How to perform a Return Event Study with EventStudyTools

Event studies can be difficult to program — for example, if test statistics are needed at the level of grouped events from the overall sample. We built EventStudyTools to simplify conducting such event studies.

EventStudyTools is an online project that hosts analytical capabilities for event studies and makes them available through different channels, such as form-based interfaces on the website, R packages, or application programmable interfaces. For most users, the interfaces on our website are the preferred channel for using our capabilities.

This article describes how to perform a standard return event study right on our website. Since the website host different capabilities, you need to first navigate to the form/interface entitled Abnormal Return Calculator (ARC).

The ARC allows you to set several broad parameters of your event study (e.g., the benchmark model) right on its form. Parameters related to individual events of your event study need to be specified in one of the three files you need to upload to ARC — namely the request file (1). The other two files you need to provide are a file with stock price data (2) and a file with capital market price data (3). Allowed file formats are CSV, XLS, and XLSX.

Let’s shortly discuss the contents of the three files and common questions we receive: The request file (1) should be composed of the following columns with each line representing a new event (please visit our instructions page for the allowed values and formats): Event ID; Firm ID; Market ID; Event Date; Grouping Variable; Start Event Window; End Event Window; End of Estimation Window; Estimation Window Length.

We regularly receive questions about the Grouping Variable. In our sample data, the Grouping Variable holds the value “addition” since the events in our data describe index reconstitution events where companies were added to the S&P 500 index. The variable defines the sub-groups of your overall sample for which you want to have average values calculated — i.e., Average Abnormal Returns (AARs) and Cumulative Average Abnormal Returns (CAARs). Let’s assume the sample data would also cover the companies which were deleted from the S&P 500 when the index was reconstituted. In this case, the lines with these companies should have the value “deletion” as the Grouping Variable value.

The files with stock prices data (2) and capital market data (3) must hold the financial data that is required to perform the event study with the events and estimation and event windows as described in your request file — meaning, closing prices of the stocks and the referenced stock market index or indices. You can compile these files with data from a source of your choice or make use of our firm and market data file compilation service.

After uploading these files to ARC, your event study is performed. This will take several seconds depending on the size of your event study. Upon completion, a total of five files will be prompted to you — as downloads on the form and per email, if you have entered an email address (note: entering an email address is optional, ARC also works without).

Four files contain the results of your analysis, the fifths file (labeled analysis report) holds meta-information about your analysis. It tells you as per each event and respective abnormal return calculations, whether the calculations were performed successfully or not. It also points to reasons why calculations may have failed, such as due to missing financial data.

In some cases, you may receive an error message after uploading your files. In this case, your event study could not be started because either your file format or internal file structure did not match the input requirements of ARC — for error codes and guidance on how to fix the problem, visit our table with error codes on the EventStudyTools website.

We hope that this article helps you to get started using our Abnormal Return Calculator. Once you have compiled your input files, you will find the process convenient and flexible enough to accommodate your research needs.

All the best with your publication!