Serving the Quantitative Finance Community

 
User avatar
burnrate00
Topic Author
Posts: 12
Joined: July 3rd, 2015, 7:08 am

Modelling M&A performance

June 22nd, 2017, 8:35 am

Hello! I am trying to compare the performance of M&A transaction between two sectors ( sector A buys sector B companies vs sector B buys sector A companies). I want to do that from a shareholder perspective, so I thought I could use an event study approach and calculate the excess return for shareholders after the transaction. Ideally, I would like also calculate premiums for these transactions and use a factor model to explain such premiums. 
How large should be the pool of transactions? Would a simple market model be enough to isolate excess return? What basic mistakes should I avoid?

Thanks a lot 
 
User avatar
Alan
Posts: 2958
Joined: December 19th, 2001, 4:01 am
Location: California
Contact:

Re: Modelling M&A performance

June 24th, 2017, 6:29 pm

Hello! I am trying to compare the performance of M&A transaction between two sectors ( sector A buys sector B companies vs sector B buys sector A companies). I want to do that from a shareholder perspective, so I thought I could use an event study approach and calculate the excess return for shareholders after the transaction. Ideally, I would like also calculate premiums for these transactions and use a factor model to explain such premiums. 
How large should be the pool of transactions? Would a simple market model be enough to isolate excess return? What basic mistakes should I avoid?

Thanks a lot 
Interesting idea. Off-hand, one objection (say from a referee) might be that the short-term reaction of the market (as measured in an event study), might not reflect the longer-term results from the transaction. 

Since it sounds like a student project, one advice would be to spend a *lot* of time on the design issues with both your adviser and knowledgeable people in your school. A lot of it will also depend on the databases you have access to, so you might want to nail that down first. 

Another  would be to study very carefully related articles in the Journal of Finance: probably the gold standard in terms of quality for this kind of thing.   
 
User avatar
burnrate00
Topic Author
Posts: 12
Joined: July 3rd, 2015, 7:08 am

Re: Modelling M&A performance

June 26th, 2017, 2:49 pm

Off-hand, one objection (say from a referee) might be that the short-term reaction of the market (as measured in an event study), might not reflect the longer-term results from the transaction
Thanks for your answer! Well this is one of my concerns. As I was trying to gain some knowledge on event studies, I found out that one can actually extend somehow the analysis to keep track of longer results. 
At the moment, I am starting to define a time span for the transactions to consider - I thought maybe 5 years are enough - and adjust that according to the number of transactions I find. My objective is to analyse at least 150 transactions for both clusters, in order to have at least some statistical robustness. 
I decided to use two models to calculate ARs, the market model and Fama/French with momentum. I will use the MSCI World Index for the market model, and use an estimation window of 120 days. Then, according to results, I will try to breakdown the ARs using a factor model, and I want to do that to test significance of Intangibles, P/BV and strategy (dummy, if the company acquires a digital(industrial) target for the first time or not).
A lot of it will also depend on the databases you have access to, so you might want to nail that down first. 
Fortunately, my university grants me access to several databases, including Bloomberg, WRDS, Thomson Reuters, CRPS Compustat and so on.. data accessibility should not be a problem. 
Anyways,  comments are highly appreciated!