Spanning Tests for Replicable Small-Cap Indexes as Separate Asset Classes

Abstract

Empirical tests of different asset combinations show that the composition of a benchmark portfolio determines whether a replicable G–7 small-cap portfolio can expand the original efficient frontier. Interaction among all assets in a portfolio is key to the effectiveness of a small-cap index in efficient portfolios, and constraints do not always reduce diversification benefits of the small-cap assets. Only a few small-cap portfolios of G-7 countries appear to behave as separate asset classes with portfolio performance-enhancing characteristics when an investor benchmarks these portfolios against the U.S. equity market or an international large-cap portfolio.