| Course | GB792 Advanced Topics in Finance |
|---|---|
| Unit | Unit 3 |
| Paper type | Event study appraisal |
| Length | About 1,110 words, 7 pages double-spaced |
| Format | APA 7 student paper |
| School | Purdue University Global |
| Program | MBA |
| Updated | October 2026 |
Free sample paper for GB792 Unit 3
Month Zero Is the Wrong Month: A Unit 3 Appraisal of the 1969 Stock Split Event Study
[Student Name]
Purdue University Global
GB792: Advanced Topics in Finance
Unit 3 Event Study Appraisal
[Instructor Name]
[Date]
This appraisal concerns a real published study; its proposed alternatives are the candidate's.
The Study
The study appraised here is Fama et al. (1969), cited in full in the references.
Part One: The Design
The study asked how quickly stock prices adjust to new information, using stock splits as the event. Its sample consisted of 940 splits on the New York Stock Exchange between 1927 and 1959. For each stock, the authors estimated a market model, relating the stock's monthly return to the market's, and used it to compute residuals: the part of each month's return not explained by the market. They then averaged residuals across all splits for each month relative to the split and cumulated them from about thirty months before to thirty months after.
The result became one of the most reproduced figures in finance. Cumulative residuals rose steadily and substantially in the months before the split and were roughly flat afterward. When the sample was divided by what happened to dividends, firms that later raised their dividends showed slightly positive residuals after the split, and those that cut them showed negative ones. The authors concluded that splits are associated with expectations of higher dividends, that the market anticipates this before the split takes effect, and that prices fully adjust by the end of the split month. The paper became the founding example of the event study method.
Part Two: The Event Date
The event in month zero was the month in which the split became effective, not the month in which the board announced it. Splits are typically announced weeks or months before they take effect. Any price reaction to the announcement therefore falls before month zero and is mixed into the pre-split rise.
That choice matters for interpretation. The rise before the split combines two different things: the strong performance that led firms to split in the first place, since firms tend to split after their share prices have risen, and the market's reaction to the split announcement itself. With the effective date as month zero, the study cannot separate them. A reader may see the pre-split rise as the market anticipating good news, when much of it may simply be the run-up in price that caused the split.
If the date changed: using announcement dates would locate the information effect precisely. Grinblatt et al. (1984) did this and found a positive abnormal return of a few percent in the days around split announcements, which suggests that the announcement carried news. The long pre-split rise would then be read mostly as selection: firms split because their prices had risen.
Part Three: The Window
The study used monthly returns and a window of about sixty months. Monthly data were what was available, but they limit what the study could show about speed. A claim that prices adjust fully by the end of the split month cannot distinguish adjustment within a day from adjustment within four weeks. A sixty-month window also allows other events, such as earnings announcements and dividend changes, to fall inside it and contribute to the residuals.
If the window changed: a short window around the announcement, measured in days, would test the speed of adjustment directly and limit contamination by other events. Brown and Warner (1985) showed that daily returns give event studies much greater power to detect abnormal performance, and that simple methods work well with daily data.
Part Four: The Benchmark
The market model was estimated using months outside a period around the split, to avoid having the split's own effects bias the estimates. But the sample was selected on a pattern of high returns before the split. If the period used to estimate the model includes some of those high-return months, the model's intercept will be biased upward, and abnormal returns after the split will be biased downward. If the period excludes them, the model may describe the stock in a period unlike the one being studied.
If the benchmark changed: comparing split stocks with matched non-split stocks of similar size and prior returns would control for the run-up without relying on a model estimated across different periods. That approach would also address whether split stocks behave differently from other stocks with similar past performance.
The study answered a question about the speed of adjustment with an event date, a window and a benchmark that each made the answer harder to read.
Part Five: What Later Work Found
Later research that used announcement dates and daily data confirmed that split announcements carry news, with positive abnormal returns concentrated in a short window (Grinblatt et al., 1984). It complicated the conclusion about complete adjustment. Ikenberry et al. (1996) studied splits from 1975 to 1990 and found positive abnormal returns of several percent in the year after the split, a drift inconsistent with immediate full adjustment and consistent with underreaction to the signal a split conveys. The 1969 study's monthly design, with its effective dates and long window, could not have detected that pattern clearly.
A Summary of Design Choices
The study's choices and the alternatives proposed here can be set side by side. For the event date, the study used the effective month; the alternative is the announcement day. For the window, it used about sixty months of monthly returns; the alternative is a few days of daily returns around the announcement. For the benchmark, it used a market model estimated around the event; the alternative is matched control stocks with similar prior returns. For the sample, it used all splits; the alternative divides them by split ratio and dividend change before testing.
The Sample Itself
One further choice deserves mention. The 940 splits were pooled regardless of split ratio, so a three-for-two split and a four-for-one split counted equally. Larger splits may convey stronger signals, and pooling can hide differences among them. Dividing the sample by ratio before testing would show whether the result depends on the kind of split.
The Verdict
The 1969 study deserves its standing. It introduced a method that, refined, became the main tool for studying how prices respond to information, and its broad conclusion, that prices incorporate news quickly, has been supported by a great deal of later work. Its specific findings about splits are less secure. The rise before splits is better read as selection than as anticipation, and the claim of complete adjustment by the split month rests on data too coarse to test it.
References
Brown, S. J., & Warner, J. B. (1985). Using daily stock returns: The case of event studies. Journal of Financial Economics, 14(1), 3-31. https://doi.org/10.1016/0304-405X(85)90042-X
Fama, E. F., Fisher, L., Jensen, M. C., & Roll, R. (1969). The adjustment of stock prices to new information. International Economic Review, 10(1), 1-21. https://doi.org/10.2307/2525569
Grinblatt, M., Masulis, R. W., & Titman, S. (1984). The valuation effects of stock splits and stock dividends. Journal of Financial Economics, 13(4), 461-490. https://doi.org/10.1016/0304-405X(84)90011-4
Ikenberry, D. L., Rankine, G., & Stice, E. K. (1996). What do stock splits really signal? Journal of Financial and Quantitative Analysis, 31(3), 357-375. https://doi.org/10.2307/2331396
GB792 Unit 3 instructions, in plain terms
The third GB792 assignment asks you to appraise a classic event study by its design choices. Cite the study and restate its design: sample, event definition, return frequency, window and benchmark model. Then take each design choice in turn and explain how it shapes what the result can mean. For each, state what would change if that one choice changed, and support the statement with later research where it exists. Consider how the sample was pooled and whether different kinds of events were mixed together. Summarize the choices beside the alternatives you propose. Close with a verdict that separates the method's lasting contribution from the specific findings that later work revised. Six pages in five parts with a closing summary of choices is a common format in this course.
How this GB792 Unit 3 example is built
The appraisal restates the 1927 to 1959 New York Stock Exchange sample, the market model residuals and the cumulative pattern, including the dividend split. The event date section shows that using effective months folds announcement effects and pre-split run-ups together, and uses Grinblatt and colleagues' announcement returns to separate them. The window section uses Brown and Warner on daily returns. The benchmark section explains how selection on past returns can bias a market model and proposes matched controls, with the highlighted sentence summing up the three choices. Later work, especially Ikenberry and colleagues' post-split drift, follows. A short section on pooling split ratios, a summary of choices and alternatives, and a verdict on the method's legacy close the paper.
GB792 Unit 3 rubric: what earns full points
Points in the GB792 event study appraisal follow the analysis of design choices. Instructors reward a paper that explains how each choice shapes interpretation, rather than listing modern improvements. The event date discussion tends to carry the largest share of points, since it decides whether the result shows anticipation or selection. Statements about what would change under an alternative earn credit when supported by later studies. Understanding why daily data increase power, and how a benchmark can be biased by the selection of events, shows methodological command. Closing with the four choices set beside their alternatives gives the instructor an easy check of the whole argument. The verdict is graded on balance: crediting the method while revising the specific claims.
GB792 Unit 3 help: mistakes that cost points
Many GB792 event study appraisals criticize the 1969 paper for using monthly data, as if the authors had a choice, and stop there. Others miss the difference between effective and announcement dates, which is the appraisal's most important point. Benchmark problems are frequently described in general terms without connecting them to the fact that splits follow price increases. Later research is cited as overturning the study rather than refining it. Students rarely consider whether pooling different kinds of events matters. Restate the design, take each choice in turn, say what would change, support it with later work, consider pooling and give a balanced verdict. Send the citation of the event study on your syllabus, and a writer can work through its event date, window and benchmark.
Get a GB792 Unit 3 example written to your instructions
Share the event study your section assigned, whether a classic like this one or a recent paper, together with the Unit 3 instructions and rubric. What returns appraises the event date, window and benchmark in turn and says how the result shifts under each alternative. Expect it within 24-48h; the first custom sample carries no charge. The paper above is an original model document written by our desk, not a submitted student paper and not an official Purdue University Global document.
More GB792 papers and MBA samples
- GB792 Unit 1: Discussion Board Post
- GB792 Unit 2: Empirical Study Critique
- GB792 Unit 4: Asset Pricing Memo
- GB792 Unit 5: Capital Structure Analysis
- GB792 Unit 6: Seminar Reflection
- GB792 Unit 7: Replication Review
- GB792 Unit 8: Valuation Model Critique
- GB792 Unit 9: Research Design Proposal
- GB792 Unit 10: Finance Research Paper
- GB791 Unit 3: Standard Setting Analysis
- GB740 Unit 3: Entry and Contracting Memo
- GB730 Unit 3: Question Alignment Exercise
- GB720 Unit 3: Institutional Theory Application
GB792 Unit 3 questions, answered
Why appraise the original split study instead of a modern event study?
Because nearly every modern event study inherits its structure, and its constraints are easy to see. Monthly data, an effective date as event zero and a long window make each design choice visible. If your section assigns a recent paper, the same five-part appraisal applies, and newer studies often raise harder benchmark questions, such as factor models over long windows.
Is the appraisal saying the 1969 conclusions were wrong?
No. It argues the conclusions about dividend information hold up well, while the long pre-event rise was mostly selection rather than reaction. That distinction sharpens the original instead of overturning it. A strong appraisal usually reaches a mixed verdict of this kind, stating exactly which inference survives each alternative design and which one weakens under it.
What does a benchmark model do in an event study?
It supplies the return a stock would have earned without the event, so the difference can be called abnormal. The market model does this by fitting each stock's returns against the market over a quiet period. If that period is not quiet, or the model omits a relevant factor, the abnormal return inherits the error, which is why the appraisal checks both.
Where can I find a free GB792 Unit 3 sample paper?
This page holds a full GB792 Unit 3 event study appraisal of the 1969 stock split study, examining its event date, window and benchmark and comparing it with later research.
Why does the choice of event date matter in an event study?
Returns before the date are read as anticipation. If the date is the effective day rather than the announcement, the announcement reaction and the run-up that caused the event are mixed together.