In this article, Hui Gho shows that, if stock volatility follows an AR(1) process, stock market returns relate positively to past volatility but relate negatively to contemporaneous volatility in Merton’s (1973) Intertemporal Capital Asset Pricing Model. The model helps explain the recent finding that stock market volatility drives out returns in forecasting real gross domestic product growth because the predictive power of returns is hampered by their positive correlation with past volatility. If the positive relation between returns and past volatility is controlled for, however, the author finds that volatility provides no additional information beyond returns in forecasting output in the post-World War II sample.