Event-Study Project Ideas for MFE Students

Event-Study Project Ideas for MFE Students

The best student projects test a clear mechanism on clean data. Here are event-study topics that are tractable, defensible, and quietly impressive.

Why Event Studies Make Strong Projects

An event study has a built-in advantage for a student project: the hypothesis is concrete and the timing is known. Instead of hunting for a vague predictive signal, you measure how prices behave around a specific, dated corporate event. That structure makes the result easy to interpret, easy to defend, and hard to fake, which is exactly what a reviewer wants to see. The methodology, including event windows and abnormal returns, is the backbone of Systematic Event-Driven Trading.

The key is to pick an event where the mechanism is clear, the event is disclosed in filings, and the data is obtainable. The topics below all satisfy those constraints.

Topics That Are Tractable and Defensible

Equity offerings and dilution. Measure abnormal returns around announced offerings. The mechanism, supply hitting the float, is clean, and the events are disclosed in SEC filings. Lock-up expirations. Study the drift around the date insiders can first sell, a classic supply event. De-SPAC float dynamics. Examine post-merger underperformance as redemptions and unlocks reshape the float, an approach detailed in our data-driven approach to shorting de-SPACs.

Index reconstitution. Measure the effect of additions and deletions, a well-studied event with a clear demand shock. Toxic or convertible financing. Study price behavior after dilutive convertible deals. Each of these has a documented mechanism in the book and a dated event you can build a window around.

What Each Topic Needs From Data

Whatever the topic, the same discipline applies: survivorship-free coverage and point-in-time data, as covered in our piece on survivorship bias, plus a defensible source mapped in our guide to market data sources for systematic research.

The Part That Takes the Most Time

For most of these projects, the bottleneck is building the event dataset, turning filings into dated, machine-readable events without lookahead. Alphanume's dilution events dataset does this for financing events, and the historical market cap dataset adds the point-in-time size that scales each event's impact. Starting from a ready event feed lets you spend your time on the study design instead of the data engineering.

Scoping One Idea End to End

Take the equity-offering study as a concrete example. You would define an event window around each announced offering, assemble a survivorship-free universe of issuers including those later delisted, align prices and size to what was known on each date, compute abnormal returns against a benchmark, and then stress the result with realistic borrow and trading costs. Each step maps to a specific data requirement, which is what makes the project tractable in a single term.

Scoping the idea this tightly is what separates a finishable project from an open-ended one. By naming the window, the universe rule, and the cost model in advance, you convert a vague interest in dilution into a bounded study you can actually complete and defend.

Turning an Idea Into a Submission

Once you have chosen a topic, the path to a finished project is short if the data is in hand. Define the windows, assemble the survivorship-free sample, compute abnormal returns, and write up both the effect and its sensitivity to costs. The topics above are deliberately bounded so that this sequence fits a single term, which is part of what makes them strong choices for a student rather than a research career.

A useful habit is to pre-register your own analysis informally, writing down the windows and rules before you look at results. It guards against the temptation to tune the study toward a clean finding, and it gives you a defensible story about why the result is real rather than fitted.

How to Choose

Pick the event whose mechanism you can explain in one sentence, then build a clean, survivorship-free, point-in-time dataset around it. A tractable topic tested honestly impresses more than an ambitious topic tested carelessly, and an event study gives you that honest structure by design.