Question

if both the dollar /sterling (£)exchange rate and the euro dollar exchange rate(y) were I (1) but there was in fact no relationship between the two variables, what would u expect the result would be of performimg the following regression using OLS

E(t)=a+bY(t)+e(t).

()= lower variable not multiplication.

Answer 1

When in fact there is no relationship between two variable and running a regression may causes show a statistically significant coefficient that is spurious correlation between them. In the time series regression when variable is non stationary that is mean and variance changing overtime. Running regression will show significant relationship due to the nature of data. Therefore we need to take first difference and then run the model. In real they are supposed to to statistical insignificant coefficient.