EViews.com

Hello everybody,

For the past weeks I've been trying to figure out a problem, unsuccessfully thusfar. I have collected data on M&A transactions. I made four datasets:
1. All transactions per month
2. Transactions in the USA per month
3. Transactions in the UK per month
4. Transactions in Continental Europe per month
Each dataset consists of the same variables and the same number of observations.
In all datasets I've estimated equation: y c interest quantum high low under over. Obviously, each dataset has different estimation results.

My question is: how can I test whether the differences between the coefficients of the variables of each dataset are significant? Currently I can see that the influence of interest in the Continental European is much stronger (e.g. larger coefficient) than its USA and UK counterparts, but I want to know if this difference is significant.

Eventually I want to be able to make statements as: the influence of variable A in the UK is significantly different from the influence of variable A in the USA, but not from Continental Europe and so on for all variables.

Your help is greatly appreciated as this is inevitable for my research, but I am really struggling with it !

As has already been mentioned, the best way to achieve this is to create a set of three dummy variables. One for USA (equal to 1 if a transaction is in the USA, and zero otherwise), one for UK (likewise), and one for Continental Europe (likewise). Then estimate the equation

Y D_USA D_UK D_CE D_USA*INTEREST D_UK*INTEREST D_CE*INTEREST D_USA*QUANTUM D_UK* QUANTUM D_CE* QUANTUM .........

Then you can perform a wald test, testing whether the coefficients on each variable are the same between the regions.

As has already been mentioned, the best way to achieve this is to create a set of three dummy variables. One for USA (equal to 1 if a transaction is in the USA, and zero otherwise), one for UK (likewise), and one for Continental Europe (likewise). Then estimate the equation

Y D_USA D_UK D_CE D_USA*INTEREST D_UK*INTEREST D_CE*INTEREST D_USA*QUANTUM D_UK* QUANTUM D_CE* QUANTUM .........

Then you can perform a wald test, testing whether the coefficients on each variable are the same between the regions.

Thanks Gareth, to my knowledge the Wald test hasn't been mentioned in the other thread but I think now we're getting somewhere. I ran the equation you suggested and then I performed a Wald test on the coefficients. Is it correct to enter for instance C(6)=C(8) to see if variable interest differs for the UK and Continental Europe? In other words, I manually enter the coefficients I want to compare?

The Wald test result then shows:
Test Statistic / Value / df / Probability

F-statistic / 1.199492 / (1, 476) / 0.2740
Chi-square / 1.199492 / 1 / 0.2734

Is it correct to conclude that if the probabilities are below 0.05 that the coefficients are equal (at 5% significance level), i.e. in my case the coefficients are significantly different? But at which significance level or is this irrelevant?

Could anyone please reply to the above as this may be the final hurdle in solving my problem - thanks.