In the following linear AR ( 1 ) model, rt is the log price of a security at time t . r _ t = 0 . 0 0 4 r _ ( t - 1 ) a _ t , here at is a random process obeying normal distribution N ( , \ sigma 2 ) with mean and variance \ sigma 2 , such as the following a _ t iid N ( 0 , 0 . 3 6 ) , Assume further that r 1 0 0 = 3 . 8 6 5 . Compute the 9 5 % interval forecast for r 1 0 1 at the forecast origin t = 1 0 0