Read the summary and the most important questions on Event Studies (based on ppt slides)

• 1 Event Studies (based on ppt slides)

• 1.1 Motivation and abnormal returns

  This is a preview. There are 8 more flashcards available for chapter 1.1
  Show more cards here

• Event studies studies

  The effect of an identifiable event on a financial variable

• If you do an event study the event should

  Occur regularly, thus you can't use one event you want to explore some systematic effect of the event

• First thing you should take into account when dealing with event study

  Transform event time data into calendar time data

• After transforming calendar time data into event time data, we should

  Think about our sample. What kind of data do we want to have? Choice is not easy. Small or large firms to include? Domestic or foreign firms?

• We assume that each firm has 1 event in this book notation wise.

  N=J

• The crucial ingredient to do an event study is

  That you'd like to calculate the abnormal return. Because you'd like to separate the impact of the event from all kind of other unrelated movements in the price. All these other unrelated movements we can hardly observe. We cant go back in time and observe without the event. Therefore we have to know the expected return, the return without the effect of the event.

• You can ofcourse include more factors than only the market return

  , maybe some of the Fama French factors, maybe some liquidity, the CARhart model, but this as just explained is the foundation of event studies. Abnormal returns. Having established abnormal returns in the event window, now we can do various tests.

• 1.2 The testing procedure

  This is a preview. There are 11 more flashcards available for chapter 1.2
  Show more cards here

• After testing, secondly we would like to know how to explain a cross-sectional regression

  Which x-variables explain the difference between Coca Cola and apple.

• 5 important aspects: Aspect one: We assume that the abnormal return is uncorrelated in the cross section.  (Ari,t not correlated with Ari,t+1 and if more firms Ari,t not correlated with ARj,t)

  - Reasonable assumption if event windows do not overlap
  - allows us to calculate variances of (cumulative) abnormal returns without taking into account the covariances
  - if they do overlap, possible solution is to put them into portfolios with weights, thus keeping same number of event 
  - this does not work if weights vary heavily by variance 

• Aspect two: the variance of return usually will increase over the event window.

  Boehmer: compute crolssectional variance all firms at particular time period