Merging datasets

[Remco van Velden]

Hello,

I have two datasets, 1 with M&A deals from SDC Platinum from 2000-2023 en another with information about different executives of the acquirors of these deals, also from 2000-2023. In both datasets, there are multiple observations for the two identifiers that I use to merge the datasets (being CUSIP-code and Year).

When merging, I get the following result (see attachment).

None of the data is matched. How do I fix this?

[Rich Goldstein]

first, m:m merging should (virtually) never be done

second, please read the FAQ and supply -dataex- data examples, within CODE blocks, so that people can have something to base an answer on

third, and this is a guess, you might want the -joinby- command rather than the -merge- command; but no one can tell without data examples (realistic data is called for so if your actual data are confidential just make something up in the same structure)

[Devra Golbe]

I echo Rich's comments. As he notes, you may want to supply artificial data. If you do, pay special attention to the definition and format of your merge variables. CUSIPs come in different formats, with either 6, 8, or 9 digits. Your two data sources may not use the same format. Indeed, be careful about the year variable, too. It may be fiscal year in one dataset and calendar year in the other.