Project description
please see my paper and according to the feedback to modify
Here is some feedback:
1. You may want to look at the models in Mankiw to provide some economic motivation of what are the determinants of growth. For example the IS curve in Mankiw section 11.3, has the interest rate impact economic activity indirectly through its impact on investment. You could use the indirect channel to justify why you are testing the impact of interest rates directly on economic activity (ie change in Y). Alternatively you could use the Output equation in section 15-1, where economic activity is a function of the real interest rate.
2. You should check for robustness the various interest rates, by separately estimating an economic growth equation which includes (1) DY = f(r, ie, TP); (2) DY = f( ie, TP); (3) DY = f(r). Where ie is expected interest rates (signalling channel) and TP is the term premium (portfolio balance channel). The interest rates may be correlated, which a correlation matrix of dependent and independent variables would indicate to you.
3. You could also test the real expected interest rate and/or TP, by subtracting inflation from ie and TP.
4. You should double check you are using OLS estimation, since it looks like you have specified ARCH residuals which is unnecessary.
5. Include a graph of DY
6. You may also want to include a dummy variable (DUM) that interacts with ie and TP, during the periods of QE. For example the dummy is equal to zero before QE and 1 after QE starts. Your most basic regression would be:
DY = b0 +b1*r + b2*DUM*r
This would indicate whether QE means there is a different impact on growth (ie if the estimated coefficient b2 is statistically different from zero, then QE is interacting with the real interest rate).
Hope this helps.
Best wishes