We develop a model of the aggregate stock market featuring both dividend-paying and no-dividend stocks within a familiar, parsimonious consumption-based equilibrium framework. We find that such a simple feature leads to profound qualitative implications that support several empirical regularities on the stock market which leading consumption-based asset pricing models have difficulty in reconciling. We show that the presence of no-dividend stocks in the stock market leads to a lower correlation between the stock market return and aggregate consumption growth rate, a non-monotonic and even a negative relation between the stock market risk premium and its volatility, and a downward sloping term structure of equity risk premia. We also find that no-dividend stocks command lower mean returns, but have higher return volatilities and higher market betas than comparable dividend-paying stocks, consistently with empirical evidence. We provide straightforward intuition for all these results and the underlying economic mechanisms at play.