EMPIRICAL METHODS
IN FINANCE – PART 2:
SUMMARY
@ECOsummaries
→ 20% discount
1
,Table of contents
Topic 1_________________________________________________page 3-9
Topic 2_________________________________________________page 10-16
Topic 3_________________________________________________page 17-23
Topic 3_________________________________________________page 24-31
Topic 4_________________________________________________page 32-37
Topic 5_________________________________________________page 38-42
2
,Topic 1 – Event studies
Efficient market:
Efficient market hypothesis (EMH): prices always fully reflect available and relevant information
- Prices only change when new information arrives.
- Information, by definition, cannot be predicted ahead of time.
→ Therefore, price changes cannot be predicted ahead of time.
Efficiency: speed and quality (direction & magnitude) of the adjustment.
Weak market efficiency:
- Prices reflect all information of the past.
→ Test for return predictability.
Semi-strong market efficiency:
- Prices reflect all information of the past + published information.
- Adjustments are direct & complete.
→ Use event studies.
Strong market efficiency:
- Prices reflect all information that might be relevant.
→ Test for insider information.
Event studies:
Event study: test of the change in stock prices around specific events
Example: earning reports, merger announcements, etc.
Intuition: examine the impact of an event on wealth of security holder and to test the efficiency
of the market.
- Announcement is more important than the actual event, as with the announcement the
information becomes public, and investors can react accordingly.
Conducting an event study – steps:
1. Identify the event of interest and its timing.
2. Specify a benchmark model for normal stock returns.
3. Analyse abnormal returns around the event date.
3
, Timing:
Estimation window: time period that
represent ‘normal’ times of the stock.
Event window: time period before and
after event.
Event: always at t=0
Post-event window: within event window,
after the event date.
Abnormal returns:
Idea: an event study measures the abnormal returns around an event.
→Abnormal returns = realized return – normal return
Normal return: returns expected in a normal situation without any event.
Benchmark models for normal return:
1. Own average return (mean adjusted)
* Inaccurate measure, noisy measures
* Normal return: note: T2 – T1 + 1 is for the # of days
2. Market return (market adjusted)
* Normal return:
3. Market model:
* Normal return = αi + βiRmt
* Abnormal return = εit
* Idea: estimate the average return and coefficients during the estimation window.
→ Use these estimations to construct normal/abnormal returns in the event window.
* Normal return:
4. CAPM: market model + restriction on alpha
* CAPM =
* Alpha restriction =
* Note: using macro events is dangerous with a market model / CAPM, as the whole market is
affected by these events instead of just the individual company.
* Normal return:
4
IN FINANCE – PART 2:
SUMMARY
@ECOsummaries
→ 20% discount
1
,Table of contents
Topic 1_________________________________________________page 3-9
Topic 2_________________________________________________page 10-16
Topic 3_________________________________________________page 17-23
Topic 3_________________________________________________page 24-31
Topic 4_________________________________________________page 32-37
Topic 5_________________________________________________page 38-42
2
,Topic 1 – Event studies
Efficient market:
Efficient market hypothesis (EMH): prices always fully reflect available and relevant information
- Prices only change when new information arrives.
- Information, by definition, cannot be predicted ahead of time.
→ Therefore, price changes cannot be predicted ahead of time.
Efficiency: speed and quality (direction & magnitude) of the adjustment.
Weak market efficiency:
- Prices reflect all information of the past.
→ Test for return predictability.
Semi-strong market efficiency:
- Prices reflect all information of the past + published information.
- Adjustments are direct & complete.
→ Use event studies.
Strong market efficiency:
- Prices reflect all information that might be relevant.
→ Test for insider information.
Event studies:
Event study: test of the change in stock prices around specific events
Example: earning reports, merger announcements, etc.
Intuition: examine the impact of an event on wealth of security holder and to test the efficiency
of the market.
- Announcement is more important than the actual event, as with the announcement the
information becomes public, and investors can react accordingly.
Conducting an event study – steps:
1. Identify the event of interest and its timing.
2. Specify a benchmark model for normal stock returns.
3. Analyse abnormal returns around the event date.
3
, Timing:
Estimation window: time period that
represent ‘normal’ times of the stock.
Event window: time period before and
after event.
Event: always at t=0
Post-event window: within event window,
after the event date.
Abnormal returns:
Idea: an event study measures the abnormal returns around an event.
→Abnormal returns = realized return – normal return
Normal return: returns expected in a normal situation without any event.
Benchmark models for normal return:
1. Own average return (mean adjusted)
* Inaccurate measure, noisy measures
* Normal return: note: T2 – T1 + 1 is for the # of days
2. Market return (market adjusted)
* Normal return:
3. Market model:
* Normal return = αi + βiRmt
* Abnormal return = εit
* Idea: estimate the average return and coefficients during the estimation window.
→ Use these estimations to construct normal/abnormal returns in the event window.
* Normal return:
4. CAPM: market model + restriction on alpha
* CAPM =
* Alpha restriction =
* Note: using macro events is dangerous with a market model / CAPM, as the whole market is
affected by these events instead of just the individual company.
* Normal return:
4
