Systematic Peer Selection : A Comparative Study of Methods Used In Relative Valuation
Sammanfattning: With the purpose to provide insights for practical use and to contribute to an arguably under-served area of valuation research, this study performs a comparative analysis of systematic peer selection methods used in relative valuation. By using U.S. data between 2010-2020 and estimating valuations with the forward P/E multiple, we compare prediction errors of three methods presented in prior research with an additional comparison of the industry classifications SIC, GICS and the previously unexplored TNIC. We find that considering the business factors of industry classification (GICS) and product similarity (TNIC), is most important to improve valuation accuracy. As a result, we encourage future research to expand the use of different business factors in systematic methods. Further, we do not find support of the suggested notion that considering several fundamental factors in peer selection is able to replace the information provided by business factors. With support from theory, our results rather suggest these factors are not comparatively priced across industries. When considering several fundamental factors within industries, we find marginally increasing benefits on accuracy when using a non-linear multivariate method. Lastly, by analysing results over time, we find that since relative valuation is based on market values, its valuation accuracy is dependent on the stability of the overall market.
HÄR KAN DU HÄMTA UPPSATSEN I FULLTEXT. (följ länken till nästa sida)