Event study and basics Flashcards
Name 3 ways to deal with outliers
Data transformation (e.g. logs)
Winsorizing (replace extreme values with upper/lower cutoff value (typically 1st and 99th
percentile values)
Truncating = deleting extreme observations
- After treating outliers, recalculate descriptive statistics
- None of these are applied to returns!
4 Important notes on Event Date
- Firms may announce their events at strategic moments, e.g., when the stock price is high > shows up in abnormal returns before the announcement.
- Stock prices may react before the event due to information leakage
- Firms may announce their events jointly with other announcements > confounding events
- But there can also be a delay in stock price reaction due to illiquid markets, more time needed to process information, limits to arbitrage, misclassified timezones.
Name 3 ways to compute the expected returns
Constant-mean-market model = historical average return
Market-adjusted-return model = market return as proxy for normal return
Market model
The choice of estimation window is a tradeoff between precision and timeliness
What are three benefits of taking the logarithm?
- taking a logarithm can often help to rescale the data so that their variance is more constant, which overcomes a common statistical problem known as heteroscedasticity, discussed in detail in Chapter 5.
- Second, logarithmic transforms can help to make a positively skewed distribution closer to a normal distribution
- Third, taking logarithms can also be a way to make a non-linear, multiplicative relationship between variables into a linear, additive one. These issues will also be discussed in some detail in Chapter 5.
inverse of an exponental
As part of an event study, you estimate a regression model explaining AR around the
announcement of a merger. The AR is regressed on a dummy equal to 1 when 2 firms are in
the same industry and 0 when they are in different industries. H0 states that the performance
of mergers where both firms are in the same industry does not differ from the performance of
mergers where the companies are from different industries. The estimated coeff is -0.52 (in %)
with a t stat of -3.5. Which statement is correct?
1) H0 is rejected
2) we have statistical evidence that same industry mergers perform better
- Statement 1 is true
- Statement 1 and 2 are true
- Statement 2 is true
- None is true
- As part of an event study, you estimate a regression model explaining AR around the
announcement of a merger. The AR is regressed on a dummy equal to 1 when 2 firms are in
the same industry and 0 when they are in different industries. H0 states that the performance
of mergers where both firms are in the same industry does not differ from the performance of
mergers where the companies are from different industries. The estimated coeff is -0.52 (in %)
with a t stat of -3.5. Which statement is correct?
1) H0 is rejected
2) we have statistical evidence that same industry mergers perform better
1. Statement 1 is true
- What is the most interesting time window in an event study?
- The post-event window
- The event window
- The estimation window
- The announcement window
menti
- What is the most interesting time window in an event study?
B, the event window
Why does an event study focus on abnormal return?
- To adjust for serial correlation in stock returns
- To control for the gradual impact of an event
- To control for overall market movements
- To improve the standard errors
Why does an event study focus on abnormal return?
C to control for overall market movements
Why do we often need a cross-sectional analysis in an event study?
- Because it helps to understand the sources of the abnormal returns
- Because it helps to understand how abnormal returns vary across firms
- Because it helps to understand how abnormal returns depend upon characteristics of the event
- Because aggregate abnormal returns may be zero, despite clear announcement effects
all correct
Event study have a long history in financial research. In an event study, scholars investigate
patterns in AR around an event date. Which is correct?
1. In a proper event study, it is important that there are no confounding events
2. In a proper event study, it is important that the announcement of the events takes place
during trading hours
3. In a proper event study, we need to be able to separate the sample into good news and
bad news firms
4. In a proper event study, a single firm cannot experience multiple events
Event study have a long history in financial research. In an event study, scholars investigate
patterns in AR around an event date. Which is correct?
1. In a proper event study, it is important that there are no confounding events
One of the potential problems in an event study is that of event clustering: the event dates
for many firms are very close together. Which way of calculating AR would suffer most from
this issue?
1. Return in deviation from the historical average of the same firm
2. Return in deviation from the industry average
3. Return in deviation from the market return
4. Return in excess of those predicted by the market model
One of the potential problems in an event study is that of event clustering: the event dates
for many firms are very close together. Which way of calculating AR would suffer most from
this issue?
- Return in deviation from the historical average of the same firm
Name 5 limitations / drawbacks of event studies
- We assume that the benchmark model is correct. Otherwise AR will be incorrect
- We assume that event windows of firms do not overlap. Otherwise there’s EVENT CLUSTERING and the observations are not independent across securities. The var(avgCAR) is underestimated. Solution = form portfolios.
- We assume that a firm’s beta remains constant after the announcement
- Choice of event window and estimation window is a bit arbitrary
- W assume abnormal returns to follow a normal distribution > use non-parametric tests
In an event study the focus is on ‘abnormal’ returns. What is NOT an appropriate definition of an
abnormal return?
Return minus the return predicted by the market model.
Returns minus the historical average of the same firm.
Return minus the market return
Return minus the average return in the event window of the same firm.
In an event study the focus is on ‘abnormal’ returns. What is NOT an appropriate definition of an
abnormal return?
Return minus the average return in the event window of the same firm.
Consider the excess returns on a stock (r_rf), and the excess returns on the market (rm_rf), observed over a period of 300 months. We estimate a CAPM regression, explaining r_ rf from rm_ rf and a constant. The estimated intercept term is 0.02 with a standard error of 0.04. The estimated slope coefficient is 1.10, with a standard error of 0.10. The R2 is 0.90. Assume we are interested in testing the null hypothesis that the slope coefficient (market beta) of this stock is 1, with a 95% confidence.
What is the appropriate t-stat?
In an event study we often see stock prices moving in the expected direction, even before the event has takes place. What is a potential reason for this?
The event is anticipated by the market.
Assume we wish to investigate whether the stock market crash of October 1987 has fundamentally
changes the risk-return relationship gives by the CAPM. Consider monthly excess returns on a stock (r_rf), and let d=1 since October 1987 and d=0 before. Which regression model would you use to test whether the stock’s market beta has changed since October 1987?
Regress r _ rf upon rm _ rf, d, and d interacted with rm_rf.
