Unbiased Forecast

Hello Everyone.

I have a question from Liner Regression section. For unbiased forecast which of the following conditions are required

  • expected value of the forecast error to be zero
  • the intercept (bo) equal to zero and slope (b1) equal to one
  • both the above conditions

can anyone shed some light on this matter?

Thanks in Advance.

The first condition is correct. The second condition is wrong because it will result in a unit root problem

If forecast error’s expected value is not zero, you’ might as well not use the regression since it’ll produce a non-zero error.

the second condition is not correct, as thanhnguyen suggests, it’s a unit root problem.

In addition, the values of the intercept and slope can be of any number as long as it is built from Ordinary Least Square (OLS) regression and fulfills all 6 assumptions of a linear regression (you better memorize or understand these as to how or why they impact regression’s usefulness)

Thanks for the assistance