how to produce F-statistic when using all dummies in the eq.

[jwcy45]

Hello, i would like to ask is there any way to produce the f-statistics if i used all dummies in the estimated equation (GJR_GARCH_1,1).
Btw, i am testing the seasonality of a stock index.
Here is the equation:
mean equation: return jan feb mar apr may jun jul aug sep oct nov dec
, where (jan feb mar apr may jun jul aug sep oct nov dec) are the dummies.

actually i want to test whether the coeff. of each dummy are different from each other.

As the estimated output didnt show the f-statistics, therefore, i want to ask is there any way to compute it without including the constant term in the mean equation.

Thanks

[startz]

The F-statistic tests whether the coefficients equal zero, not whether they are all equal.

You can probably use the Wald test to do what you want.

[jwcy45]

The F-statistic tests whether the coefficients equal zero, not whether they are all equal.

You can probably use the Wald test to do what you want.

so you mean i should use test c(1)=c(2)=c(3)=c(4)=c(5)=c(6)=c(7)=c(8)=c(9)=c(10)=c(11)=c(12)=0 in the Wald test???

[startz]

Yes, but leave off the last zero.

[jwcy45]

Yes, but leave off the last zero.

really?
because i tried c(1)=c(2)=c(3)=c(4)=c(5)=c(6)=c(7)=c(8)=c(9)=c(10)=c(11)=c(12) before, and i think it means testing the restriction on c(1) - c(12) =0, c(2) - c(12) =0 ...etc.

sorry mate i am confused.

[EViews Gareth]

Typing:

Code: Select all

c(1)=c(2)=c(3)=c(4)=c(5)=c(6)=c(7)=c(8)=c(9)=c(10)=c(11)=c(12)

tests the restriction c(1)=c(2)=c(3)=c(4)=c(5)=c(6)=c(7)=c(8)=c(9)=c(10)=c(11)=c(12).

If you wanted to test the restriction: c(1) - c(12) =0, c(2) - c(12) =0, you would type:

Code: Select all

c(1) - c(12) =0, c(2) - c(12) =0

[startz]

Yes, but leave off the last zero.

really?
because i tried c(1)=c(2)=c(3)=c(4)=c(5)=c(6)=c(7)=c(8)=c(9)=c(10)=c(11)=c(12) before, and i think it means testing the restriction on c(1) - c(12) =0, c(2) - c(12) =0 ...etc.

sorry mate i am confused.

If c(1)-c(12)=0, then c(1)=c(12).
if c(1)-c(12)=0,c(2)-c(12)=0, then c(1)=c(2)=c(12).

But if c(1)=c(2)=c(12)=0, then c(1)=0, c(2)=0, and c(12)=0.