FACULTÉ & RECHERCHE

Using a partially revealing dynamic equilibrium model, investors adjust their estimates of the expected returns through the price discovery process (past price dynamics) and consequently implement price contingent portfolios based on these estimates. We implement the price contingent portfolio on the U.S. stock market and compare its performance with other common portfolio strategies. We also consider the price-volume contingent strategy, estimating the expected return and covariance matrix from both the past price and observed volume dynamics. We find that these signal-based portfolios outperform the capitalization and equal weighted strategies. They also provide appealing diversification benefits compared to common optimization-based portfolios.