Using Textual Sentiment in Key Audit Matters Reporting as a Proxy for Financial Health
- Autores: Orathai Dusadeedumkoeng
- Directores de la Tesis: Juan Luis Gandía Cabedo (codir. tes.), David Huguet Benavent (codir. tes.)
- Lectura: En la Universitat de València ( España ) en 2026
- Idioma: inglés
- Número de páginas: 375
- Tribunal Calificador de la Tesis: Begoña Giner Inchausti (presid.), Belén Gill de Albornoz Noguer (secret.), Carlos Serrano Cinca (voc.)
- Programa de doctorado: Programa de Doctorado en Contabilidad y Finanzas Corporativas por la Universitat de València (Estudi General)
- Calificación: Sobresaliente
- Materias:
- Enlaces
- Resumen
This doctoral thesis examines whether Key Audit Matters (KAMs) disclosed in audit reports can serve as an early warning signal regarding the financial health of companies. The study has three objectives: to identify the factors influencing the number and type of KAMs; to analyse how auditors express risks through their textual characteristics; and to assess whether KAMs, both quantitative and textual, can be used as relevant indicators of companies’ financial health.
The analysis is based on an unbalanced panel dataset comprising 3,052 firm-year observations and 5,687 KAM-level observations corresponding to 682 listed and delisted companies on the Stock Exchange of Thailand (SET) and the Market for Alternative Investment Market (mai) during the period 2016–2020. Readability is measured using two widely used indices: the Fog and Flesch indices, whilst sentiment is captured through lexical analysis based on the Loughran and McDonald dictionaries. Readability is measured using the Fog and Flesch indices, whilst sentiment is assessed using the Loughran and McDonald dictionaries. Financial health is estimated using Altman’s Z-score. Descriptive, univariate and correlation analyses are employed, as well as regression models, including a GMM estimator to address potential endogeneity issues.
The results of the study show that, on average, the auditors disclose approximately two KAMs per year, mainly relating to matters of high uncertainty such as revenue recognition, inventory, impairment, leases and financial instruments.
Furthermore, auditors tend to use complex language and a relatively negative tone, particularly in KAMs concerning accounting-level risks. The difficulty in reading the text is associated with a more negative tone, and these characteristics vary between firms. The number, type and wording of KAMs are influenced by business factors such as size, complexity and risk, as well as by audit characteristics such as firm size, experience, type of opinion and the auditor’s gender. The evidence suggests that firms with weaker financial health present more KAMs, particularly those related to impairment, asset valuation, leases, internal controls, litigation and going concern. A more negative tone is significantly associated with greater financial problems, supporting the idea that auditor communication is a relevant indicator in emerging markets.
