Testing the Dividend Discount Model: Comparative Evidence on Accuracy, Stability, and Sensitivity

Abstract

This review examines the empirical performance of the Dividend Discount Model (DDM) in terms of three key dimensions: accuracy, stability, and sensitivity. The DDM, widely used in equity valuation, estimates the present value of expected future dividends. While its theoretical appeal is strong, real-world applications often highlight significant challenges. In the “accuracy” dimension we aggregate evidence from multiple markets that compare intrinsic values derived from DDMs with realised market prices or subsequent returns. The “stability” dimension explores how DDM performance varies across time, markets and firm types, investigating whether the model’s error metrics remain consistent. The “sensitivity” dimension analyses how small changes in input assumptions—particularly dividend growth rate (g) and discount rate (r)—can produce large swings in value estimates. In the comparative evidence reviewed, mature, dividend-paying firms in stable markets show relatively better accuracy and stability, whereas growth-oriented or irregular–dividend firms display larger deviations and instability. Sensitivity analyses reveal that a ±1 % variation in assumed discount or growth rate commonly yields ±20 % or more change in valuation, underlining the fragility of the model’s output. Based on this review, the DDM remains a useful heuristic for income-oriented, dividend-stable companies but is less robust as a universal valuation tool. The model’s heavy reliance on precise inputs and the observed variation in accuracy and stability suggest caution: analysts should treat DDM outputs as ranges rather than point estimates, and complement them with other valuation frameworks (e.g., free-cash-flow or comparables). The review also identifies research gaps — notably, insufficient comparative empirical evidence across emerging vs developed markets, and limited study of multi-stage DDMs in volatile firms.