Question

In terms of econometric perspective, what are the assumptions or elements that you need to consider in developing the model for estimation

Answer 1

The following are some of the most important assumptions that are to be kept in mind while developing a model for estimation -

These assumptions are for a Classical Linear Regression Model (CLRM) using Ordinary Least Squares (OLS):

1. Linearity: The dependent variable should be a linear function of the indepedent variables.
2. The mean of the error terms should be zero
3. The independent variables should not b related to each other (no multicollinearity)
4. The dependent variable should not depend on its own previous values (no autocorrelation)
5. The error variances should be constant over the entire data (no heteroscedasticity)
6. The error terms should be correlated with the independent variables - COV(x,e) = 0
7. The number of observations in the data set should be more than the number of independent variables.
8. The sample is random, and normally distributed

Violation of any of these assumptions will give an ill-fitted model which will not be a good fit and will give biased results on forecasting.