In smart beta, we find that factor returns—net of changes in valuation levels—are much lower than recent performance suggests. In fact, many of the most popular new factors (some 458 at last count) have succeeded solely because they have become more expensive.