Addressing Puzzle About Equity Valuation Using Multiples: How Earnings Forecasts Outperform Residual Income Model in Multiples

Abstract

Since Liu, Nissim and Thomas (LNT, 2002), researchers have been perplexed by how simple earnings forecasts using multiples, apparently outperform the theory-based residual income model in terms of pricing error. This paper explains mathematically how LNT (2002) find this curious result and demonstrates that, in terms of pricing error, the majority of residual income models in fact outperform earnings forecasts using multiples. The explanation for the LNT (2002) result is in their selection of comparators: they choose residual income models that perform the worst among residual income models, and compare them with the best performing multiples. In terms of future return generation, this paper reports that the majority of residual income models again outperform earnings forecasts using multiples, further supporting the superiority of theory-based valuation models to rule-of-thumb based models in price estimation and future return generation. The paper resolves a decade-old puzzle in equity valuation and demonstrates that theory-based valuation models are empirically superior to rule-of-thumb based valuation models.