Regression with dpendent variable as a ratio

[Sara JM]

Hey
I want to do a regression where my dependent variable is a ratio of number of acquisition in a year divided by total acquisition during my sample period. Which type of regression is suitable in this case?

[Andrew Musau]

If the outcome is a proportion, i.e., bounded within the range of 0 to 1, then you may want to consider using fracreg.

help fracreg

[Nick Cox]

That's the reciprocal of the mean. Either you have good reason to use it, or perhaps you should not do that.

Ratios can be very unstable and associated with high skewness (often awkward) or serious nonlinearity (usually problematic). Sometimes a logarithmic or even reciprocal link function helps.

A weak answer but the best I can give without context.