Which of the following problems is not caused by multicollinearity?

It is difficult or even impossible to interpret the estimates of the regression coefficients.

Inordinately small t values for the regression coefficients may result

The standard deviations of the regression coefficients are underestimated

The algebraic sign of estimated regression coefficients may be the opposite of what would be expected for a predictor variable.

The variance of a given regression coefficient is larger than it would otherwise have been if the variable had been completely uncorrelated with all the other variables in the model.