This paper examines whether the "diversification discount" can be attributed to diversification itself, using recent econometric developments about causal inference. The author, Belén Villalonga, from the University of California, Los Angeles, matches diversified and specialized firms on the propensity score—predicted values from a probit model of the propensity to diversify—to unbiasedly estimate the value effect of diversification. The study finds that when a more comparable benchmark based on propensity scores is used, the diversification discount disappears or even turns into a premium. This suggests that the diversification discount may be caused by factors other than diversification itself, such as pre-diversification characteristics or firm performance. The paper also discusses the theoretical and empirical puzzles surrounding the diversification discount and provides a detailed methodology for constructing comparable control groups using propensity scores.This paper examines whether the "diversification discount" can be attributed to diversification itself, using recent econometric developments about causal inference. The author, Belén Villalonga, from the University of California, Los Angeles, matches diversified and specialized firms on the propensity score—predicted values from a probit model of the propensity to diversify—to unbiasedly estimate the value effect of diversification. The study finds that when a more comparable benchmark based on propensity scores is used, the diversification discount disappears or even turns into a premium. This suggests that the diversification discount may be caused by factors other than diversification itself, such as pre-diversification characteristics or firm performance. The paper also discusses the theoretical and empirical puzzles surrounding the diversification discount and provides a detailed methodology for constructing comparable control groups using propensity scores.