Calendar-Time Portfolios for Long-Term Event Studies

We discuss long-term event studies in finance, which assess how specific events impact asset prices over extended periods, often years. It highlights their use in evaluating investment value and the effectiveness of strategies, such as those based on insider behavior. We focus on the calendar-time portfolio (CTIME) approach, also known as the Jensen's alpha method, explaining how it constructs and analyzes rolling portfolios of firms that have experienced an event. This methodology is contrasted with the buy-and-hold abnormal return (BHAR) method, with arguments presented regarding the statistical reliability of the CTIME approach due to its handling of cross-sectional dependence. We outline the two main steps of the CTIME calculation: determining average excess returns of the portfolio and performing a time-series regression of these returns on various risk factors to find the alpha, which represents abnormal return. Different risk-factor models, such as the one-factor, three-factor, and four-factor models, are discussed as options for this regression analysis. Finally, the text briefly mentions calculating long-short portfolios and touches upon criticisms of the CTIME method related to managers' potential timing of corporate events.