This article explores the use of information-theoretic explanatory variables in stock price forecasting regressions. Stock price changes, as the dependent variable, is run against traditional explanatory variables such as lags of price changes and lags of the squares of the price changes. The addition of lags of a variable related to the information content of the price changes is shown to significantly improve the explanatory power. This exercise should encourage the use of information related variables in forecasting financial markets.