Equity valuation using accounting numbers : empirical analysis on different valuation methods estimates and how trimming affects explanatory power

• Autor(a) principal: Afonso, Fábio Joaquim Gonçalves
• Data de Publicação: 2021
• Tipo de documento: Dissertação
• Idioma: eng
• Título da fonte: Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)
• Texto Completo: http://hdl.handle.net/10400.14/36797
• topic: Valuation; Discounted dividend model; Discounted cash flow model; Residual income valuation model; Multiples based model; R&D intensity; R&D expenditures, and trimming; Avaliação financeira; Domínio/Área Científica::Ciências Sociais::Economia e Gestão

This dissertation aims to investigate the performance of the Dividend Discount Model (DDM), the Residual Income Valuation Model (RIVM), the Discounted Cash-Flow Model (DCFM), and the multiples-based models (MBM), specifically the Forward Price to Earnings (P/E), when analysing a large sample analysis of US-listed firms. First analysis on the results indicates MBM as the most accurate model. A sensitivity analysis exposed the RIVM as the model most sensitive to changes in assumptions. Further sensitivity was made on the sample selection process where trimming was tested and proved to be reliable. Additionally, the role of Research and Development (R&D) was analysed as a possible reason for the difference between the outputs on the different valuation models, and how low-high-intensive R&D companies provide accurate intrinsic values. The analysis based on smaller samples for low-high R&D intensive firms found that high-intensive firms provided more accurate estimates. Further analysis on a small sample was conducted, where the investigation on the valuation methods used by analysts in equity valuation was compared with what literature suggests and with the large sample analysis results. The analysis on those reports indicated the MBM as the most essential for analysts on their valuations. After the multiples, DCFM was the model that comes up to be used more times in the reports. The analysis shows that the multiples model is the best model after both analyses of the large sample and analysts' reports.