Regression with binary independent variables


I have a multiple regression with 200 binary independent variables. The dependent variable is not binary. I know that there should not be more than 10-20 independent variables in a multiple regression. However, only 6 of them take the value 1 and the rest take the value 0, i.e. that only 6 of the independent variables have a value.

Is it possible to run a regression with 200 indepent variables or should I run a simple linear regression for each independent variable?

I am looking forward to your response.

Kind regards


TS Contributor
What does that mean? You have 194 predictors with only zero values, and 6 predictors
with exclusively value "1"?

What is the topic of your research, what is the precise question, how were data collected,
what are the dependent and the independen variables and how were they measured, and
how large is the sample size?

With kind regards

Hi Karabiner,

Thanks a lot for your answer and effort.

Data set: The data set consists of 1000 conference calls with the participatation of 100 banks.
I dummy coded which banks participate in the conference call (1=participate, 0= does not particapte).

Research question: what is the marginal effect of a single bank on the cumulative return, i.e. does the abnormal return of the firm X increase if JP Morgan participated in the earnings call?

Cumulative abnormal return = α + β1ParticipationBrokerhouse1 +...+βkParticipationBrokerhousek+ Controls + ε

Running this regression, I have 100 binary indepent variables. Is it possible?