The paper presents a new model for analyzing changes in stock market volatility, called exponential ARCH. This model allows for a correlation between returns and volatility changes, removes the need for certain constraints on parameters, and provides a clear understanding of how shocks to volatility persist over time. Compared to the commonly used GARCH model, this new approach is considered an improvement. The researchers applied this model to study volatility changes and risk premiums in the stock market from 1962 to 1987.