Consider the time series data for the real consumption of households and government from year 1950 to year 2019. Five autoregressive models, AR(1) to AR(5), are estimated as follows, where the t statistics for the corresponding estimated coefficients are in the parentheses, and the size of the samples, n, used for each estimation is also shown in the table. According to the significant lag approach, which model is the optimal under the 5% significance level? AR(1) AR(2) AR(3) AR(4) AR(5) intercept 5854.66 5041.07 5365.55 5689.50 7681.51 t stat (2.44) (2.05) (2.13) (2.13) (2.82) rcont-1 0.82 0.92 0.92 0.92 0.92 t stat (315.53) (9.49 (9.41) (9.33) (9.57) rcont-2 -0.18 -0.28 -0.27 -0.26 t stat (-1.38) (-1.38) (-1.31) (-1.32) rcont-3 0.08 0.02 0.07 t stat (0.64) (0.09 (0.32) rcont-4 0.06 -0.32 t stat (0.38) (-1.47) rcont-5 0.34 t stat (2.32) 65 65 65 65 65 AIC 18.30 18.30 18.33 18.36 18.29 BIC 18.37 18.40 18.46 18.52 18.50 O A. AR(1). O B. AR(2). C. AR(5). D. none