The (Apparent) Usefulness of Brand Values for Predicting Cash Flows and Earnings

Abstract

We evaluate the extent to which brand value estimates (BVE) provided by brand value consulting firms have a predictive ability for brandowner cash flows. Our sample consists of BVE of publicly-listed U.S. firms published by Interbrand, Brand Finance and BrandZ between 2006 and 2021. While we find a positive incremental impact of BVE on future cash flows in in-sample regressions, out-of-sample predictions based on linear regressions and Machine Learning methods show that
BVE have no incremental forecasting power over standard accounting information. An analysis of abnormal returns on zero-cost portfolios of brand owners with high and low forecasted cash flows supports this conclusion. Our findings, which extend to earnings forecasts, contribute to discussions on the accounting treatment and disclosure of intangible assets and the usefulness of externally-provided BVE. Moreover, our results demonstrate the importance of relying on out-of-sample tests when evaluating the incremental forecasting ability of intangibles.

Keywords: Brand value, intangible assets, forecasting ability, out-of-sample, Machine Learning, cash flows, earnings