“If the exchange rate series is a random walk, then the first-differenced series will yield b0 = 0 and b1 = 0 and the error terms will not be serially correlated. (CFA text)
I have been knowing that a time series with a Unit root will have b0 = 0 and b1 = 1. Then why it has b1 = 0?
Hey Juhi, not 100% sure, I think if a series is a random walk & has a unit root then both the betas i.e., B0 & B1 will be 0.
As said I’m not 100% sure so if you want we can discuss further. Thank you!
Hi Juhi.
The statement talks about first differenced time series and not a time series with a random walk. If a time series model is a random walk i.e it has a unit root present in it, then we can solve the same by first differencing which would make b0=b1=0. Hence, the problem of unit root shall be solved. The same has been mentioned in the notes as well. You may go through it for better understanding. Thank you!