The ABS publishes a series on Australia?s international merchandise imports and exports. The table displays a portion of the data for imports of chemicals and related products from January of year 1 to December of year 5.

(a) Use time-series decomposition methods to develop the seasonal indices for these data.

(b) Use the seasonal indices to deseasonalise the data.

(c) Fit linear, quadratic and exponential trend models to the data.

(d) Which of the models would you recommend as giving the best forecast? Explain your answer.

Transcribed Image Text:

Chemical and Chemical and Chemical and related product imports ($ million) related product imports ($ million) related product imports (S million) Time Time Time period period period (Year 5) January February March (Year 1) January February March (Year 3) January February 1301 986 1218 1048 1223 1198 1245 1226 March 1408 April May June April May June July August September 996 April May 1401 1386 1273 1162 1290 1087 June 997 1343 July August September 1102 1259 July August September 1393 1114 1201 1383 1201 1281 1497 October 1174 October 1237 October 1422 1180 November 1181 November November 1388 December 1035 December 1146 December 1433 (Year 2) January February March (Year 4) January February March 1152 1259 1120 1294 1315 1344 April May 1385 April May June July August September 1347 1327 1261 June 1156 1148 July August September 1177 1205 1285 1214 1108 1173 October 1319 October 1474 November 1245 November 1039 1061 1112 December December