Universität zu Köln

Brunke, Luca (2024). Three Essays on Earnings Forecasting. PhD thesis, Universität zu Köln.

All available versions of this item

• Three Essays on Earnings Forecasting. (deposited 19 Nov 2024 10:46) [Currently Displayed]

PDF Dissertation_Brunke.pdf Download (1MB)

Abstract

Earnings are key indicators of a company’s financial performance as its value depends on future earnings. Accurate earnings forecasts are vital for investors, analysts, and firms to assess future financial prospects (e.g., Azevedo, Bielstein and Gerhart (2021); Tian, Yim and Newton (2021)). Beyond mean earnings, higher moments of future earnings hold significance for stakeholders as they are related to the value of debt and equity (e.g., Merton (1987); Barberis and Huang (2008); Konstantinidi and Pope (2016); Chang, Monahan, Ouazad and Vasvari (2021)). This thesis examines factors affecting the accuracy and bias of analyst and model-based earnings forecasts, focusing on accounting conservatism and earnings management (EM). It emphasizes integrating EM information into models to enhance forecast precision and improve the reliability of implied cost of capital (ICC) as an expected return proxy. Furthermore, it introduces a novel approach to earnings variance forecasting and evaluation methods. Chapter 2 investigates accounting conservatism, defined as “anticipate no profits and provide for all losses” (e.g., Bliss (1924)), and its effects on forecast reliability. Findings reveal that both conditional and unconditional conservatism negatively impact forecast accuracy, increase bias, and widen analyst forecast dispersion for horizons up to three years. Chapter 3 examines the relationship between EM, measured via discretionary accruals (e.g., Dechow, Sloan and Sweeney (1995)), and forecast accuracy. Higher EM reduces forecast precision and impairs ICC reliability. Incorporating EM data improves forecast accuracy and increases the economic relevance of ICCs, evidenced by superior investment returns. Chapter 4 introduces a squared residuals-based method for earnings variance forecasting and compares it with quantile regression approaches (e.g., Konstantinidi and Pope (2016); Chang, Monahan, Ouazad and Vasvari (2021)). While the new method excels at industry-level variance predictions, quantile-based methods perform better at the firm level. Two novel evaluation methods highlight the importance of aggregation levels in variance forecasting. This thesis advances understanding of earnings forecasting by addressing the roles of accounting conservatism, EM, and variance prediction, providing insights for improved financial forecasting models and decision-making.

Item Type:	Thesis (PhD thesis)
Creators:	Creators Email ORCID ORCID Put Code Brunke, Luca luca.brunke@icloud.com UNSPECIFIED UNSPECIFIED
URN:	urn:nbn:de:hbz:38-743639
Date:	2024
Language:	English
Faculty:	Faculty of Management, Economy and Social Sciences
Divisions:	Faculty of Management, Economics and Social Sciences > Business Administration > Finance > Professorship for Business Administration and Corporate Finance
Subjects:	Social sciences Economics Management and auxiliary services
Uncontrolled Keywords:	Keywords Language Earnings Forecasting English Higher Moments English Accounting Conservatism English Earnings Management English Earnings Variance English
Date of oral exam:	30 October 2024
Referee:	Name Academic Title Hess, Dieter Prof. Dr. Schradin, Heinrich Prof. Dr. Hartmann-Wendels, Thomas Prof. Dr.
Refereed:	Yes
URI:	http://kups.ub.uni-koeln.de/id/eprint/74363