Estimating Beta for unlisted accounting firms with financial metrics

Abstract

This thesis examines the estimation of systematic risk in unlisted firms by comparing three beta estimation methods: Accounting Beta, Fundamental Beta, and Bottom-up Beta. The study focuses on ten large unlisted accounting firms operating in Finland between 2014 and 2023. The aim is to assess the accuracy and practical applicability of each method in the absence of market-based return data.

The research addresses a real-world need to evaluate firm-specific risk where public share prices are unavailable. Three research questions guide the analysis, focusing on methodological differences, theoretical foundations, and practical usability. Methodologically, the study combines financial statement analysis, cross-sectional regression, and capital structure adjustments, implemented through a structured spreadsheet model. Peer data from eight listed firms and sector-level benchmarks from Professor Aswath Damodaran's datasets were used for calibration and validation.

The findings show that Bottom-up Beta produced the most consistent and theoretically grounded estimates. Accounting Beta was highly sensitive to annual accounting anomalies, while Fundamental Beta yielded more balanced outcomes but remained dependent on model specification and peer comparability.

The study concludes that systematic risk can be estimated with reasonable reliability for unlisted firms, provided that the estimation method is carefully matched to the available data and its inherent limitations are acknowledged. The results contribute to the field of private firm valuation by offering a structured comparative framework for selecting an appropriate beta estimation model.