Explanation

An autoregressive model of order p (AR(p)) uses p past values of the variable to predict the current value. Therefore:

• AR(1): Uses 1 past value
• AR(2): Uses 2 past values
• AR(3): Uses 3 past values

For a third-order autoregressive model (AR(3)), the general form is: y_t = b_0 + b_1y_{t-1} + b_2y_{t-2} + b_3y_{t-3} + ε_t

This uses three past values (y_{t-1}, y_{t-2}, y_{t-3}) to predict the current value y_t, making option A correct.