Abstract: The long-term growth forecasts of equity analysts do not have well-defined horizons, an ambiguity of substantial import for many applications. I propose an empirical valuation model, derived from the Campbell-Shiller dividend-price ratio model, in which the forecast horizon used by the "market" can be deduced from linear regressions. Specifically, in this model, the horizon can be inferred from the elasticity of the price-earnings ratio with respect to the long-term growth forecast. The model is estimated on industry- and sector-level portfolios of S&P 500 firms over 1983-2001. The estimated coefficients on consensus long-term growth forecasts suggest that the market applies these forecasts to an average horizon of at least 6 years, and as many as 10 years.

Full paper (238 KB PDF)