Event Studies for Your Thesis & Research

A comprehensive guide for students and researchers. From defining your research question to reporting publication-ready results -- we walk you through every step of the event study process.

Used in bachelor's, master's, and doctoral theses worldwide

Workflow

Step-by-Step Guide

Follow these five steps to conduct a rigorous event study for your thesis or research paper.

1

Define Your Research Question

Start with a clear, testable hypothesis. What event are you studying? What is the expected direction of the effect? A well-defined research question determines your entire study design.

  • Formulate a null and alternative hypothesis before collecting data
  • Review prior literature to understand expected effect sizes
  • Define your event precisely -- date, type, and scope
  • Consider whether a short-run or long-run study is more appropriate
2

Collect and Prepare Data

Gather daily adjusted closing prices for all firms and market indices in your sample. Ensure clean data with consistent date formats and no gaps in the estimation window.

  • Use adjusted closing prices to account for splits and dividends
  • Common sources: Yahoo Finance, CRSP, Refinitiv, Bloomberg
  • Check for data quality: missing values, delistings, thin trading
  • Prepare your event list with firm identifiers and exact event dates
3

Choose Your Methodology

Select the expected return model, event window, estimation window, and statistical tests. Your choices should align with your research question and follow established conventions in your field.

  • Market Model is the standard choice for most thesis topics
  • Estimation window: 120-250 trading days before the event window
  • Event window: typically (-1, +1) for short-run, with robustness checks on (-5, +5) and (-10, +10)
  • Report at least one parametric (BMP) and one non-parametric (Rank) test
4

Run the Analysis

Execute your event study using the R package, web app, or Google Sheets template. Start with the example dataset to validate your setup, then run on your actual data.

  • Run the example dataset first to verify your workflow
  • Check the diagnostics: residual plots, Q-Q plots, R-squared
  • Test robustness with alternative models and windows
  • Save all intermediate results for reproducibility
5

Interpret and Report Results

Present your findings with publication-quality tables and charts. Discuss statistical significance, economic significance, and robustness across alternative specifications.

  • Report AARs and CAARs with significance levels (*, **, ***)
  • Include a CAR plot with confidence bands for visual presentation
  • Discuss both statistical and economic significance
  • Address potential limitations: confounding events, thin trading, sample selection

Tools

Which Tool Should You Use?

Compare our three tools to find the best fit for your thesis requirements, coding experience, and study complexity.

FeatureGoogle SheetsR PackageWeb App
Coding RequiredNoYes (R)No
Return Models31510
Test Statistics31211
Multi-Event StudiesNoYesYes
Panel DiDNoYes (6 estimators)No
Synthetic ControlNoYesNo
Power AnalysisNoYesNo
Interactive ChartsNoNoYes
LaTeX ExportNoYesNo
Excel ExportManualYesYes
Best ForQuick single-eventFull researchExploration
PricePaid templateFree (MIT)Free

Our recommendation for thesis work

Start with the Web App to explore your data and get a quick sense of the results. Then switch to the R Package for your final analysis -- it gives you full control, reproducibility, and LaTeX export for your thesis document.

Inspiration

Popular Thesis Topics

Event studies are used across many areas of finance and economics. Here are five popular topics with suggested approaches.

Mergers & Acquisitions

Study the wealth effects of M&A announcements on acquirer and target shareholders. One of the most researched topics in empirical finance with well-established benchmarks.

Example hypothesis

"Acquirer shareholders experience negative abnormal returns around M&A announcements."

Suggested model

Market Model or Fama-French 3-Factor

Earnings Announcements

Analyze how stock prices react to quarterly earnings surprises. Test the post-earnings announcement drift (PEAD) or examine information asymmetry around announcements.

Example hypothesis

"Positive earnings surprises generate significant positive CAARs over (-1, +1)."

Suggested model

Fama-French 3-Factor or Carhart 4-Factor

ESG & Sustainability Events

Measure the market reaction to ESG ratings changes, environmental disasters, sustainability reports, or climate policy announcements. A growing research area with high relevance.

Example hypothesis

"ESG rating downgrades lead to significant negative abnormal returns."

Suggested model

Market Model or Fama-French 5-Factor

Regulatory Changes

Assess the impact of new regulations, policy announcements, or legislative changes on affected industries. Often involves studying an entire sector rather than individual firms.

Example hypothesis

"Banking stocks exhibit negative CARs following announcement of stricter capital requirements."

Suggested model

Market Model with industry-adjusted benchmark

Cryptocurrency Events

Study price reactions to exchange listings, regulatory announcements, protocol upgrades, or security breaches in cryptocurrency markets. Consider 24/7 trading and higher volatility.

Example hypothesis

"Token listings on major exchanges generate significant positive abnormal returns on the listing day."

Suggested model

Market Adjusted or Mean Adjusted (due to limited factor data)

FAQ

Thesis-Specific Questions

Answers to the questions students ask most often when conducting event studies for their thesis.

Which tool should I use for my thesis?

For a bachelor's thesis with a single event type, the Google Sheets template or web app may be sufficient. For master's and doctoral theses requiring robustness checks, multiple models, or advanced methods like DiD, use the R package. The R package also allows you to document your analysis as a reproducible script, which many advisors require.

How many events do I need for a valid study?

For multi-event studies, aim for at least 30 events to ensure parametric tests perform reliably. Studies with fewer than 20 events should rely more heavily on non-parametric tests (Sign Test, Rank Test). Single-event studies are also valid but require longer estimation windows and careful interpretation. Use the power analysis feature in the R package to determine the sample size needed for your expected effect size.

How should I handle firms listed on different exchanges?

Use the relevant local market index for each firm (e.g., DAX for German firms, S&P 500 for US firms). The EventStudy R package and web app allow you to specify different market indices for different events in your event list. Ensure all price data is in the same currency, or convert to a common currency using daily exchange rates.

Can I study events in emerging markets?

Yes, but emerging markets introduce additional challenges: thin trading, non-normality of returns, and market microstructure effects. Consider using non-parametric tests (which do not assume normality) and the Scholes-Williams or Dimson beta adjustments for thin trading. Also check whether your estimation window is free from other major market events.

How do I cite EventStudy in my thesis?

For the R package, run citation("EventStudy") in R to get the BibTeX entry. For the web app, cite the EventStudy website (eventstudy.de) and mention the version used. Also cite the seminal event study methodology papers: MacKinlay (1997), Campbell, Lo, and MacKinlay (1997), and the specific test statistic papers (e.g., Patell 1976, Boehmer et al. 1991) that you use.

My advisor requires reproducible code. What should I use?

Use the EventStudy R package within an R Markdown or Quarto document. This lets you combine your event study code, results, and written analysis in a single reproducible document. Your advisor can re-run the entire analysis by knitting the document. The R package also supports exporting results to LaTeX tables for inclusion in your thesis.

Ready to Start Your Thesis Event Study?

Begin with our documentation to learn the methodology, then choose the tool that fits your thesis requirements.

Ready to run your first event study?

Jump straight into the Event Study App and follow along with your own data.

We use cookies for analytics to improve this site. See our Privacy Policy.