables to capture short-term financial distress and cumulative magnitude variables to track long-term financial deterioration through compounded changes.
The results demonstrate that goodwill productivity and intensity ratios and the difference between goodwill amortization and impairment, rather than broad economic and financial performance indicators, are more likely to indicate a more precise foundation and predictive power for goodwill impairment. In contrast, goodwill keywords in 10K show limited predictive utility. Sectoral goodwill impairments and temporal delays significantly moderate the relationship between the key predictors and these outcomes.
The study enhances the predictive model with additional textual analysis, focusing on managerial language in earnings call transcripts beyond goodwill-related texts in 10K filings. Findings reveal that the managerial tone, which is subjective and emotional, is negatively associated with impairment magnitude, indicating more transparent communication to justify a substantial impending impairment. However, constraining words, which are formal and objective language, show an inverse relationship with both impairment occurrence and magnitude. This finding highlights that managers strategically use constraining words to manage stakeholder expectations and protect managerial interests.
This research addresses limitations in current accounting practices and contributes to the literature with a comprehensive framework for robust goodwill impairment prediction models by effectively bridging quantitative and qualitative approaches. The study's findings also offer investors, auditors, and regulators practical implications with improved tools to mitigate financial risks, strengthen financial reporting and enhance disclosure transparency.
Recommended Citation
Ahn, Paul Bumchan, "PREDICTING GOODWILL IMPAIRMENT WITH FINANCIAL DATA AND TEXTUAL ANALYSIS" (2025). Open Access Dissertations. Paper 4432.
https://digitalcommons.uri.edu/oa_diss/4432
