Question 3 (40 points). An analyst for JP Morgan decides that a simple 5 year averaging method is good enough and makes predictions on exchange rates. The table below provides actual values for exchange rates (Y4) along with associated predicted values (Y) for the past 12 months. T Yt Yt Jan 110.9 111.9 Feb 107.9 112.1 Mar 106.1 111.5 Apr 107.7 110.1 May 109.7 109.1 Jun 110.1 108.4 Jul 111.5 108.3 Aug 111.0 109.0 Sep 112.1 110.0 Oct 112.7 110.9 Nov 113.3 111.5 Dec 112.2 112.1 (a) Compute the forecast residuals based on the information given in the table. (b) Compute MAD, MSE, RMSE, MAPE and MPE of the forecast. Give a brief interpre- tation for each value you compute. (c) Based on the report and correlogram of the forecast residuals given below, state your opinion about the adequacy of the forecast method chosen by the analyst. x Autocorrelation Function: RES Lag I ACF Autocorrelation for RES LBQ - 1 0.940560 6.78 48.71 Autocorrelation Function for RES 20.865181 3.75 90.75 3 0.773379 2.70 125.02 4 0.702605 2.17 153.90 5 0.650081 1.85 179.15 6 0.597553 1.60 200.95 7 0.561007 1.43 220.58 8 0.511664 1.26 237.29 9 0.479047 1.14 252.28 10 0.433590 1.01 264.85 Autocorrelation for RES Question 3 (40 points). An analyst for JP Morgan decides that a simple 5 year averaging method is good enough and makes predictions on exchange rates. The table below provides actual values for exchange rates (Y4) along with associated predicted values (Y) for the past 12 months. T Yt Yt Jan 110.9 111.9 Feb 107.9 112.1 Mar 106.1 111.5 Apr 107.7 110.1 May 109.7 109.1 Jun 110.1 108.4 Jul 111.5 108.3 Aug 111.0 109.0 Sep 112.1 110.0 Oct 112.7 110.9 Nov 113.3 111.5 Dec 112.2 112.1 (a) Compute the forecast residuals based on the information given in the table. (b) Compute MAD, MSE, RMSE, MAPE and MPE of the forecast. Give a brief interpre- tation for each value you compute. (c) Based on the report and correlogram of the forecast residuals given below, state your opinion about the adequacy of the forecast method chosen by the analyst. x Autocorrelation Function: RES Lag I ACF Autocorrelation for RES LBQ - 1 0.940560 6.78 48.71 Autocorrelation Function for RES 20.865181 3.75 90.75 3 0.773379 2.70 125.02 4 0.702605 2.17 153.90 5 0.650081 1.85 179.15 6 0.597553 1.60 200.95 7 0.561007 1.43 220.58 8 0.511664 1.26 237.29 9 0.479047 1.14 252.28 10 0.433590 1.01 264.85 Autocorrelation for RES