Is “trading up” in the NFL and NHL drafts a winning strategy?

We find that NFL teams that trade lower draft picks to acquire higher ones lose in player performance, whereas NHL teams break even. Pre-sample regressions of performance on pick position predict NFL teams to lose 98% and NHL teams to lose 69% of those trades. Our simple model predicts both the short- and long-run outcomes in the NFL. However, GMs who trade up do better than we predict and better than trades involving adjacent picks in the draft would have done, strongly suggesting that the teams that do trade up possess additional skills in scouting and developing players. Avoiding the trades, our model deems ‘inadvisable’ would have collectively saved NHL teams the equivalent of one Hall of Fame career and NFL teams about 30 times the career of Barry Sanders.