matching with control firms with some conditions.

[nzrdufiu]

I am having a new problem. 

My imaginary data:

Sample firms:

yr       tick sic  asset

2011 AB 20 2000
2011 BC 21 3000
2011 CD 22 4000
2011 DE 23 5000
2011 EF 24 6000

Control firms:

yr       tick sic  asset

2011 DE 20 900
2011 XS 24 3300
2011 SA 23 2600
2011 FR 22 2000

2011 TH 22 1500

2011 YU 23 2500

My question:

I want to match firms on the basis of sic and asset. My condition is sample firms' asset may be 50% of control firms or vice verse. WHat should be my coding? 

Thank you.

[RW9]

Hi,

You say nothing about what you want in the output dataset, what the linking variables should be, what you mean by "matching" etc.  You should be able to construct a simple SQL statement to do this, condition, you could have where (sample.asset /100)*50 <= control.asset <= (sample.asset/100)* 150.

Joins are (note these examples would need to be in a proc sql; quite; block):