According to French and Roll (1986), the three sources of market volatility are public information, private information, and error in estimating the value of public information, which is known as “pricing error”. Even though pricing error can generate a significant amount of stock volatility and cause peculiar patterns of short-term price movement, there is little analysis on the effects of pricing error. This paper models a price discovery process that investors reduce the pricing error of a stock by consulting the prices of other stocks affected by the same information (i.e., “peer stocks”). The model shows that a stock’s pricing error and volatility are decreasing in peer stock trading activity as investors learn more from a larger peer stock price data. I test the implication of the model using NASDAQ stocks and find that higher peer stock activity is correlated with smaller pricing error and lower volatility. This learning framework is useful for understanding some empirical patterns of short-term price movement, such as high volatility in market opening prices and cross-sectional mean-reversion of stock returns.