Steps in Conducting an Event Study

There are several steps involved in conducting an event study. 
flowchart TD
  A[Define the event and sample] --> B(Determine the event window and estimation window)
  B(Determine the event window and estimation window) --> C[Calculate normal returns]
  C[Calculate normal returns] --> D[Compute abnormal returns]
  D[Compute abnormal returns] --> E(Aggregate abnormal returns)
Figure 1

There are several steps involved in conducting an event study:

  1. Define the event and sample: Clearly specify the event of interest and identify the sample of firms or securities affected by the event. The sample should be representative and large enough to draw meaningful conclusions.

  2. Determine the event window and estimation window: The event window is the period during which the event’s impact on security returns is observed. The estimation window is used to estimate the normal, or expected, returns in the absence of the event. The choice of event and estimation windows should be based on the research question and the expected duration of the event’s impact.

  3. Calculate normal returns: Using historical data, estimate the normal returns for each security in the sample, typically using a market model or a factor model (applicable models). This step involves selecting an appropriate benchmark, such as a market index or a matched control group, to estimate the expected returns.

  4. Compute abnormal returns: Calculate the abnormal returns by subtracting the normal returns from the actual returns observed during the event window. The abnormal returns represent the portion of returns that can be attributed to the event.

  5. Aggregate abnormal returns: To draw inferences about the overall impact of the event, aggregate the abnormal returns across the sample of firms or securities. This can be done using various techniques, such as averaging the abnormal