0
$\begingroup$

I have a data set of the daily share prices of 163 companies with the time frame between 2009 and 2016.

I have collected 50 different major economic and geopolitic events that happened during this time frame Ex:(fukushima reactor explosion), and I want to select the 3 most important events.

what is the right statistical approach to do this?

I was thinking about regression analysis with dummy variables or factor analysis.

Would appreciate your opinions.

Improve this question
$\endgroup$
1
  • $\begingroup$ I think I am looking for something like "change in stock indices following the event" $\endgroup$
    – Kingindanord
    Commented Jun 14, 2017 at 13:29

1 Answer 1

Reset to default
0
$\begingroup$

Perhaps the most simple approach may be to measure the difference of all 163 share prices before and after the event (say a week before and a week after). You can normalize them so you will have relative change before and after as percentage of before.

Out of this, 50 163-dimensional vectors will be available $x_{1},\dots x_{50}$ with components $x_{1,1}\dots x_{50,163}$. These may be aggregated with respect to what you understand as the importance, e.g:

  • Max average relative change - what increased the prices most
  • Min average relative change - what decreased the prices most
  • Max average absolute relative change - what changed the prices most
  • Span of relative changes - what caused most changes in dynamics among the companies $$\max_{i=1\dots50} \left(\max_{j=1\dots 163 }x_{i,j}-\min_{j=1\dots 163 }x_{i,j}\right)$$
Improve this answer
$\endgroup$
1
  • $\begingroup$ Thanks for the suggestion. what method should I use to measure the "Span of relative changes"? $\endgroup$
    – Kingindanord
    Commented Jun 14, 2017 at 13:56

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.