# Collinearity among explanatory variables

My correlation matrix is not showing very high correlations (less than 0.5), but still, when I am regressing panel fixed effects models, all my variables are omitted, and it shows it is due to collinearity among the variables. What is the way out of it?

• This site must already have a number of closely related questions. Please search multicollinearity, singularity. – ttnphns Jan 17 '14 at 13:27