‫ﻋﺜﻤﺎﻥ ﻨﻘﺎﺭ – ﻤﻨﺫﺭ ﺍﻟﻌﻭﺍﺩ‬ ‫ﻤﺠﻠﺔ ﺠﺎﻤﻌﺔ ﺩﻤﺸﻕ ﻟﻠﻌﻠﻭﻡ ﺍﻻﻗﺘﺼﺎﺩﻴﺔ ﻭﺍﻟﻘﺎﻨﻭﻨﻴﺔ – ﺍﻟﻤﺠﻠﺩ ‪ - 28‬ﺍﻟﻌﺩﺩ ﺍﻟﺜﺎﻨﻲ ‪2012-‬‬

‫ﻓﻲ ﺍﻟﺘﻨﺒﺅ ﻭﺩﺭﺍﺴﺔ ﺍﻟﻌﻼﻗﺔ ﺍﻟﺴﺒﺒﻴﺔ‬ ‫‪VAR‬‬ ‫ﺍﺴﺘﺨﺩﺍﻡ ﻨﻤﺎﺫﺝ‬

‫ﺒﻴﻥ ﺇﺠﻤﺎﻟﻲ ﺍﻟﻨﺎﺘﺞ ﺍﻟﻤﺤﻠﻲ ﻭﺇﺠﻤﺎﻟﻲ ﺍﻟﺘﻜﻭﻴﻥ ﺍﻟﺭﺃﺴﻤﺎﻟﻲ‬
‫ﻓﻲ ﺴﻭﺭﻴﺔ‬

‫ﺍﻟﺩﻜﺘﻭﺭ ﻤﻨﺫﺭ ﺍﻟﻌﻭﺍﺩ‬ ‫ﺍﻟﺩﻜﺘﻭﺭ ﻋﺜﻤﺎﻥ ﻨﻘﺎﺭ‬

‫ﻜﻠﻴﺔ ﺍﻻﻗﺘﺼﺎﺩ‬
‫ﺠﺎﻤﻌﺔ ﺩﻤﺸﻕ‬

‫ﺍﻟﻤﻠﺨﺹ‬
‫ﻤﻥ ﺍﻟﻤﺅﺸﺭﺍﺕ ﺍﻻﻗﺘﺼﺎﺩﻴﺔ ﺍﻟﻤﻬﻤﺔ ﻷﻱ ﺩﻭﻟﺔ‪ ،‬ﺇﺫﹾ ﻴﻌﻜﺱ ﺇﺠﻤﺎﻟﻲ ﻨﺸﺎﻁ‬ ‫‪GDP‬‬ ‫ﻴﻌﺩ‪ ‬ﺍﻟﻨﺎﺘﺞ ﺍﻟﻤﺤﻠﻲ ﺍﻹﺠﻤﺎﻟﻲ‬
‫ﺍﻟﺩﻭﻟﺔ ﻭﺃﺩﺍﺀﻫﺎ ﺍﻻﻗﺘﺼﺎﺩﻱ ﺨﻼل ﺴﻨﺔ ﻤﺎ‪ ،‬ﻟﺫﻟﻙ ﻻﺒﺩ‪ ‬ﻤﻥ ﺍﻻﻫﺘﻤﺎﻡ ﺒﺎﻷﺴﺎﻟﻴﺏ ﺍﻟﺼﺤﻴﺤﺔ ﻟﻠﺘﻨﺒـﺅ ﺒﺤﺠﻤـﻪ‬
‫ﻟﻠﺴﻨﻭﺍﺕ ﺍﻟﻘﺎﺩﻤﺔ ﺒﻌﺩ ﺃﺨﺫ ﺍﻟﻤﺘﻐﻴ‪‬ﺭﺍﺕ ﺍﻟﻤﺅﺜﺭﺓ ﻓﻴﻪ ﺍﻟﺤﺴﺒﺎﻥ‪ ،‬ﻭﺨﺼﻭﺼﻴﺔ ﻫﺫﺍ ﺍﻟﺘﺄﺜﻴﺭ ﻤـﻥ ﺤﻴـﺙ ﺇﻨﱠـﻪ‬
‫ﻤﺘﺒﺎﻁﺊ )ﺁﺠل( ﻭﻟﻴﺱ ﺁﻨﻴﺎﹰ‪.‬‬

‫(‬ ‫‪Vectorial AutoRegressive) VAR‬‬ ‫ﻫﺩﻓﺕ ﻫﺫﻩ ﺍﻟﺩﺭﺍﺴﺔ ﺇﻟﻰ ﺍﺴﺘﻨﺘﺎﺝ ﻨﻤﻭﺫﺝ ﻗﻴﺎﺴﻲ ﻤﺒﻨﻲ ﻋﻠﻰ ﻨﻤـﺎﺫﺝ‬
‫ﻓﻲ ﺴﻭﺭﻴﺔ‪ ،‬ﻭﻜﺫﻟﻙ ﺇﺠﻤﺎﻟﻲ ﺍﻟﺘﻜﻭﻴﻥ ﺍﻟﺭﺃﺴـﻤﺎﻟﻲ ﺍﻟﺜﺎﺒـﺕ ‪،GFCF‬‬ ‫‪GDP‬‬ ‫ﻟﻠﺘﻨﺒﺅ ﺒﺈﺠﻤﺎﻟﻲ ﺍﻟﻨﺎﺘﺞ ﺍﻟﻤﺤﻠﻲ‬
‫ﻭﺩﺭﺍﺴﺔ ﻋﻼﻗﺔ ﺍﻟﺘﺄﺜﻴﺭ ﻓﻴﻤﺎ ﺒﻴﻨﻬﻤﺎ‪.‬‬

‫ﻓـﻲ‬ ‫‪GDP‬‬ ‫ﺨﻠﺼﺕ ﺍﻟﺩﺭﺍﺴﺔ ﺇﻟﻰ ﺍﺴﺘﻨﺘﺎﺝ ﻨﻤﻭﺫﺝ ﻴﻤﻜﻥ ﺍﺴﺘﺨﺩﺍﻤﻪ ﻓﻲ ﺍﻟﺘﻨﺒﺅ ﺒﺈﺠﻤﺎﻟﻲ ﺍﻟﻨـﺎﺘﺞ ﺍﻟﻤﺤﻠـﻲ‬
‫ﺴﻭﺭﻴﺔ ﻭﻜﺫﻟﻙ ﺇﺠﻤﺎﻟﻲ ﺍﻟﺘﻜﻭﻴﻥ ﺍﻟﺭﺃﺴﻤﺎﻟﻲ ﺍﻟﺜﺎﺒﺕ ‪ ،GFCF‬ﺇﺫﹾ ﺘﻡ ﺍﻟﺘﻨﺒﺅ ﺒﺤﺠﻤﻬﻤﺎ ﺤﺘﻰ ﻋﺎﻡ ‪ .2015‬ﻭﻫﺫﺍ‬
‫ﻤﺎ ﻴﺸﻜل ﻗﺎﻋﺩﺓ ﻋﻠﻤﻴﺔ ﻟﻭﻀﻊ ﺨﻁﻁ ﺍﻟﺘﻨﻤﻴﺔ ﺍﻻﻗﺘﺼﺎﺩﻴﺔ ﻭﺍﻻﺠﺘﻤﺎﻋﻴﺔ‪.‬‬

‫‪337‬‬
 ‫ﺍﺴﺘﺨﺩﺍﻡ ﻨﻤﺎﺫﺝ ‪ VAR‬ﻓﻲ ﺍﻟﺘﻨﺒﺅ ﻭﺩﺭﺍﺴﺔ ﺍﻟﻌﻼﻗﺔ ﺍﻟﺴﺒﺒﻴﺔ ﺒﻴﻥ ﺇﺠﻤﺎﻟﻲ ﺍﻟﻨﺎﺘﺞ ﺍﻟﻤﺤﻠﻲ ﻭﺇﺠﻤﺎﻟﻲ ﺍﻟﺘﻜﻭﻴﻥ ﺍﻟﺭﺃﺴﻤﺎﻟﻲ ﻓﻲ ﺴﻭﺭﻴﺔ‬

‫ﺍﻟﻤﻘﺩﻤﺔ‪:‬‬
‫)‪ (Gross Domestic Product‬ﻤﻥ ﺍﻟﻤﺅﺸﺭﺍﺕ ﺍﻻﻗﺘـﺼﺎﺩﻴﺔ ﺍﻟﻤﻬﻤـﺔ ﻷﻱ‬ ‫‪GDP‬‬ ‫ﻴﻌﺩ‪ ‬ﺍﻟﻨﺎﺘﺞ ﺍﻟﻤﺤﻠﻲ ﺍﻹﺠﻤﺎﻟﻲ‬
‫ﺩﻭﻟﺔ‪ ،‬ﺇﺫﹾ ﻴﻌﻜﺱ ﺇﺠﻤﺎﻟﻲ ﻨﺸﺎﻁ ﺍﻟﺩﻭﻟﺔ ﻭﺃﺩﺍﺀﻫﺎ ﺍﻻﻗﺘﺼﺎﺩﻱ ﺨﻼل ﺴﻨﺔ ﻤﺎ ﺍﻟﺫﻱ ﺒـﺩﻭﺭﻩ ﻴـﻨﻌﻜﺱ ﻋﻠـﻰ‬
‫ﺍﻟﻭﻀﻊ ﺍﻻﻗﺘﺼﺎﺩﻱ ﺍﻟﻌﺎﻡ ﺒﺎﻟﺩﻭﻟﺔ‪ :‬ﻤﺴﺘﻭﻯ ﺍﻟﺘﺸﻐﻴل‪ ،‬ﻭﺍﻟﻤﺴﺘﻭﻯ ﺍﻟﻤﻌﻴـﺸﻲ‪ ،‬ﻭﺍﻟـﺼﺎﺩﺭﺍﺕ ﻭﺍﻟـﻭﺍﺭﺩﺍﺕ‪،‬‬
‫ﻭﺍﻟﺘﺭﺍﻜﻡ ﺍﻟﺭﺃﺴﻤﺎﻟﻲ‪....‬‬

‫ﺒﻌﻼﻗﺘﻪ ﺍﻟﻤﺘﺒﺎﺩﻟﺔ ﻤﻊ ﺤﺠﻡ ﺍﻻﺴﺘﺜﻤﺎﺭﺍﺕ ﺍﻟﺭﺃﺴﻤﺎﻟﻴﺔ‬ ‫‪GDP‬‬ ‫ﻤﻥ ﺠﻬﺔ ﺃﺨﺭﻯ ﻴﺘﻤﻴﺯ ﺍﻟﻨﺎﺘﺞ ﺍﻟﻤﺤﻠﻲ ﺍﻹﺠﻤﺎﻟﻲ‬
‫(‪ ،‬ﻭﻴﺘـﺄﺜﺭ ﺍﻟﻨـﺎﺘﺞ ﺍﻟﻤﺤﻠـﻲ‬ ‫‪Gross Fixed Capital Formation - GFCF‬‬ ‫)ﻤﺠﻤل ﺘﻜﻭﻴﻥ ﺭﺃﺱ ﺍﻟﻤﺎل ﺍﻟﺜﺎﺒﺕ‬
‫ﺍﻹﺠﻤﺎﻟﻲ ﺒﺤﺠﻡ ﺍﻻﺴﺘﺜﻤﺎﺭﺍﺕ‪ ،‬ﻓﻜﻠﻤﺎ ﺍﺯﺩﺍﺩﺕ ﻫﺫﻩ ﺍﻻﺴﺘﺜﻤﺎﺭﺍﺕ ﻭﺍﺯﺩﺍﺩﺕ ﻓﻌﺎﻟﻴﺘﻬﺎ ﺍﺯﺩﺍﺩ ﺍﻟﻨـﺎﺘﺞ ﺍﻟﻤﺤﻠـﻲ‬
‫ﺍﻹﺠﻤﺎﻟﻲ‪ ،‬ﻭﻓﻲ ﺍﻟﻭﻗﺕ ﻨﻔﺴﻪ ﺇﻥ‪ ‬ﺍﻟﻨﺎﺘﺞ ﺍﻟﻤﺤﻠﻲ ﺍﻹﺠﻤﺎﻟﻲ ﻴﺅﺜﺭ ﻓﻲ ﺤﺠﻡ ﺍﻻﺴﺘﺜﻤﺎﺭﺍﺕ ﻷﻥ‪ ‬ﻫـﺫﺍ ﺍﻷﺨﻴـﺭ‬
‫ﺠﺯﺀ ﻤﻥ ﺍﻟﻨﺎﺘﺞ ﺍﻟﻤﺤﻠﻲ ﺍﻹﺠﻤﺎﻟﻲ‪ ،‬ﺴﻭﺍﺀ ﺤ‪‬ﺴِﺏ‪ ‬ﺒﺎﻟﺤﺴﺎﺒﺎﺕ ﺍﻟﻘﻭﻤﻴﺔ ﺒﻁﺭﻴﻘﺔ ﺍﻟﻤﺩﺨﻼﺕ ﻭﺍﻟﻤﺨﺭﺠﺎﺕ‪ ،‬ﻫﺫﺍ‬
‫ﻤﻥ ﺠﻬﺔ‪ ،‬ﻭﻤﻥ ﺠﻬﺔ ﺃﺨﺭﻯ ﻫﻨﺎﻙ ﺨﺼﻭﺼﻴﺔ ﻟﺘﺄﺜﻴﺭ ﺍﻻﺴﺘﺜﻤﺎﺭﺍﺕ ﻓﻲ ﺍﻟﻨﺎﺘﺞ ﺍﻟﻤﺤﻠﻲ ﺍﻹﺠﻤﺎﻟﻲ ﻭﻫـﻲ ﺃﻥ‬
‫ﻫﺫﺍ ﺍﻟﺘﺄﺜﻴﺭ ﻻ ﻴﻅﻬﺭ ﻓﻲ ﺍﻟﻌﺎﻡ ﻨﻔﺴﻪ‪ ،‬ﻭﺇﻨﻤﺎ ﻗﺩ ﻴﻤﺘﺩ ﻋﺩﺓ ﺴﻨﻭﺍﺕ ﺒﺤﺴﺏ ﺴﺭﻋﺔ ﺇﺩﺨﺎل ﺍﻟﻤﺸﺎﺭﻴﻊ ﺍﻟﺠﺩﻴﺩﺓ‬
‫ﺒﺎﻻﺴﺘﺜﻤﺎﺭ‪ .‬ﻫﺫﺍ ﻜﻠﻪ ﻴﺴﺘﻠﺯﻡ ﻤﻨﺎ ﺍﻟﺒﺤﺙ ﻋﻥ ﺍﻷﺴﺎﻟﻴﺏ ﺍﻟﻤﻨﺎﺴﺒﺔ ﻋﻨﺩ ﺍﻟﺘﻨﺒـﺅ ﺒﺤﺠـﻡ ﺍﻟﻨـﺎﺘﺞ ﺍﻟﻤﺤﻠـﻲ‬
‫ﺍﻹﺠﻤﺎﻟﻲ‪.‬‬

‫ﺃﻭﻻﹰ‪ :‬ﺍﻹﻁﺎﺭ ﺍﻟﻌﺎﻡ ﻟﻠﺒﺤﺙ‪:‬‬

‫‪ -1-1‬ﺃﻫﻤﻴﺔ ﺍﻟﺒﺤﺙ ‪:‬‬
‫ﺘﻜﻤﻥ ﺃﻫﻤﻴﺔ ﺍﻟﺒﺤﺙ ﺒﺎﺴﺘﻨﺘﺎﺝ ﻨﻤﻭﺫﺝ ﻗﻴﺎﺴﻲ ﻴﺴﺘﺨﺩﻡ ﻟﻠﺘﻨﺒﺅ ﺒﺤﺠﻤﻲ ﺍﻟﻨﺎﺘﺞ ﺍﻟﻤﺤﻠﻲ ﺍﻹﺠﻤﺎﻟﻲ ﻭﺍﻟﺘﻜـﻭﻴﻥ‬
‫ﺍﻟﺭﺃﺴﻤﺎﻟﻲ ﺒﺎﻟﺠﻤﻬﻭﺭﻴﺔ ﺍﻟﻌﺭﺒﻴﺔ ﺍﻟﺴﻭﺭﻴﺔ ‪ ،‬ﻭﺫﻟﻙ ﺒﺎﺴﺘﺨﺩﺍﻡ ﻤﻨﻬﺞ ﺍﻟﺘﺤﻠﻴل ﺍﻟﺤﺩﻴﺙ ﻟﻠـﺴﻼﺴل ﺍﻟﺯﻤﻨﻴـﺔ‬
‫‪ (Vectorial‬ﺍﻻﻨﺤـﺩﺍﺭ ﺍﻟـﺫﺍﺘﻲ ﺍﻟـﺸﻌﺎﻋﻲ ﺃﻭ‬ ‫‪AutoRegressive) VAR‬‬ ‫ﺍﻟﻤﺒﻨﻲ ﻋﻠﻰ ﺍﺴﺘﺨﺩﺍﻡ ﻨﻤـﺎﺫﺝ‬
‫ﺍﻻﺘﺠﺎﻫﻲ‪ ،‬ﻭﻤﻥ ﺜﻡ ﺍﻟﺘﻨﺒﺅ ﺒﺤﺠﻡ ﺍﻟﻨﺎﺘﺞ ﺍﻟﻤﺤﻠﻲ ﺍﻹﺠﻤﺎﻟﻲ ﻭﻜﺫﻟﻙ ﺒﺤﺠﻡ ﺍﻻﺴﺘﺜﻤﺎﺭﺍﺕ ﺤﺘﻰ ﻋﺎﻡ ‪. 2015‬‬

‫‪ -2-1‬ﺃﻫﺩﺍﻑ ﺍﻟﺒﺤﺙ ‪:‬‬

‫ﻫﺩﻑ ﺍﻟﺒﺤﺙ ﺇﻟﻰ‪:‬‬

‫( ﻓﻲ ﺘﺤﻠﻴل ﺍﻟﺴﻼﺴل ﺍﻟﺯﻤﻨﻴﺔ‪.‬‬ ‫‪Vectorial AutoRegressive) VAR‬‬ ‫ﺍﻟﺘﻌﺭﻴﻑ ﺒﻨﻤﺎﺫﺝ‬ ‫‪-‬‬

‫ﻜﺸﻑ ﺍﻟﻌﻼﻗﺔ ﺍﻟﺴﺒﺒﻴﺔ ﻏﻴﺭ ﺍﻵﻨﻴﺔ ﺒﻴﻥ ﺍﻟﻨﺎﺘﺞ ﺍﻟﻤﺤﻠﻲ ﺍﻹﺠﻤﺎﻟﻲ ﻭﺤﺠﻡ ﺍﻻﺴﺘﺜﻤﺎﺭﺍﺕ‪.‬‬ ‫‪-‬‬

‫‪338‬‬
‫ﻋﺜﻤﺎﻥ ﻨﻘﺎﺭ – ﻤﻨﺫﺭ ﺍﻟﻌﻭﺍﺩ‬ ‫ﻤﺠﻠﺔ ﺠﺎﻤﻌﺔ ﺩﻤﺸﻕ ﻟﻠﻌﻠﻭﻡ ﺍﻻﻗﺘﺼﺎﺩﻴﺔ ﻭﺍﻟﻘﺎﻨﻭﻨﻴﺔ – ﺍﻟﻤﺠﻠﺩ ‪ - 28‬ﺍﻟﻌﺩﺩ ﺍﻟﺜﺎﻨﻲ ‪2012-‬‬

‫ﻓﻲ ﺘﺤﻠﻴل ﺍﻟﺴﻼﺴل ﺍﻟﺯﻤﻨﻴﺔ ﻓﻲ ﺍﻟﺘﻨﺒﺅ ﺒﺤﺠﻡ ﺍﻟﻨﺎﺘﺞ ﺍﻟﻤﺤﻠﻲ ﺍﻹﺠﻤﺎﻟﻲ‪.‬‬ ‫‪VAR‬‬ ‫ﺍﺨﺘﺒﺎﺭ ﻗﺩﺭﺓ ﻨﻤﺎﺫﺝ‬ ‫‪-‬‬
‫ﺍﺴﺘﻨﺘﺎﺝ ﻨﻤﻭﺫﺝ ﻗﻴﺎﺴﻲ ﻟﻠﺘﻨﺒﺅ ﺒﺤﺠﻡ ﺍﻟﻨﺎﺘﺞ ﺍﻟﻤﺤﻠﻲ ﺍﻹﺠﻤﺎﻟﻲ ﻭﻜﺫﻟﻙ ﺒﺤﺠﻡ ﺍﻻﺴﺘﺜﻤﺎﺭﺍﺕ‪.‬‬ ‫‪-‬‬
‫ﺍﻟﺘﻨﺒﺅ ﺒﺤﺠﻡ ﺍﻟﻨﺎﺘﺞ ﺍﻟﻤﺤﻠﻲ ﺍﻹﺠﻤﺎﻟﻲ‪ ،‬ﻭﻜﺫﻟﻙ ﺒﺤﺠﻡ ﺍﻻﺴﺘﺜﻤﺎﺭﺍﺕ ﻓﻲ ﺴﻭﺭﻴﺔ ﺤﺘﻰ ﻋﺎﻡ ‪. 2015‬‬ ‫‪-‬‬

‫‪ -3-1‬ﻤﻨﻬﺠﻴﺔ ﺍﻟﺒﺤﺙ‪:‬‬
‫ﺍﺴﺘﹸﺨﺩِﻡ‪ ‬ﺍﻟﻤﻨﻬﺞ ﺍﻟﻭﺼﻔﻲ ﺍﻟﺘﺤﻠﻴﻠﻲ ﻓﻲ ﺇﻨﺠﺎﺯ ﻫﺫﺍ ﺍﻟﺒﺤﺙ ﻤﻥ ﺨﻼل ﺍﻻﻁﻼﻉ ﻋﻠﻰ ﻋـﺩﺩ ﻤـﻥ ﺍﻟﻤﺭﺍﺠـﻊ‬
‫‪Vectorial‬‬ ‫)‬ ‫‪VAR‬‬ ‫)ﺒﺎﻟﻠﻐﺔ ﺍﻟﻌﺭﺒﻴـﺔ‪ ،‬ﻭﺍﻟﻔﺭﻨـﺴﻴﺔ‪ ،‬ﻭﺍﻹﻨﻜﻠﻴﺯﻴـﺔ( ﺍﻟﺘـﻲ ﺘﻨﺎﻭﻟـﺕ ﺍﺴـﺘﺨﺩﺍﻡ ﻨﻤـﺎﺫﺝ‬
‫( ﻓﻲ ﺘﺤﻠﻴل ﺍﻟﺴﻼﺴل ﺍﻟﺯﻤﻨﻴﺔ‪ ،‬ﻭﻤﻥ ﺜﻡ ﺘﻡ ﺍﻟﺤﺼﻭل ﻋﻠﻰ ﺒﻴﺎﻨﺎﺕ ﺍﻟﺴﻠـﺴﻠﺔ ﺍﻟﺯﻤﻨﻴـﺔ‬ ‫‪AutoRegressive‬‬

‫ﻟﺤﺠﻡ ﺍﻟﻨﺎﺘﺞ ﺍﻹﺠﻤﺎﻟﻲ‪ ،‬ﻭﺤﺠﻡ ﺘﻜﻭﻴﻥ ﺭﺃﺱ ﺍﻟﻤﺎل ﺍﻟﺜﺎﺒﺕ ﻤﻥ ﺍﻟﻤﺠﻤﻭﻋﺔ ﺍﻹﺤﺼﺎﺌﻴﺔ ﺍﻟﺴﻭﺭﻴﺔ‪ ،‬ﻭﺃُﺠـﺭﻱ‬
‫ﺍﻟﺘﻁﺒﻴﻕ ﻋﻠﻴﻬﺎ ﺒﺎﺴﺘﺨﺩﺍﻡ ﺍﻟﺤﺯﻤﺔ ﺍﻟﺒﺭﻤﺠﻴﺔ ‪ EViews 6‬ﻓﻲ ﺘﺤﻠﻴﻠﻬﺎ‪.‬‬

‫ﺜﺎﻨﻴﺎﹰ‪ :‬ﺍﻟﻨﻤﻭﺫﺝ ‪:(Vectorial AutoRegressive) VAR‬‬

‫ﻴﺭﻯ ﺃﻥ ﺍﻟﻁﺭﻴﻘﺔ ﺍﻟﺘﻘﻠﻴﺩﻴﺔ ﻓﻲ ﺒﻨﺎﺀ ﺍﻟﻨﻤﺎﺫﺝ‬ ‫‪Sims‬‬ ‫ﻓﻲ ﻋﺎﻡ ‪ ،19811‬ﻜﺎﻥ‬ ‫‪Sims‬‬ ‫ﻗﺎﻡ ﺒﺎﻗﺘﺭﺍﺡ ﻫﺫﺍ ﺍﻟﻨﻤﻭﺫﺝ‬
‫ﺍﻟﻘﻴﺎﺴﻴﺔ ﺍﻵﻨﻴﺔ ﺘﻌﺘﻤﺩ ﻭﺠﻬﺔ ﺍﻟﻨﻅﺭ ﺍﻟﺘﻔﺴﻴﺭﻴﺔ‪ ،‬ﺇﺫﹾ ﺘﺘﻀﻤﻥ ﻜﺜﻴﺭﺍﹰ ﻤﻥ ﺍﻟﻔﺭﻀﻴﺎﺕ ﻏﻴـﺭ ﺍﻟﻤﺨﺘﺒـﺭﺓ ﻤﺜـل‪:‬‬
‫ﻤﻘﺒـﻭل‬ ‫)‪(Identification‬‬ ‫ﺍﺴﺘﺒﻌﺎﺩ ﺒﻌﺽ ﺍﻟﻤﺘﻐﻴ‪‬ﺭﺍﺕ ﻤﻥ ﺒﻌﺽ ﺍﻟﻤﻌﺎﺩﻻﺕ ﻤﻥ ﺃﺠل ﺍﻟﻭﺼﻭل ﺇﻟﻰ ﺘﺸﺨﻴﺹ‬
‫‪ ،‬ﻭﺸـﻜل ﺘﻭﺯﻴـﻊ ﻓﺘـﺭﺍﺕ‬ ‫)‪(Exogenes‬‬ ‫ﻟﻠﻨﻤﻭﺫﺝ‪ ،‬ﻭﻜﺫﻟﻙ ﺍﻷﻤﺭ ﻓﻴﻤﺎ ﻴﺘﻌﻠﻕ ﺒﺎﺨﺘﻴﺎﺭ ﺍﻟﻤﺘﻐﻴ‪‬ﺭﺍﺕ ﺍﻟﺨﺎﺭﺠﻴﺔ‬
‫ﺍﻹﺒﻁﺎﺀ ﺍﻟﺯﻤﻨﻲ‪.‬‬

‫ﻓﻲ ﻨﻤﻭﺫﺠﻪ ﻤﻌﺎﻤﻠﺔ ﺍﻟﻤﺘﻐﻴ‪‬ﺭﺍﺕ ﺠﻤﻴﻌﻬﺎ ﺒﺎﻟﻁﺭﻴﻘﺔ ﻨﻔﺴﻬﺎ ﺩﻭﻥ ﺃﻴﺔ ﺸﺭﻭﻁ ﻤﺴﺒﻘﺔ )ﺍﺴﺘﺒﻌﺎﺩﻫﺎ‬ ‫‪Sims‬‬ ‫ﻴﻘﺘﺭﺡ‬
‫ﺃﻭ ﻋﺩ‪‬ﻫﺎ ﺨﺎﺭﺠﻴﺔ(‪ ،‬ﻭﺇﺩﺨﺎﻟﻬﺎ ﺠﻤﻴﻌﺎﹰ ﻓﻲ ﺍﻟﻤﻌﺎﺩﻻﺕ ﺒﻌﺩﺩ ﻤﺩﺩ ﺍﻹﺒﻁﺎﺀ ﺍﻟﺯﻤﻨﻲ ﻨﻔﺴﻬﺎ‪.‬‬

‫ﺍﻟﺫﻱ ﻴﻤﻜﻥ ﻜﺘﺎﺒﺘﻪ ﺒﺎﻟـﺸﻜل‬ ‫‪(VAR) Vectorial AutoRegressive‬‬ ‫ﺍﻟﻨﻤﻭﺫﺝ ﺍﻟﻌﺎﻡ ﺍﻟﺫﻱ ﺍﻗﺘﺭﺤﻪ ﻫﻭ ﻨﻤﻭﺫﺝ‬
‫ﺍﻵﺘﻲ‪:‬‬

‫‪Φ ( B)Yt = ε t‬‬

‫ﺇﺫﹾ‪:‬‬

‫‪: Yt‬ﺴﻴﺎﻕ ﻋﺸﻭﺍﺌﻲ ﺫﻭ ‪ n‬ﺒﻌﺩ‪ ،‬ﻤﺴﺘﻘﺭ ﻤﻥ ﺍﻟﻤﺭﺘﺒﺔ ﺍﻟﺜﺎﻨﻴﺔ‪.2‬‬

‫‪1‬‬
‫‪SIMS C. A. (1981) "Macroeconomics and Reality". Econometrica, n48: pp 1-48.‬‬
‫اﻧﻈﺮ اﻟﻔﻘﺮة ‪1-1-2‬‬ ‫‪2‬‬

‫‪339‬‬
 ‫ﺍﺴﺘﺨﺩﺍﻡ ﻨﻤﺎﺫﺝ ‪ VAR‬ﻓﻲ ﺍﻟﺘﻨﺒﺅ ﻭﺩﺭﺍﺴﺔ ﺍﻟﻌﻼﻗﺔ ﺍﻟﺴﺒﺒﻴﺔ ﺒﻴﻥ ﺇﺠﻤﺎﻟﻲ ﺍﻟﻨﺎﺘﺞ ﺍﻟﻤﺤﻠﻲ ﻭﺇﺠﻤﺎﻟﻲ ﺍﻟﺘﻜﻭﻴﻥ ﺍﻟﺭﺃﺴﻤﺎﻟﻲ ﻓﻲ ﺴﻭﺭﻴﺔ‬

‫ﻴﻜﺘﺏ ﻜﻤﺎ ﻴﺄﺘﻲ‪:‬‬ ‫‪B‬‬ ‫)‪ : Φ (B‬ﻜﺜﻴﺭ ﺤﺩﻭﺩ ﻤﺼﻔﻭﻓﻲ ﻤﻥ ﺍﻟﺩﺭﺠﺔ ‪ p‬ﺒﻤﻌﺎﻤل ﺍﻹﺒﻁﺎﺀ ﺍﻟﺯﻤﻨﻲ‬

‫‪Φ ( B) = φ 0 − Bφ1 − B 2φ 2 − ... − B p φ p‬‬

‫‪n‬‬ ‫‪ : φ0‬ﻤﺼﻔﻭﻓﺔ ﺃﺤﺎﺩﻴﺔ ﻤﻥ ﺍﻟﻤﺭﺘﺒﺔ‬

‫ﺒﻌﺩ‪ ،‬ﻤﺼﻔﻭﻓﺔ ﺘﻐﺎﻴﺭﺍ ﺘﻪ ﻫﻲ‪ . Ω :‬ﻭ ﻴﻔﺴﺭ ﻫﺫﺍ ﺍﻟﺴﻴﺎﻕ ﺒﺄﻨﱠﻪ ﺘﺠﺩﻴﺩ‬ ‫‪n‬‬ ‫‪ : ε t‬ﺴﻴﺎﻕ ﺍﻟﻀﺠﺔ ﺍﻟﺒﻴﻀﺎﺀ ﺫﻭ‬
‫ﻟﻠﺴﻴﺎﻕ ﺍﻟﻌﺸﻭﺍﺌﻲ ‪. Yt‬‬ ‫)‪(innovation‬‬

‫ﻋﻠﻰ ﺸﻜل ﻤﺠﻤﻭﻋﺔ ﻤﻥ ﺍﻟﻤﻌﺎﺩﻻﺕ ﻜﻤﺎ ﻴﺄﺘﻲ‪:3‬‬ ‫‪VAR‬‬ ‫ﻴﻤﻜﻥ ﺃﻴﻀﺎﹰ ﻜﺘﺎﺒﺔ ﺍﻟﻨﻤﻭﺫﺝ‬

‫‪y1t = φ11(1) y1,t −1 + ... + φ11( p ) y1,t − p + ... + φ1(n1) yn ,t −1 + ... + φ1(np ) yn ,t − p + ε 1,t‬‬

‫‪ynt = φ n(11) y1,t −1 + ... + φ n(1p ) y1,t − p + ... + φ nn(1) yn ,t −1 + ... + φ nn( p ) yn ,t − p + ε n,t‬‬

‫ﻴﻅﻬﺭ ﻟﻨﺎ ﺠﻠﻴﺎﹰ ﻓﻲ ﺍﻟﻜﺘﺎﺒﺔ ﺍﻷﺨﻴﺭﺓ ﺃﻥ ﻜل ﻤﻌﺎﺩﻟﺔ ﻫﻲ ﻋﺒﺎﺭﺓ ﻋﻥ ﻤﻌﺎﺩﻟﺔ ﺍﻨﺤﺩﺍﺭ ﻟﻌﻨﺼﺭ ﻤﻥ ﺍﻟﺸﻌﺎﻉ ‪Yt‬‬
‫ﻋﻠﻰ ﻤﺎﻀﻴﻪ ﻭﻤﺎﻀﻲ ﺍﻟﻌﻨﺎﺼﺭ ﺍﻷﺨﺭﻯ ﻤﻥ ﺍﻟﺸﻌﺎﻉ‪ .‬ﻨﺭﻯ ﻓﻲ ﻫﺫﻩ ﺍﻟﻤﻌﺎﺩﻻﺕ ﻨﻭﻋﺎﹰ ﻤﻥ ﺍﻻﻨﺘﻅﺎﻡ‬
‫ﺍﻹﺤﺼﺎﺌﻲ ﻓﻲ ﺇﺩﺨﺎل ﺍﻟﻤﺘﻐﻴ‪‬ﺭﺍﺕ‪ ،‬ﻭﺒﺸﻜل ﺨﺎﺹ ﺃﺨﺫ ﺍﻟﺘﺄﺜﻴﺭﺍﺕ ﺍﻟﺩﻴﻨﺎﻤﻴﻜﻴﺔ ﺍﻟﻤﺘﺒﺎﺩﻟﺔ ﺒﻴﻥ ﻫﺫﻩ ﺍﻟﻤﺘﻐﻴ‪‬ﺭﺍﺕ‬
‫ﺒﺎﻟﺤﺴﺒﺎﻥ‪.‬‬

‫ﺇﻥ ﺘﻘﺩﻴﺭ ﺍﻟﻨﻤﻭﺫﺝ ﺍﻟﺴﺎﺒﻕ ﻴﻤﻜﻥ ﺃﻥ ﻴﺘﻡ ﺒﺎﺴﺘﺨﺩﺍﻡ ﻁﺭﻴﻘﺔ ﺍﻟﻤﺭﺒﻌﺎﺕ ﺍﻟﺼﻐﺭﻯ ﻤﻁﺒﻘﺔ ﻋﻠﻰ ﻜل ﻤﻌﺎﺩﻟﺔ‬
‫ﺤﺩﺓ ﺍﻟﺨﺼﺎﺌﺹ ﺍﻟﺘﻘﺎﺭﺒﻴﺔ ﻟﻠﻤﻘﺩﺭﺍﺕ ﺍﻟﺘﻲ ﻴﻤﻜﻥ ﺃﻥ ﻨﺤﺼل ﻋﻠﻴﻬﺎ ﻫﻲ ﺍﻟﺨﺼﺎﺌﺹ ﺍﻟﻤﻌﺘﺎﺩﺓ‪ ،‬ﺇﺫﺍ ﻜﺎﻥ‬
‫ﺍﻟﺴﻴﺎﻕ ﺍﻟﻌﺸﻭﺍﺌﻲ ‪ Yt‬ﻤﺴﺘﻘﺭﺍﹰ ﻤﻥ ﺍﻟﻤﺭﺘﺒﺔ ﺍﻟﺜﺎﻨﻴﺔ‪.‬‬

‫‪ : 1-2‬ﺒﻨﺎﺀ ﺍﻟﻨﻤﻭﺫﺝ ‪: VAR‬‬

‫ﻴﺘﻁﻠﺏ ﻤﺎ ﻴﺄﺘﻲ ‪:‬‬ ‫‪VAR‬‬ ‫ﺇﻥ ﺒﻨﺎﺀ ﺍﻟﻨﻤﻭﺫﺝ‬

‫ﺃﻥ ﺘﻜﻭﻥ ﺍﻟﺴﻼﺴل ﺍﻟﺯﻤﻨﻴﺔ ﺍﻟﻤﺴﺘﺨﺩﻤﺔ ﻤﺴﺘﻘﺭﺓ‪ ،‬ﺃﻱ ﻻ ﺘﺤﻭﻱ ﺠﺫﺭ ﺍﻟﻭﺤﺩﺓ‪.‬‬ ‫‪-1‬‬
‫ﺘﺤﺩﻴﺩ ﻋﺩﺩ ﻤﺩﺩ ﺍﻹﺒﻁﺎﺀ ﺍﻟﺯﻤﻨﻲ ﺍﻟﺘﻲ ﺴﺘﹸﻌﺘﹶﻤ‪‬ﺩ‪ ‬ﻓﻲ ﺍﻟﻨﻤﻭﺫﺝ‪.‬‬ ‫‪-2‬‬
‫ﺩﺭﺍﺴﺔ ﻋﻼﻗﺔ ﺍﻟﺴﺒﺒﻴﺔ ﺒﻴﻥ ﺍﻟﻤﺘﻐﻴ‪‬ﺭﺍﺕ‪.‬‬ ‫‪-3‬‬

‫‪3‬‬
‫‪SHUMWAY R.H. and STOFFER D.S. (2006) "Time Series Analysis and Its Applications".‬‬
‫‪SPRINGER, New York. pp. 303-304.‬‬

‫‪340‬‬
‫ﻋﺜﻤﺎﻥ ﻨﻘﺎﺭ – ﻤﻨﺫﺭ ﺍﻟﻌﻭﺍﺩ‬ ‫ﻤﺠﻠﺔ ﺠﺎﻤﻌﺔ ﺩﻤﺸﻕ ﻟﻠﻌﻠﻭﻡ ﺍﻻﻗﺘﺼﺎﺩﻴﺔ ﻭﺍﻟﻘﺎﻨﻭﻨﻴﺔ – ﺍﻟﻤﺠﻠﺩ ‪ - 28‬ﺍﻟﻌﺩﺩ ﺍﻟﺜﺎﻨﻲ ‪2012-‬‬

‫ﺴﻨﺒﺩﺃ ﺍﻵﻥ ﺒﻌﺭﺽ ﻨﻅﺭﻱ ﻤﻭﺠﺯ ﻟﻬﺫﻩ ﺍﻟﻤﺴﺎﺌل ﻗﺒل ﺃﻥ ﻨﻘﻭﻡ ﺒﺘﻁﺒﻴﻘﻬﺎ ﻋﻠﻰ ﺍﻟﺤﺎﻟﺔ ﺍﻟﺩﺭﺍﺴﻴﺔ ‪.‬‬
‫‪ : 1-1-2‬ﺍﻻﺴﺘﻘﺭﺍﺭ ‪: 4Stationarity‬‬
‫ﻨﻘﻭل ﻋﻥ ﺍﻟﺴﻴﺎﻕ ﺍﻟﻌﺸﻭﺍﺌﻲ ‪ Yt‬ﺇﻨﻪ ﻤﺴﺘﻘﺭ ﻤﻥ ﺍﻟﻤﺭﺘﺒﺔ ﺍﻟﺜﺎﻨﻴﺔ ﺇﺫﺍ ﻜﺎﻥ ﺍﻟﺘﻭﻗـﻊ ﺍﻟﺭﻴﺎﻀـﻲ ﻟــ ‪Yt‬‬
‫ﻭ ‪ Yt + h‬ﻫﻭ ﻨﻔﺴﻪ ﻤﻥ ﺃﺠل ﻜل ‪ t‬ﻭﻜل ﻋﺩﺩ ﺼﺤﻴﺢ ‪ ،h‬ﻭ ﻜﺫﻟﻙ ﺍﻷﻤﺭ ﺒﺎﻟﻨﺴﺒﺔ ﺇﻟﻰ ﺍﻟﺘﺒﺎﻴﻥ‪:‬‬
‫) ‪E (Yt ) = E (Yt + h‬‬
‫) ‪V (Yt ) = V (Yt + h‬‬
‫ﻭﺇﺫﺍ ﻜﺎﻥ ﺍﻟﺘﺒﺎﻴﻥ ﺍﻟﻤﺸﺘﺭﻙ ﻟـ ‪ Yt‬ﻭ ‪ Yt + h‬ﻤﺴﺘﻘﻼﹰ ﻋﻥ ﺍﻟﺯﻤﻥ‪ .‬ﺃﻱ ﺇﺫﺍ ﻜﺎﻥ‪:‬‬
‫]) ‪d [Cov(Yt , Yt + h‬‬
‫‪=0‬‬
‫‪dt‬‬
‫‪: 1-1-1-2‬ﺍﺨﺘﺒﺎﺭ ﺍﻻﺴﺘﻘﺭﺍﺭ ‪: Stationarity test‬‬
‫ﺇﻥ ﻋﺩﻡ ﺍﻻﺴﺘﻘﺭﺍﺭ ﻓﻲ ﺍﻟﺴﻼﺴل ﺍﻟﺯﻤﻨﻴﺔ ﻴﺭﺠﻊ ‪-‬ﻓﻲ ﻜﺜﻴﺭ ﻤﻥ ﺍﻷﺤﻴﺎﻥ ‪ -‬ﺇﻟﻰ ﻭﺠﻭﺩ ﺠﺫﺭ ﺍﻟﻭﺤﺩﺓ‪ ،‬ﻭﻗـﺩ‬
‫ﺍﺨﺘﺒﺎﺭﺍﹰ ﻴﻜﺸﻑ ﻭﺠﻭﺩ ﺠﺫﺭ ﺍﻟﻭﺤﺩﺓ ﺃﻭ ﻋﺩﻡ ﻭﺠﻭﺩﻫﺎ‪.‬‬ ‫‪Dickey & Fuller‬‬ ‫ﺍﻗﺘﺭﺡ‬

‫)‪:5(Dickey & Fuller‬‬ ‫‪D.F.‬‬ ‫‪ -1‬ﺍﺨﺘﺒﺎﺭ‬

‫ﺒﻔﺭﺽ ﺃﻨﻪ ﻟﺩﻴﻨﺎ ﺍﻟﺴﻴﺎﻕ ﺍﻟﻌﺸﻭﺍﺌﻲ ‪ ، Xt‬ﻨﻘﻭﻡ ﺒﺤﺴﺎﺏ ﺍﻻﻨﺤﺩﺍﺭ ﺍﻟﺘﺎﻟﻲ‪:‬‬

‫‪P‬‬
‫‪Xt = ρXt −1 + ∑ a i ( Xt −i − Xt −i −1 ) + ε t‬‬
‫‪i =1‬‬

‫ﻋﻠﻰ ﺍﺨﺘﺒﺎﺭ ﻗﻴﻤﺔ ‪ ρ‬ﻓﻴﻤﺎ ﺇﺫﺍ ﻜﺎﻨﺕ ‪ ρ = 1‬ﺃﻱ ﺇﻥ‪ ‬ﻓﺭﻀﻴﺎﺕ ﺍﻻﺨﺘﺒﺎﺭ ﻫﻲ‪:‬‬ ‫‪D.F.‬‬ ‫ﻴﻘﻭﻡ ﺍﺨﺘﺒﺎﺭ‬

‫‪H0 : ρ = 1‬‬ ‫ﻗﺒﻭل ﻫﺫﻩ ﺍﻟﻔﺭﻀﻴﺔ ﻴﻌﻨﻲ ﻗﺒﻭل ﻋﺩﻡ ﺍﻻﺴﺘﻘﺭﺍﺭ ﻭﻭﺠﻭﺩ ﺠﺫﺭ ﺍﻟﻭﺤﺩﺓ‬

‫‪H1 : ρ < 1‬‬ ‫ﻗﺒﻭل ﻫﺫﻩ ﺍﻟﻔﺭﻀﻴﺔ ﻴﻌﻨﻲ ﻗﺒﻭل ﺍﻻﺴﺘﻘﺭﺍﺭ ﻭ ﻋﺩﻡ ﻭﺠﻭﺩ ﺠﺫﺭ ﺍﻟﻭﺤﺩﺓ‬

‫‪.‬‬ ‫‪Dickey & Fuller‬‬ ‫ﺍﻟﺠﺩﻭﻟﻴﺔ ﺍﻟﺘﻲ ﻗﺩﻤﻬﺎ‬ ‫‪t‬‬ ‫ﺜﻡ ﻨﻘﻭﻡ ﺒﻤﻘﺎﺭﻨﺘﻬﺎ ﻤﻊ‬ ‫*‪t‬‬ ‫ﻨﺤﺴﺏ ﺇﺤﺼﺎﺌﻴﺔ ﺍﻻﺨﺘﺒﺎﺭ‬

‫‪4‬‬
‫‪Kirchgässner G. and Wolters J. (2007) "Introduction to Modern Time Series Analysis",‬‬
‫‪SPRINGER-Verlag, Berlin Heidelberg. pp. 13-14.‬‬
‫‪5‬‬
‫‪Dickey D. and Fuller W.(1979), " Distribution of the estimators for Autoregressive Time Series‬‬
‫‪With a unit Root ", Journal of the American Statistical Association, n74: pp .427-431.‬‬

‫‪341‬‬
 ‫ﺍﺴﺘﺨﺩﺍﻡ ﻨﻤﺎﺫﺝ ‪ VAR‬ﻓﻲ ﺍﻟﺘﻨﺒﺅ ﻭﺩﺭﺍﺴﺔ ﺍﻟﻌﻼﻗﺔ ﺍﻟﺴﺒﺒﻴﺔ ﺒﻴﻥ ﺇﺠﻤﺎﻟﻲ ﺍﻟﻨﺎﺘﺞ ﺍﻟﻤﺤﻠﻲ ﻭﺇﺠﻤﺎﻟﻲ ﺍﻟﺘﻜﻭﻴﻥ ﺍﻟﺭﺃﺴﻤﺎﻟﻲ ﻓﻲ ﺴﻭﺭﻴﺔ‬

‫ﻴﻤﻜﻥ ﺃﻥ ﻨﻀﻴﻑ ﺇﻟﻰ ﺍﻟﻤﻌﺎﺩﻟﺔ ﺍﻟﺴﺎﺒﻘﺔ ﺤﺩﺍﹰ ﺜﺎﺒﺘﺎﹰ‪ ،‬ﺃﻭ ﺤﺩﺍﹰ ﺜﺎﺒﺘﺎﹰ ﻭﺍﻨﺤﺩﺍﺭﺍﹰ ﺨﻁﻴﺎﹰ ﺒﺎﻟﺯﻤﻥ ﺒﺤـﺴﺏ ﻁﺒﻴﻌـﺔ‬
‫ﺍﻟﺴﻠﺴﻠﺔ ﺍﻟﺯﻤﻨﻴﺔ ﺍﻟﻤﺩﺭﻭﺴﺔ‪ ،‬ﻭ ﻓﻲ ﻫﺫﻩ ﺍﻟﺤﺎﻟﺔ ﺘﻜـﻭﻥ ﺍﻟﻘـﻴﻡ ﺍﻟﺠﺩﻭﻟﻴـﺔ ﻤﺨﺘﻠﻔـﺔ ﺒﺤـﺴﺏ ﺍﻟﻤﻌﺎﺩﻟـﺔ‬
‫ﺍﻟﻤﺴﺘﺨﺩﻤﺔ‪.‬‬
‫‪6‬‬
‫)‪: ( Augmanted Dickey & Fuller‬‬ ‫‪A.D.F.‬‬ ‫‪ -2‬ﺍﻻﺨﺘﺒﺎﺭ ﺍﻟﻤﺤﺴﻥ‬
‫‪P‬‬
‫ﻨﻘﻭﻡ ﺒﺤﺴﺎﺏ ﺍﻻﻨﺤﺩﺍﺭ‪∆Xt = ϕXt −1 + ∑ a i ( Xt −i − Xt −i −1 ) + ε t :‬‬
‫‪i =1‬‬

‫ﺇﺫﹾ‪ϕ = ρ − 1 :‬‬
‫‪H0 : ϕ = 0‬‬ ‫ﻓﻲ ﻫﺫﻩ ﺍﻟﺤﺎﻟﺔ ﻓﺭﻀﻴﺎﺕ ﺍﻻﺨﺘﺒﺎﺭ‪:‬‬
‫‪H1 : ϕ < 0‬‬
‫ﺇﺫﺍ ﺘﺒﻴ‪‬ﻥ ﻟﻨﺎ ﻨﺘﻴﺠﺔ ﺍﻻﺨﺘﺒﺎﺭ ﺃﻥ ﺍﻟﺴﻠﺴﻠﺔ ﺍﻟﺯﻤﻨﻴﺔ ﻏﻴﺭ ﻤﺴﺘﻘﺭﺓ ﻭ ﺘﺤﻭﻱ ﺠﺫﺭ ﺍﻟﻭﺤﺩﺓ ﻨﻘﻭﻡ ﺒﺘﺤﻭﻴﻠﻬﺎ ﺇﻟﻰ‬
‫ﺴﻠﺴﻠﺔ ﻤﺴﺘﻘﺭﺓ ﺒﺘﻁﺒﻴﻕ ﻤﺭﺸﺢ ﺍﻟﻔﺭﻭﻕ ﺍﻷﻭﻟﻰ‪ ، ∆ = (1 − B) 7‬ﺜﻡ ﻨﻘﻭﻡ ﺒﺎﺨﺘﺒﺎﺭ ﺍﻟﺴﻠﺴﻠﺔ ﺍﻟﻨﺎﺘﺠـﺔ‬
‫ﻓﺈﻥ ﻟﻡ ﺘﻜﻥ ﻤﺴﺘﻘﺭﺓ ﻨﻁﺒﻕ ﻤﺭﺸﺢ ﺍﻟﻔﺭﻭﻕ ﺍﻷﻭﻟﻰ ﻤﺭﺓ ﺜﺎﻨﻴﺔ‪ ،‬ﻭﻨﻌﻴﺩ ﺍﻟﻜـﺭﺓ ﺤﺘـﻰ ﺘـﺼﺒﺢ ﺍﻟﺴﻠـﺴﻠﺔ‬
‫ﻤﺴﺘﻘﺭﺓ‪.‬‬
‫ﻭﻨﺸﻴﺭ ﻫﻨﺎ ﺃﻴﻀﺎﹰ ﺇﻟﻰ ﺃﻨﻪ ﻴﻤﻜﻥ ﺃﻥ ﻨﻀﻴﻑ ﺇﻟﻰ ﺍﻟﻤﻌﺎﺩﻟﺔ ﺍﻟﺴﺎﺒﻘﺔ ﺤﺩﺍﹰ ﺜﺎﺒﺘﺎﹰ‪ ،‬ﺃﻭ ﺤﺩﺍﹰ ﺜﺎﺒﺘﺎﹰ ﻭﺍﻨﺤﺩﺍﺭﺍﹰ ﺨﻁﻴﺎﹰ‬
‫ﺒﺎﻟﺯﻤﻥ‪ ،‬ﻭﻓﻲ ﻫﺫﻩ ﺍﻟﺤﺎﻟﺔ ﺘﻜﻭﻥ ﺍﻟﻘﻴﻡ ﺍﻟﺠﺩﻭﻟﻴﺔ ﻤﺨﺘﻠﻔﺔ ﺒﺤﺴﺏ ﺍﻟﻤﻌﺎﺩﻟﺔ ﺍﻟﻤﺴﺘﺨﺩﻤﺔ‪.‬‬

‫‪ : 2-1-2‬ﺘﺤﺩﻴﺩ ﻋﺩﺩ ﻤﺩﺩ ﺍﻟﺘﺒﺎﻁﺅ ﺍﻟﺯﻤﻨﻲ ﻓﻲ ﻨﻤﻭﺫﺝ ‪: VAR‬‬

‫ﻨﻤﻭﺫﺠﻪ ﻟﻡ ﻴﻌﻁِ ﺃﻱ ﺘﺤﺩﻴﺩ ﻓﻴﻤﺎ ﻴﺘﻌﻠﻕ ﺒﻁﻭل ﻤﺩﺓ ﺍﻟﺘﺒﺎﻁﺅ ﺍﻟﺯﻤﻨﻲ ﺍﻟﺘﻲ ﻴﻤﻜﻥ ﺘﻁﺒﻴﻘﻬـﺎ‬ ‫‪Sims‬‬ ‫ﻋﻨﺩﻤﺎ ﻗﺩﻡ‬
‫ﻋﻠﻰ ﻤﺘﻐﻴ‪‬ﺭﺍﺕ ﺍﻟﻨﻅﺎﻡ‪ .‬ﻤﻥ ﺃﺠل ﺘﺠﻨﺏ ﺍﻟﺤﻜﻡ ﺍﻟﺸﺨﺼﻲ ﻓﻴﻤﺎ ﻴﺘﻌﻠﻕ ﺒﻁﻭل ﻤﺩﺓ ﺍﻟﺘﺒﺎﻁﺅ ﺍﻟﺯﻤﻨﻲ ﺍﻗﺘﹸﺭِﺤ‪‬ـﺕﹾ‬
‫ﻋﺩﺓ ﻤﻌﺎﻴﻴﺭ ﻜﻤﻴﺔ ﻴﻤﻜﻥ ﺍﻻﺴﺘﻌﺎﻨﺔ ﺒﻬﺎ‪:‬‬
‫‪8‬‬
‫)‪:(Final Predictor Error Criterion‬‬ ‫‪FPE‬‬ ‫‪ -1‬ﻤﻌﻴﺎﺭ ﺨﻁﺄ ﺍﻟﺘﻨﺒﺅ ﺍﻟﻨﻬﺎﺌﻲ‬

‫ﻭﻫﻭ ﻴﻌﻁﻰ ﺒﺎﻟﻌﻼﻗﺔ ﺍﻵﺘﻴﺔ‪:‬‬

‫‪6‬‬
‫‪Dickey D. and Fuller W.(1981) 'The likelihood Ratio Statistics for Autoregressive Time Series‬‬
‫‪With a unit Root", Econometrica ,n49: pp .1057-1072‬‬
‫‪7‬‬
‫‪HENIN P.Y. (1989), "Bilans et essais sur la non-Stationnarité des séries Macroéconomiques" révue‬‬
‫‪d' économie politique – n5-pp 661-691‬‬
‫‪8‬‬
‫‪Cromwell, J. B., Hannan M. J., Labys W. C. and Terraza M. (1994), "Multivariate tests for Time‬‬
‫‪Series Models", SAGE publications, Inc. California. pp. 73-75.‬‬

‫‪342‬‬
‫ﻋﺜﻤﺎﻥ ﻨﻘﺎﺭ – ﻤﻨﺫﺭ ﺍﻟﻌﻭﺍﺩ‬ ‫ﻤﺠﻠﺔ ﺠﺎﻤﻌﺔ ﺩﻤﺸﻕ ﻟﻠﻌﻠﻭﻡ ﺍﻻﻗﺘﺼﺎﺩﻴﺔ ﻭﺍﻟﻘﺎﻨﻭﻨﻴﺔ – ﺍﻟﻤﺠﻠﺩ ‪ - 28‬ﺍﻟﻌﺩﺩ ﺍﻟﺜﺎﻨﻲ ‪2012-‬‬

‫‪ N+P −n‬‬

‫‪n‬‬

‫‪FPE = ‬‬ ‫)‪ . det Ω( p‬‬

‫‪ n − P .n ‬‬
‫ﺇﺫﹾ‪:‬‬
‫‪ : Ω‬ﻤﺼﻔﻭﻓﺔ ﺍﻟﺘﺒﺎﻴﻨﺎﺕ ﻭ ﺍﻟﺘﻐﺎﻴﺭﺍﺕ ﺍﻟﻤﻘﺩﺭﺓ ﻟﻠﺒﻭﺍﻗﻲ‬
‫‪ :‬ﻋﺩﺩ ﺍﻟﻤﺘﻐﻴ‪‬ﺭﺍﺕ ﺍﻟﺩﺍﺨﻠﻴﺔ‬ ‫‪n‬‬

‫‪ :‬ﻋﺩﺩ ﺍﻟﻤﺸﺎﻫﺩﺍﺕ ﺍﻟﻜﻠﻴﺔ‬ ‫‪N‬‬

‫‪N‬‬
‫= ‪ K‬ﻋﻠﻰ ﺍﻷﻜﺜﺭ ﻭ ﻤﻥ ﺜﻡ ﺘﺤـﺩ‪‬ﺩ‪ ‬ﻗﻴﻤـﺔ‬ ‫ﻨﻘﻭﻡ ﺒﺤﺴﺎﺏ ‪ FPE‬ﺍﻟﻤﺘﻌﻠﻘﺔ ﺒﺎﻟﻘﻴﻡ ﺍﻟﻤﺘﺘﺎﻟﻴﺔ ﻟـ‪ P‬ﺤﺘﻰ‬
‫‪10‬‬
‫ﺍﻷﺼﻐﺭ ﻭﺃﺨﺫ ﻋﺩﺩ ﻤﺩﺩ ﺍﻟﺘﺒﺎﻁﺅ ﺍﻟﺯﻤﻨﻲ ﻤﻨﻬﺎ ﺃﻱ‪:‬‬ ‫‪FPE‬‬
‫‪K‬‬
‫) ‪FPE ( P0 ) = Min FPE ( P‬‬
‫‪P =1‬‬

‫)‪:(AKAIKE Information Criterion‬‬ ‫‪9‬‬

‫ﻟـ‪AKAIKE‬‬ ‫‪ -2‬ﻤﻌﻴﺎﺭ ﺍﻟﻤﻌﻠﻭﻤﺎﺕ‬
‫ﻭ ﻫﻭ ﻴﺤﺴﺏ ﺒﺎﻟﻌﻼﻗﺔ ﺍﻵﺘﻴﺔ‪:‬‬
‫‪n2P‬‬
‫(‪AIC ( P ) = Log (det Ω( P )) + 2‬‬ ‫)‬
‫‪N‬‬
‫ﺍﻟﺭﻤﻭﺯ ﻟﻬﺎ ﻨﻔﺱ ﺍﻟﻤﻌﻨﻰ ﺍﻟﻤﺸﺎﺭ ﺇﻟﻴﻪ ﺴﺎﺒﻘﺎﹰ‪ .‬ﻨﺨﺘﺎﺭ ﺒﻌﺩ ﺫﻟﻙ ‪ P0‬ﺍﻟﺘﻲ ﺘﺤﻘﻕ ﺍﻟﻌﻼﻗﺔ‪:‬‬
‫‪K‬‬
‫) ‪AIC ( P0 ) = Min AIC ( P‬‬
‫‪P =1‬‬
‫‪10‬‬
‫)‪:(Bayesian Information Criterion‬‬ ‫‪BIC‬‬ ‫‪ -3‬ﻤﻌﻴﺎﺭ ﺍﻟﻤﻌﻠﻭﻤﺎﺕ ﺍﻟﺒﺎﻴﺯﻴﺔ‬
‫ﻴﺤﺴﺏ ﻤﻥ ﺍﻟﻌﻼﻗﺔ ﺍﻵﺘﻴﺔ‪:‬‬
‫‪ n 2 .P. log N ‬‬
‫‪BIC ( P ) = Log (det Ω( P )) + ‬‬ ‫‪‬‬
‫‪‬‬ ‫‪N‬‬ ‫‪‬‬
‫ﺜﻡ ﻨﺄﺨﺫ ﻋﺩﺩ ﻤﺩﺩ ﺍﻟﺘﺒﺎﻁﺅ ﺍﻟﺯﻤﻨﻲ ﺒﺤﻴﺙ ﺘﺤﻘﻕ ‪:‬‬
‫‪K‬‬
‫) ‪BIC ( P0 ) = Min BIC ( P‬‬
‫‪P =1‬‬

‫‪ 9‬ﺍﻟﻤﺭﺠﻊ ﺍﻟﺴﺎﺒﻕ ﻨﻔﺴﻪ‪.‬‬

‫‪10‬‬
‫‪LARDIC S. et MIGNON V. ,(2002) "Econométrie des séries temporelles macroéconomiques et‬‬
‫‪financières" Ed. Economica-Paris. pp. 97.‬‬

‫‪343‬‬
 ‫ﺍﺴﺘﺨﺩﺍﻡ ﻨﻤﺎﺫﺝ ‪ VAR‬ﻓﻲ ﺍﻟﺘﻨﺒﺅ ﻭﺩﺭﺍﺴﺔ ﺍﻟﻌﻼﻗﺔ ﺍﻟﺴﺒﺒﻴﺔ ﺒﻴﻥ ﺇﺠﻤﺎﻟﻲ ﺍﻟﻨﺎﺘﺞ ﺍﻟﻤﺤﻠﻲ ﻭﺇﺠﻤﺎﻟﻲ ﺍﻟﺘﻜﻭﻴﻥ ﺍﻟﺭﺃﺴﻤﺎﻟﻲ ﻓﻲ ﺴﻭﺭﻴﺔ‬

‫)‪:(Hannan & Quinn Information Criterion‬‬ ‫‪11‬‬

‫‪HQIC‬‬ ‫‪ -4‬ﻤﻌﻴﺎﺭ ﺍﻟﻤﻌﻠﻭﻤﺎﺕ‬
‫ﻴﺤﺴﺏ ﻤﻥ ﺍﻟﻌﻼﻗﺔ ﺍﻵﺘﻴﺔ‪:‬‬
‫‪‬‬ ‫‪log . log N ‬‬
‫‪HQIC( P ) = Log (det Ω( P )) +  2n 2 Pc‬‬ ‫‪‬‬
‫‪‬‬ ‫‪N‬‬ ‫‪‬‬
‫ﺇ ﹾﺫ ‪ : c‬ﺘﻤﺜل ﻤﺅﺸﺭ ﻟﻘﻭﺓ ﺍﻟﻤﻌﻴﺎﺭ ﻭ ﻨﻌﺘﺒﺭﻩ ﻤﺴﺎﻭﻴﺎﹰ ‪ 2‬ﻓﻲ ﺍﻟﺘﻁﺒﻴﻕ ﺍﻟﻌﻤﻠﻲ‪.‬‬
‫ﻨﺄﺨﺫ ﻋﺩﺩ ﻓﺘﺭﺍﺕ ﺍﻟﺘﺒﺎﻁﺅ ﺍﻟﺯﻤﻨﻲ ﺒﺤﻴﺙ ﺘﺤﻘﻕ ‪:‬‬
‫‪K‬‬
‫) ‪HQIC ( P0 ) = Min HQIC ( P‬‬
‫‪P =1‬‬

‫ﻤﻼﺤﻅﺔ‪ :‬ﻴﻤﻜﻥ ﺃﻥ ﻨﺤﺼل ﻓﻲ ﺍﻟﺘﻁﺒﻴﻕ ﺍﻟﻌﻤﻠﻲ ﻋﻠﻰ ﻨﺘﺎﺌﺞ ﻤﺨﺘﻠﻔﺔ ﻤﻥ ﻫﺫﻩ ﺍﻟﻤﻌﺎﻴﻴﺭ‪ ،‬ﻓﻲ ﻫـﺫﻩ ﺍﻟﺤﺎﻟـﺔ‬
‫ﻨﻘﻭﻡ ﺒﺎﺨﺘﻴﺎﺭ ﺍﻟﺘﺒﺎﻁﺅ ﺍﻟﺯﻤﻨﻲ ﺍﻟﺫﻱ ﺤﺼﻠﻨﺎ ﻋﻠﻴﻪ ﻓﻲ ﺍﻟﻌﺩﺩ ﺍﻷﻜﺒﺭ ﻤﻥ ﺍﻟﻤﻌﺎﻴﻴﺭ‪.‬‬
‫‪ : 3-1-2‬ﺍﺨﺘﺒﺎﺭ ﺍﻟﺴﺒﺒﻴﺔ‪:‬‬
‫ﻤﻔﻬﻭﻡ ﺍﻟﺴﺒﺒﻴﺔ ﻓﻲ ﺍﻻﻗﺘﺼﺎﺩ ﺍﻟﻘﻴﺎﺴﻲ ﻋﺎﻡ ‪ ، 1969‬ﻭﻫﺫﺍ ﺍﻟﻤﻔﻬﻭﻡ ﻴﺴﻤﺢ ﺒﺎﻟﺘﻤﻴﻴﺯ ﺒـﻴﻥ‬ ‫‪Granger‬‬ ‫ﺃﺩﺨل‬
‫ﺍﻟﻤﺘﻐﻴ‪‬ﺭﺍﺕ ﺍﻟﺩﺍﺨﻠﻴﺔ ﻭﺍﻟﻤﺘﻐﻴ‪‬ﺭﺍﺕ ﺍﻟﺨﺎﺭﺠﻴﺔ‪ .‬ﺴﻨﺭﻯ ﻜﻴﻑ ﻴﻁﺒﻕ ﻫﺫﺍ ﺍﻟﻤﻔﻬﻭﻡ ﻋﻠﻰ ﻤﺴﺘﻭﻯ ﻜل ﻤﻌﺎﺩﻟﺔ ﻤﻥ‬
‫‪.‬‬ ‫‪VAR‬‬ ‫ﻤﻌﺎﺩﻻﺕ ﺍﻟﻨﻤﻭﺫﺝ‬

‫ﺴﻨﺒﺩﺃ ﺒﺘﻌﺭﻴﻑ ﺍﻟﺴﺒﺒﻴﺔ ﻭﻤﻥ ﺜﻡ ﺍﻟﻁﺭﺍﺌﻕ ﺍﻟﺘﻲ ﺘﺴﻤﺢ ﻟﻨﺎ ﺒﺎﺨﺘﺒﺎﺭﻫﺎ‪.‬‬

‫‪ : 1-3-1-2‬ﺘﻌﺭﻴﻑ ‪ Granger‬ﻟﻠﺴﺒﺒﻴﺔ‪:12‬‬
‫ﻨﻘﻭل ﻋﻥ ﺍﻟﻤﺘﻐﻴ‪‬ﺭ ﺍﻟﻌﺸﻭﺍﺌﻲ ‪ X‬ﺃﺇﻨﻪ ﻴﺴﺒﺏ ﺍﻟﻤﺘﻐﻴ‪‬ﺭ ﺍﻟﻌﺸﻭﺍﺌﻲ ‪ Y‬ﺇﺫﺍ ﻜﺎﻨﺕ ﻫﻨﺎﻙ ﻤﻌﻠﻭﻤﺎﺕ ﻓﻲ ﻤﺎﻀﻲ‬
‫‪.‬‬ ‫‪Y‬‬ ‫‪ X‬ﻤﻔﻴﺩﺓ ﻓﻲ ﺍﻟﺘﻨﺒﺅ ﺒـ‪ ، Y‬ﻭ ﻫﺫﻩ ﺍﻟﻤﻌﻠﻭﻤﺎﺕ ﻏﻴﺭ ﻤﻭﺠﻭﺩﺓ ﻓﻲ ﻤﺎﻀﻲ‬
‫ﻟﺩﻴﻨﺎ ﻫﻨﺎ ﻤﺴﻠﻤﺘﺎﻥ ﻫﻤﺎ‪:‬‬
‫‪ - 1‬ﺍﻟﺴﺒﺒﻴﺔ ﻻ ﺘﻁﺒﻕ ﺇﻻ ﻋﻠﻰ ﻤﺘﻐﻴ‪‬ﺭﺍﺕ ﻋﺸﻭﺍﺌﻴﺔ‪.‬‬
‫‪ - 2‬ﺍﻟﻤﺎﻀﻲ ﻭﺍﻟﺤﺎﻀﺭ ﻴﻤﻜﻥ ﺃﻥ ﻴﺴﺒﺏ ﺍﻟﻤﺴﺘﻘﺒل ﻭﺍﻟﻌﻜﺱ ﻏﻴﺭ ﻤﻤﻜﻥ‪.‬‬
‫ﻓﺈﺫﺍ ﺭﻤﺯﻨﺎ ﻟﻠﻤﻌﻠﻭﻤﺎﺕ ﺍﻟﻤﺤﺘﻭﺍﺓ ﻓﻲ ﻤﺎﻀﻲ ﺍﻟﺴﻴﺎﻕ ﺍﻟﻌﺸﻭﺍﺌﻲ ‪ X‬ﻭ ‪ Y‬ﻋﻠﻰ ﺍﻟﺘﻭﺍﻟﻲ ﻜﻤﺎ ﻴﺄﺘﻲ‪:‬‬
‫}‪Xt = {Xt , Xt −1 ,...‬‬
‫~‬

‫}‪Yt = {Yt , Yt −1 ,...‬‬

‫~‬

‫‪ 11‬ﺍﻟﻤﺭﺠﻊ ﺍﻟﺴﺎﺒﻕ ﻨﻔﺴﻪ‪.‬‬

‫‪12‬‬
‫‪GOURIEROUX C. et MONFORT A., (1990) "Séries Temporelles et Modèles Dynamiques " Ed.‬‬
‫‪Economica-Paris. pp. 442-446.‬‬

‫‪344‬‬
‫ﻋﺜﻤﺎﻥ ﻨﻘﺎﺭ – ﻤﻨﺫﺭ ﺍﻟﻌﻭﺍﺩ‬ ‫ﻤﺠﻠﺔ ﺠﺎﻤﻌﺔ ﺩﻤﺸﻕ ﻟﻠﻌﻠﻭﻡ ﺍﻻﻗﺘﺼﺎﺩﻴﺔ ﻭﺍﻟﻘﺎﻨﻭﻨﻴﺔ – ﺍﻟﻤﺠﻠﺩ ‪ - 28‬ﺍﻟﻌﺩﺩ ﺍﻟﺜﺎﻨﻲ ‪2012-‬‬

‫ﻭ ﺭﻤﺯﻨﺎ ﻟﺨﻁﺄ ﺍﻟﺘﻨﺒﺅ ﺒﺎﻻﻋﺘﻤﺎﺩ ﻋﻠﻰ ﺍﻟﻤﻌﻠﻭﻤﺎﺕ ﺍﻟﻤﺘﻭﻓﺭﺓ ﻜﻤﺎ ﻴﺄﺘﻲ‪:‬‬

‫)‪e( X / inf) = X − E ( X / inf‬‬
‫)‪e(Y / inf) = Y − E (Y / inf‬‬
‫ﻟﺩﺭﺍﺴﺔ ﺍﻟﺴﺒﺒﻴﺔ‪ ،‬ﻓﻬﻭ ﻴﺭﻯ ﺃﻨﻪ ﻜﻠﻤﺎ ﻜـﺎﻥ ﻫـﺫﺍ ﺍﻟﺘﺒـﺎﻴﻥ‬ ‫]‪V[e‬‬ ‫ﻋﻠﻰ ﺘﺒﺎﻴﻥ ﺨﻁﺄ ﺍﻟﺘﻨﺒﺅ‬ ‫‪Granger‬‬ ‫ﻴﻌﺘﻤﺩ‬
‫ﻀﻌﻴﻔﺎﹰ ﻜﺎﻥ ﺍﻟﻤﺘﻐﻴ‪‬ﺭ ﻤﻔﺴﺭﺍﹰ ﺘﻔﺴﻴﺭﺍﹰ ﺠﻴﺩﺍﹰ‪ ،‬ﻭﺒﻨﺎﺀ‪ ‬ﻋﻠﻰ ﺫﻟﻙ ﻓﻬﻭ ﻴﻤﻴﺯ ﺒﻴﻥ ﺃﺭﺒﻌﺔ ﺃﻨﻭﺍﻉ ﻤﻥ ﺍﻟﺴﺒﺒﻴﺔ‪:‬‬
‫ﺇﺫﺍ ﺘﺤﻘﻘﺕ ﺍﻟﻤﺘﺭﺍﺠﺤﺔ ﺍﻟﺘﺎﻟﻴﺔ‪:‬‬ ‫‪Y‬‬ ‫ﺘﺴﺒﺏ‬ ‫‪X‬‬ ‫ﺍﻟﺴﺒﺒﻴﺔ ﻭﺤﻴﺩﺓ ﺍﻻﺘﺠﺎﻩ‪ :‬ﻨﻘﻭل ﺃﻥ‬ ‫‪-1‬‬
‫~ ~‬ ‫~‬
‫]) ‪V[e(Yt / Yt −1 , Xt −1 )] < V[e(Yt / Yt −1‬‬
‫ﻓﻲ ﺍﻟﻠﺤﻅﺔ ‪ ،t‬ﻭﻫﻭ ﺃﻓﻀل ﻤﻥ ﺍﻻﻋﺘﻤﺎﺩ ﻓﻘﻁ ﻋﻠﻰ ﻤﺎﻀﻲ ‪.Y‬‬ ‫‪Y‬‬ ‫ﻴﺤﺴﻥ ﻤﻥ ﺘﻨﺒﺅ‬ ‫‪X‬‬ ‫ﻭﻫﺫﺍ ﻴﻌﻨﻲ ﺃﻥ ﻤﺎﻀﻲ‬
‫‪ ،‬ﻭﺘﻜﻭﻥ ﻟﺩﻴﻨﺎ ﺍﻟﻤﺘﺭﺍﺠﺤـﺎﺕ‬ ‫‪X‬‬ ‫ﻭ‪ Y‬ﺘﺴﺒﺏ‬ ‫‪Y‬‬ ‫ﺘﺴﺒﺏ‬ ‫‪X‬‬ ‫ﺍﻟﺴﺒﺒﻴﺔ ﺒﺎﻻﺘﺠﺎﻫﻴﻥ‪ :‬ﻭﻫﻲ ﺘﻌﻨﻲ ﺃﻥ‬ ‫‪-2‬‬
‫ﺍﻵﺘﻴﺔ ﻤﺤﻘﻘﺔ‪:‬‬
‫~ ~‬ ‫~‬
‫]) ‪V[e(Yt / Yt −1 , Xt −1 )] < V[e(Yt / Yt −1‬‬
‫~ ~‬ ‫~‬
‫]) ‪V[e( Xt / Xt −1 , Yt −1 )] < V[e( Xt / Xt −1‬‬
‫‪.‬‬ ‫‪X‬‬ ‫ﻴﺤﺴﻥ ﺘﻨﺒﺅ‬ ‫‪Y‬‬ ‫‪ ،‬ﻭﺃﻥ ﻤﺎﻀﻲ‬ ‫‪Y‬‬ ‫ﻴﺤﺴﻥ ﺘﻨﺒﺅ‬ ‫‪X‬‬ ‫ﻭﻫﺫﺍ ﻴﻌﻨﻲ ﺃﻥ ﻤﺎﻀﻲ‬
‫‪ ،‬ﻭﺘﻜـﻭﻥ‬ ‫‪Y‬‬ ‫ﺘﺴﺒﺏ ﺍﻟﻘﻴﻤﺔ ﺍﻟﺤﺎﻟﻴﺔ ﻟـ‬ ‫‪X‬‬ ‫ﺍﻟﺴﺒﺒﻴﺔ ﺍﻵﻨﻴﺔ‪ :‬ﻭﻫﻲ ﺘﻌﻨﻲ ﺃﻥ ﺍﻟﻘﻴﻤﺔ ﺍﻟﺤﺎﻟﻴﺔ ﻟـ‬ ‫‪-3‬‬
‫ﺍﻟﻤﺘﺭﺍﺠﺤﺔ ﺍﻵﺘﻴﺔ ﻤﺤﻘﻘﺔ‪:‬‬
‫~ ~‬ ‫~ ~‬
‫]) ‪V[e(Yt / Yt −1 , Xt )] < V[e(Yt / Yt −1 , Xt −1‬‬

‫ﺍﻟﺴﺒﺒﻴﺔ ﺍﻟﻤﺘﺒﺎﻁﺌﺔ )ﺍﻵﺠﻠﺔ (‪ :‬ﺍﻟﻘﻴﻡ ﺍﻟﻤﺎﻀﻴﺔ ﻟـ‪ X‬ﺘﺴﺒﺏ ﺍﻟﻘﻴﻤﺔ ﺍﻟﺤﺎﻀﺭﺓ ﻟـ‪ ،Y‬ﻭ ﻭﻤﻥ ﺜـﻡ‪‬‬ ‫‪-4‬‬
‫ﺘﻜﻭﻥ ﺍﻟﻤﺘﺭﺍﺠﺤﺔ ﺍﻵﺘﻴﺔ ﻤﺤﻘﻘﺔ‪:‬‬
‫~ ~‬ ‫~‬
‫]) ‪V[e(Yt / Yt −1 , Xt − m )] < V[e(Yt / Yt −1‬‬
‫‪ 2-3-1-2‬ﺨﻁﻭﺍﺕ ﺍﺨﺘﺒﺎﺭ ﺍﻟﺴﺒﺒﻴﺔ ‪:13Granger‬‬
‫‪ -1‬ﻨﻘﻭﻡ ﺒﺘﻘﺩﻴﺭ ﺍﻟﻤﻌﺎﺩﻟﺔ ﺍﻟﺘﺎﻟﻴﺔ ﺒﺎﺴﺘﺨﺩﺍﻡ ﻁﺭﻴﻘﺔ ﺍﻟﻤﺭﺒﻌﺎﺕ ﺍﻟﺼﻐﺭﻯ‪:‬‬
‫‪Yt = Φ1 ( B).Yt + Φ 2 ( B). Xt + ε t‬‬

‫ﺇﺫﹾ‪ Φ1 ( B) = ∑ φ1i .Bi :‬ﻭ ‪Φ 2 ( B) = ∑ φ2i .Bi‬‬

‫‪q‬‬ ‫‪p‬‬

‫‪i =1‬‬ ‫‪i =1‬‬

‫‪SCR1‬‬ ‫ﺜﻡ ﻨﺤﺴﺏ ﻤﺠﻤﻭﻉ ﻤﺭﺒﻌﺎﺕ ﺍﻨﺤﺭﺍﻓﺎﺕ ﺍﻟﻘﻴﻡ ﺍﻟﻔﻌﻠﻴﺔ ﻋﻥ ﺍﻟﻤﻘﺩﺭﺓ ﻭﻨﺭﻤﺯ ﻟﻬﺎ‪:‬‬

‫‪13‬‬
‫‪LARDIC S. et MIGNON V. ,(2002) "Econométrie des séries temporelles macroéconomiques et‬‬
‫‪financières" Ed. Economica-Paris. pp. 99-101.‬‬

‫‪345‬‬
 ‫ﺍﺴﺘﺨﺩﺍﻡ ﻨﻤﺎﺫﺝ ‪ VAR‬ﻓﻲ ﺍﻟﺘﻨﺒﺅ ﻭﺩﺭﺍﺴﺔ ﺍﻟﻌﻼﻗﺔ ﺍﻟﺴﺒﺒﻴﺔ ﺒﻴﻥ ﺇﺠﻤﺎﻟﻲ ﺍﻟﻨﺎﺘﺞ ﺍﻟﻤﺤﻠﻲ ﻭﺇﺠﻤﺎﻟﻲ ﺍﻟﺘﻜﻭﻴﻥ ﺍﻟﺭﺃﺴﻤﺎﻟﻲ ﻓﻲ ﺴﻭﺭﻴﺔ‬

‫‪ -2‬ﻨﻘﻭﻡ ﺒﺘﻘﺩﻴﺭ ﺍﻟﻤﻌﺎﺩﻟﺔ ﺍﻵﺘﻴﺔ‪:‬‬

‫‪Yt = Φ 1 ( B).Yt + ε t‬‬
‫‪.‬‬ ‫‪SCR2‬‬ ‫ﺜﻡ ﻨﺤﺴﺏ ﻤﺠﻤﻭﻉ ﻤﺭﺒﻌﺎﺕ ﺍﻨﺤﺭﺍﻓﺎﺕ ﺍﻟﻘﻴﻡ ﺍﻟﻔﻌﻠﻴﺔ ﻋﻥ ﺍﻟﻤﻘﺩﺭﺓ ﻭ ﻨﺭﻤﺯ ﻟﻬﺎ‪:‬‬
‫‪ -3‬ﻨﺤﺴﺏ ﺇﺤﺼﺎﺌﻴﺔ ﺍﻻﺨﺘﺒﺎﺭ ‪ Fc‬ﻤﻥ ﺍﻟﻌﻼﻗﺔ‪:‬‬
‫= ‪Fc‬‬
‫‪(SCR2 − SCR1) / p‬‬
‫) ‪SCR1 /( M − N‬‬
‫ﺇﺫﹾ‪ M = T − Max( p, q ) :‬ﻭ ‪N = p + q + 2‬‬
‫‪ :‬ﻋﺩﺩ ﺍﻟﻤﺸﺎﻫﺩﺍﺕ‪.‬‬ ‫‪T‬‬

‫‪ :‬ﻋﺩﺩ ﺍﻟﺘﺒﺎﻁﺅﺍﺕ ﺍﻟﺯﻤﻨﻴﺔ ﻟﻠﻤﺘﻐﻴ‪‬ﺭﺍﺕ ﺍﻟﺩﺍﺨﻠﻴﺔ‪.‬‬ ‫‪P‬‬

‫‪ :‬ﻋﺩﺩ ﺍﻟﺘﺒﺎﻁﺅﺍﺕ ﺍﻟﺯﻤﻨﻴﺔ ﻟﻠﻤﺘﻐﻴ‪‬ﺭﺍﺕ ﺍﻟﺨﺎﺭﺠﻴﺔ‪.‬‬ ‫‪q‬‬

‫‪Fα‬‬ ‫ﺍﻟﻤﺤﺴﻭﺒﺔ ﻤﻊ‬ ‫‪Fc‬‬ ‫ﺍﻟﺘﻲ ﺘﻘﻭل‪ :‬ﺇﻥ‪ Xt ‬ﻻ ﺘﺴﺒﺏ ‪ ، Yt‬ﺜﻡ ﻨﻘﺎﺭﻥ‬ ‫‪H0‬‬ ‫‪ -4‬ﻨﻀﻊ ﻓﺭﻀﻴﺔ ﺍﻟﻌﺩﻡ‬
‫ﺍﻟﺠﺩﻭﻟﻴﺔ ﻭﻨﻘﺒل ﻓﺭﻀﻴﺔ ﺍﻟﻌﺩﻡ ﺇﺫﺍ ﻜﺎﻥ‪.Fc < Fα(P, (M-N)) :‬‬
‫ﺜﺎﻟﺜﺎ ‪ -‬ﺍﻟﺤﺎﻟﺔ ﺍﻟﺘﻁﺒﻴﻘﻴﺔ‪:‬‬
‫ﺤﺼﻠﻨﺎ ﻋﻠﻰ ﺴﻠﺴﻠﺘﻴﻥ ﺯﻤﻨﻴﺘﻴﻥ ﺘﻤﺘﺩﺍﻥ ﻤﻥ ﻋﺎﻡ ‪ 1983‬ﺇﻟﻰ ﻋﺎﻡ ‪ 2009‬ﻹﺠﻤﺎﻟﻲ ﺍﻟﻨﺎﺘﺞ ﺍﻟﻤﺤﻠﻲ )‪،(GDP‬‬
‫ﻓﻲ ﺴﻭﺭﻴﺔ‪ ،‬ﻤﻘﺩﺭﺓ ﺒﻤﻼﻴﻴﻥ ﺍﻟﻠﻴﺭﺍﺕ ﺍﻟﺴﻭﺭﻴﺔ ﺒﺄﺴﻌﺎﺭ ﻋﺎﻡ‬ ‫)‪(GFCF‬‬ ‫ﻭﻹﺠﻤﺎﻟﻲ ﺘﻜﻭﻴﻥ ﺭﺃﺱ ﺍﻟﻤﺎل ﺍﻟﺜﺎﺒﺕ‬
‫‪ 2000‬ﺍﻟﺜﺎﺒﺘﺔ‪ ،‬ﻜﻤﺎ ﻴﺒﻴ‪‬ﻥ ﺍﻟﺠﺩﻭل )‪:(1‬‬
‫ﺍﻟﺠﺩﻭل )‪(1‬‬
‫ﺇﺠﻤﺎﻟﻲ ﺍﻟﻨﺎﺘﺞ ﺍﻟﻤﺤﻠﻲ ﻭ ﺇﺠﻤﺎﻟﻲ ﺘﻜﻭﻴﻥ ﺭﺃﺱ ﺍﻟﻤﺎل ﺍﻟﺜﺎﺒﺕ ﻓﻲ ﺴﻭﺭﻴﺔ ﺒﺄﺴﻌﺎﺭ ﻋﺎﻡ ‪ 2000‬ﺍﻟﺜﺎﺒﺘﺔ‪.‬‬
‫)ﻤﻘﺩﺭﺓ ﺒﻤﻼﻴﻴﻥ ﺍﻟﻠﻴﺭﺍﺕ ﺍﻟﺴﻭﺭﻴﺔ(‬
‫‪GDP‬‬ ‫‪GFCF‬‬ ‫ﺍﻟﻌﺎﻡ‬ ‫‪GDP‬‬ ‫‪GFCF‬‬ ‫ﺍﻟﻌﺎﻡ‬ ‫‪GDP‬‬ ‫‪GFCF‬‬ ‫ﺍﻟﻌﺎﻡ‬
‫‪950248‬‬ ‫‪178148‬‬ ‫‪2001‬‬ ‫‪666630‬‬ ‫‪111600‬‬ ‫‪1992‬‬ ‫‪506899.6‬‬ ‫‪160463‬‬ ‫‪1983‬‬
‫‪1006431‬‬ ‫‪196387‬‬ ‫‪2002‬‬ ‫‪700466‬‬ ‫‪122534‬‬ ‫‪1993‬‬ ‫‪486259.8‬‬ ‫‪152720‬‬ ‫‪1984‬‬
‫‪1018708‬‬ ‫‪231944‬‬ ‫‪2003‬‬ ‫‪706745‬‬ ‫‪167874‬‬ ‫‪1994‬‬ ‫‪515986‬‬ ‫‪170136‬‬ ‫‪1985‬‬
‫‪1089027‬‬ ‫‪255767‬‬ ‫‪2004‬‬ ‫‪756404‬‬ ‫‪167846‬‬ ‫‪1995‬‬ ‫‪490475.8‬‬ ‫‪150756‬‬ ‫‪1986‬‬
‫‪1156714‬‬ ‫‪288193‬‬ ‫‪2005‬‬ ‫‪830726‬‬ ‫‪167352‬‬ ‫‪1996‬‬ ‫‪499831.6‬‬ ‫‪101583.5‬‬ ‫‪1987‬‬
‫‪1215082‬‬ ‫‪308669‬‬ ‫‪2006‬‬ ‫‪872461‬‬ ‫‪158944‬‬ ‫‪1997‬‬ ‫‪566140.6‬‬ ‫‪97486.5‬‬ ‫‪1988‬‬
‫‪1284035‬‬ ‫‪283099‬‬ ‫‪2007‬‬ ‫‪931660‬‬ ‫‪164065‬‬ ‫‪1998‬‬ ‫‪515424.6‬‬ ‫‪88697.5‬‬ ‫‪1989‬‬
‫‪1341516‬‬ ‫‪266488‬‬ ‫‪2008‬‬ ‫‪898552‬‬ ‫‪159793‬‬ ‫‪1999‬‬ ‫‪510548‬‬ ‫‪99770‬‬ ‫‪1990‬‬
‫‪1422178‬‬ ‫‪297100‬‬ ‫‪2009‬‬ ‫‪903944‬‬ ‫‪156092‬‬ ‫‪2000‬‬ ‫‪569298.9‬‬ ‫‪102935‬‬ ‫‪1991‬‬

‫ﺍﻟﻤﺼﺩﺭ‪ :‬ﺍﻟﻤﺠﻤﻭﻋﺔ ﺍﻹﺤﺼﺎﺌﻴﺔ ﻓﻲ ﺴﻭﺭﻴﺔ ﻟﻸﻋﻭﺍﻡ‪2008- 1990:‬‬

‫‪346‬‬
‫ﻋﺜﻤﺎﻥ ﻨﻘﺎﺭ – ﻤﻨﺫﺭ ﺍﻟﻌﻭﺍﺩ‬ ‫ﻤﺠﻠﺔ ﺠﺎﻤﻌﺔ ﺩﻤﺸﻕ ﻟﻠﻌﻠﻭﻡ ﺍﻻﻗﺘﺼﺎﺩﻴﺔ ﻭﺍﻟﻘﺎﻨﻭﻨﻴﺔ – ﺍﻟﻤﺠﻠﺩ ‪ - 28‬ﺍﻟﻌﺩﺩ ﺍﻟﺜﺎﻨﻲ ‪2012-‬‬

‫‪.‬‬ ‫‪EViews version 6.0‬‬ ‫ﺴﻨﺴﺘﺨﺩﻡ ﻓﻲ ﺘﺤﻠﻴﻠﻨﺎ ﻟﻬﺎﺘﻴﻥ ﺍﻟﺴﻠﺴﻠﺘﻴﻥ ﺍﻟﺯﻤﻨﻴﺘﻴﻥ ﺍﻟﺤﺯﻤﺔ ﺍﻟﺒﺭﻤﺠﻴﺔ‪:‬‬
‫‪ -1-3‬ﻋﺭﺽ ﺍﻟﺴﻼﺴل ﺍﻟﺯﻤﻨﻴﺔ‪:‬‬
‫ﻨﺒﺩﺃ ﺒﺎﻟﺘﻌﺭﻑ ﻋﻠﻰ ﺍﻟﺸﻜل ﺍﻟﺒﻴﺎﻨﻲ ﻟﻜل ﻤﻥ ﺍﻟﺴﻠﺴﻠﺘﻴﻥ‪ .‬ﺘﻅﻬﺭ ﺍﻷﺸﻜﺎل )‪ (1‬ﻭ )‪ (2‬ﺒﺼﻭﺭﺓ ﻭﺍﻀﺤﺔ ﻋﺩﻡ‬
‫ﺍﺴﺘﻘﺭﺍﺭ ﺍﻟﺴﻠﺴﻠﺘﻴﻥ‪ ،‬ﻭﻟﻜﻨﻬﺎ ﻻ ﺘﺒﻴ‪‬ﻥ ﻫل ﻴﻌﻭﺩ ﻋﺩﻡ ﺍﻻﺴﺘﻘﺭﺍﺭ ﻟﻭﺠﻭﺩ ﺠﺫﺭ ﺍﻟﻭﺤﺩﺓ ﺃﻡ ﻻ‪ ،‬ﻭﻤﻥ ﺜﻡ‪ ‬ﻻﺒﺩ‪ ‬ﻟﻨﺎ‬
‫ﻤﻥ ﺍﺨﺘﺒﺎﺭ ﺠﺫﺭ ﺍﻟﻭﺤﺩﺓ‪.‬‬

‫‪GDP‬‬
‫‪1,600,000‬‬

‫‪1,400,000‬‬

‫‪1,200,000‬‬

‫‪1,000,000‬‬

‫‪800,000‬‬

‫‪600,000‬‬

‫‪400,000‬‬
‫‪84‬‬ ‫‪86‬‬ ‫‪88‬‬ ‫‪90‬‬ ‫‪92‬‬ ‫‪94‬‬ ‫‪96‬‬ ‫‪98‬‬ ‫‪00‬‬ ‫‪02‬‬ ‫‪04‬‬ ‫‪06‬‬ ‫‪08‬‬ ‫اﻟﺰﻣﻦ‬
‫ﺍﻟﺸﻜل )‪ (1‬ﺘﻁﻭﺭ ﺇﺠﻤﺎﻟﻲ ﺍﻟﻨﺎﺘﺞ ﺍﻟﻤﺤﻠﻲ ﻤﻥ ﻋﺎﻡ ‪ 1983‬ﺇﻟﻰ ‪2009‬‬

‫‪GFCF‬‬
‫‪320,000‬‬

‫‪280,000‬‬

‫‪240,000‬‬

‫‪200,000‬‬

‫‪160,000‬‬

‫‪120,000‬‬

‫‪80,000‬‬
‫‪84‬‬ ‫‪86‬‬ ‫‪88‬‬ ‫‪90‬‬ ‫‪92‬‬ ‫‪94‬‬ ‫‪96‬‬ ‫‪98‬‬ ‫‪00‬‬ ‫‪02‬‬ ‫‪04‬‬ ‫‪06‬‬ ‫‪08‬‬
‫اﻟﺰﻣﻦ‬

‫ﺍﻟﺸﻜل )‪ :(2‬ﺘﻁﻭﺭ ﺇﺠﻤﺎﻟﻲ ﺘﻜﻭﻴﻥ ﺭﺃﺱ ﺍﻟﻤﺎل ﺍﻟﺜﺎﺒﺕ ﻤﻥ ﻋﺎﻡ ‪ 1983‬ﺇﻟﻰ ‪2009‬‬

‫‪347‬‬
 ‫ ﻓﻲ ﺍﻟﺘﻨﺒﺅ ﻭﺩﺭﺍﺴﺔ ﺍﻟﻌﻼﻗﺔ ﺍﻟﺴﺒﺒﻴﺔ ﺒﻴﻥ ﺇﺠﻤﺎﻟﻲ ﺍﻟﻨﺎﺘﺞ ﺍﻟﻤﺤﻠﻲ ﻭﺇﺠﻤﺎﻟﻲ ﺍﻟﺘﻜﻭﻴﻥ ﺍﻟﺭﺃﺴﻤﺎﻟﻲ ﻓﻲ ﺴﻭﺭﻴﺔ‬VAR ‫ﺍﺴﺘﺨﺩﺍﻡ ﻨﻤﺎﺫﺝ‬

:‫ ﺍﺨﺘﺒﺎﺭ ﺍﻻﺴﺘﻘﺭﺍﺭ‬-2-3
‫ﻋﻠﻰ ﻜل ﻤـﻥ ﺍﻟﺴﻠـﺴﻠﺘﻴﻥ‬ ADF ‫ﻓﻴﻠﻠﺭ ﺍﻟﻤﻭﺴﻊ‬- ‫ﻨﻁﺒﻕ ﺍﺨﺘﺒﺎﺭ ﺠﺫﺭ ﺍﻟﻭﺤﺩﺓ ﺃﻭ ﻤﺎ ﻴﻌﺭﻑ ﺒﺎﺨﺘﺒﺎﺭ ﺩﻴﻜﻲ‬
:‫ﻓﻨﺤﺼل ﻋﻠﻰ ﺍﻟﻨﺘﺎﺌﺞ ﺍﻵﺘﻴﺔ‬
GDP ‫ﻟﻠﺴﻠﺴﻠﺔ‬ ADFِ ‫ﺍﺨﺘﺒﺎﺭ‬ (2) ‫ﺍﻟﺠﺩﻭل‬
Null Hypothesis: GDP has a unit root
Exogenous: Constant, Linear Trend
Lag Length: 2 (Fixed)

t-Statistic Prob.*

Augmented Dickey-Fuller test statistic -1.629860 0.7503

Test critical values: 1% level -4.394309
5% level -3.612199
10% level -3.243079

*MacKinnon (1996) one-sided p-values.

EViews 6 ‫ﻨﺘﺎﺌﺞ ﺤﺴﺎﺒﺎﺕ ﺒﺭﻨﺎﻤﺞ‬:‫ﺍﻟﻤﺼﺩﺭ‬
‫( ﺃﺼﻐﺭ ﻤﻥ ﺍﻟﻘﻴﻡ ﺍﻟﻤﻁﻠﻘﺔ ﻟﻘـﻴﻡ‬1.62986) ‫ﺍﻟﻤﺤﺴﻭﺒﺔ‬ ADF‫ﻟـ‬ ‫( ﺃﻥ ﺍﻟﻘﻴﻤﺔ ﺍﻟﻤﻁﻠﻘﺔ‬2) ‫ﻨﺠﺩ ﻤﻥ ﺍﻟﺠﺩﻭل‬
‫ ﻨﻘﺒـل‬:‫ ﺃﻱ‬،‫ ﺇﺫﺍﹰ ﻻ ﻨﺴﺘﻁﻴﻊ ﺭﻓﺽ ﻓﺭﻀﻴﺔ ﺍﻟﻌـﺩﻡ‬.‫ﺍﻟﻨﻅﺭﻴﺔ ﻋﻨﺩ ﻤﺴﺘﻭﻴﺎﺕ ﺍﻟﺩﻻﻟﺔ ﺍﻟﻤﺨﺘﻠﻔﺔ‬ Mackinnon

. GDP ‫ﻭﺠﻭﺩ ﺠﺫﺭ ﺍﻟﻭﺤﺩﺓ ﻓﻲ ﺍﻟﺴﻠﺴﻠﺔ‬

GFCF ‫ﻟﻠﺴﻠﺴﻠﺔ‬ ADFِ ‫ﺍﺨﺘﺒﺎﺭ‬ (3) ‫ﺍﻟﺠﺩﻭل‬
Null Hypothesis: GFCF has a unit root
Exogenous: Constant, Linear Trend
Lag Length: 1 (Automatic based on SIC, MAXLAG=6)

t-Statistic Prob.*

Augmented Dickey-Fuller test statistic -2.296223 0.4206

Test critical values: 1% level -4.374307
5% level -3.603202
10% level -3.238054

*MacKinnon (1996) one-sided p-values.

EViews 6 ‫ﻨﺘﺎﺌﺞ ﺤﺴﺎﺒﺎﺕ ﺒﺭﻨﺎﻤﺞ‬:‫ﺍﻟﻤﺼﺩﺭ‬

348
‫ﻋﺜﻤﺎﻥ ﻨﻘﺎﺭ – ﻤﻨﺫﺭ ﺍﻟﻌﻭﺍﺩ‬ ‫ﻤﺠﻠﺔ ﺠﺎﻤﻌﺔ ﺩﻤﺸﻕ ﻟﻠﻌﻠﻭﻡ ﺍﻻﻗﺘﺼﺎﺩﻴﺔ ﻭﺍﻟﻘﺎﻨﻭﻨﻴﺔ – ﺍﻟﻤﺠﻠﺩ ‪ - 28‬ﺍﻟﻌﺩﺩ ﺍﻟﺜﺎﻨﻲ ‪2012-‬‬

‫ﺍﻟﻤﺤﺴﻭﺒﺔ )‪ (2.296223‬ﺃﺼﻐﺭ ﻤﻥ ﺍﻟﻘﻴﻡ ﺍﻟﻤﻁﻠﻘﺔ ﻟﻘـﻴﻡ‬ ‫ﻟـ‪ADF‬‬ ‫ﻨﺠﺩ ﻤﻥ ﺍﻟﺠﺩﻭل )‪ (3‬ﺃﻥ ﺍﻟﻘﻴﻤﺔ ﺍﻟﻤﻁﻠﻘﺔ‬
‫ﺍﻟﻨﻅﺭﻴﺔ ﻋﻨﺩ ﻤﺴﺘﻭﻴﺎﺕ ﺍﻟﺩﻻﻟﺔ ﺍﻟﻤﺨﺘﻠﻔﺔ‪ .‬ﺇﺫﺍﹰ ﻻ ﻨﺴﺘﻁﻴﻊ ﺭﻓﺽ ﻓﺭﻀﻴﺔ ﺍﻟﻌـﺩﻡ ‪ ،‬ﺃﻱ‪ :‬ﻨﻘﺒـل‬ ‫‪Mackinnon‬‬

‫‪.‬‬ ‫‪GFCF‬‬ ‫ﻭﺠﻭﺩ ﺠﺫﺭ ﺍﻟﻭﺤﺩﺓ ﻓﻲ ﺍﻟﺴﻠﺴﻠﺔ‬

‫ﻟﻠﺘﺨﻠﺹ ﻤﻥ ﺠﺫﺭ ﺍﻟﻭﺤﺩﺓ ﻨﻁﺒﻕ ﻤﺭﺸﺢ ﺍﻟﻔﺭﻭﻕ ﺍﻷﻭﻟﻰ ﻋﻠﻰ ﻜل ﻤﻥ ﺍﻟﺴﻠﺴﻠﺘﻴﻥ ﻓﻨﺤﺼل ﻋﻠﻰ ﺴﻠﺴﻠﺘﻴﻥ‬
‫ﻤﺴﺘﻘﺭﺘﻴﻥ ﻜﻤﺎ ﻴﺒﻴ‪‬ﻥ ﺍﻟﺸﻜﻼﻥ )‪ (3‬ﻭ )‪.(4‬‬
‫)‪D(GDP‬‬
‫‪100,000‬‬

‫‪80,000‬‬

‫‪60,000‬‬

‫‪40,000‬‬

‫‪20,000‬‬

‫‪0‬‬

‫‪-20,000‬‬

‫‪-40,000‬‬

‫‪-60,000‬‬ ‫اﻟﺰﻣﻦ‬
‫‪84‬‬ ‫‪86‬‬ ‫‪88‬‬ ‫‪90‬‬ ‫‪92‬‬ ‫‪94‬‬ ‫‪96‬‬ ‫‪98‬‬ ‫‪00‬‬ ‫‪02‬‬ ‫‪04‬‬ ‫‪06‬‬ ‫‪08‬‬

‫ﺍﻟﺸﻜل )‪ :(3‬ﺴﻠﺴﻠﺔ ﺍﻟﻔﺭﻭﻕ ﺍﻷﻭﻟﻰ ﻹﺠﻤﺎﻟﻲ ﺍﻟﻨﺎﺘﺞ ﺍﻟﻤﺤﻠﻲ‬

‫)‪D(GFCF‬‬
‫‪60,000‬‬

‫‪40,000‬‬

‫‪20,000‬‬

‫‪0‬‬

‫‪-20,000‬‬

‫‪-40,000‬‬

‫‪-60,000‬‬
‫‪84‬‬ ‫‪86‬‬ ‫‪88‬‬ ‫‪90‬‬ ‫‪92‬‬ ‫‪94‬‬ ‫‪96‬‬ ‫‪98‬‬ ‫‪00‬‬ ‫‪02‬‬ ‫‪04‬‬ ‫‪06‬‬ ‫‪08‬‬ ‫اﻟﺰﻣﻦ‬

‫ﺍﻟﺸﻜل )‪ :(4‬ﺴﻠﺴﻠﺔ ﺍﻟﻔﺭﻭﻕ ﺍﻷﻭﻟﻰ ﻹﺠﻤﺎﻟﻲ ﺘﻜﻭﻴﻥ ﺭﺃﺱ ﺍﻟﻤﺎل ﺍﻟﺜﺎﺒﺕ‬

‫‪349‬‬
 ‫ﺍﺴﺘﺨﺩﺍﻡ ﻨﻤﺎﺫﺝ ‪ VAR‬ﻓﻲ ﺍﻟﺘﻨﺒﺅ ﻭﺩﺭﺍﺴﺔ ﺍﻟﻌﻼﻗﺔ ﺍﻟﺴﺒﺒﻴﺔ ﺒﻴﻥ ﺇﺠﻤﺎﻟﻲ ﺍﻟﻨﺎﺘﺞ ﺍﻟﻤﺤﻠﻲ ﻭﺇﺠﻤﺎﻟﻲ ﺍﻟﺘﻜﻭﻴﻥ ﺍﻟﺭﺃﺴﻤﺎﻟﻲ ﻓﻲ ﺴﻭﺭﻴﺔ‬

‫ﻭﺒﺘﻁﺒﻴﻕ ﺍﺨﺘﺒﺎﺭ ﺠﺫﺭ ﺍﻟﻭﺤﺩﺓ ﻤﺭﺓ ﺃﺨﺭﻯ ﻋﻠﻰ ﺴﻼﺴل ﺍﻟﻔﺭﻭﻕ ﺍﻷﻭﻟﻰ ﻨﺠﺩ ﺃﻥ ﺍﻟﻨﺘﺎﺌﺞ ﺘﺅﻜﺩ ﻋﺩﻡ ﻭﺠـﻭﺩ‬
‫ﺠﺫﺭ ﺍﻟﻭﺤﺩﺓ ﻭ ﺍﺴﺘﻘﺭﺍﺭ ﻫﺎﺘﻴﻥ ﺍﻟﺴﻠﺴﻠﺘﻴﻥ‪.‬‬
‫‪ -3-3‬ﺘﺤﺩﻴﺩ ﻋﺩﺩ ﻤﺩﺩ ﺍﻟﺘﺒﺎﻁﺅ ﺍﻟﺯﻤﻨﻲ‪:‬‬
‫ﻟﺘﺤﺩﻴﺩ ﻋﺩﺩ ﻤﺩﺩ ﺍﻟﺘﺒﺎﻁﺅ ﺍﻟﺯﻤﻨﻲ ﻨﺴﺘﺨﺩﻡ ﺍﻟﻤﻌﺎﻴﻴﺭ ﺍﻟﺘﻲ ﻋﺭﻀﻨﺎﻫﺎ ﻤﺴﺒﻘﺎﹰ ﻭﺍﻟﺘﻲ ﺘﻅﻬﺭ ﻓﻲ ﺍﻟﺠﺩﻭل ﺍﻵﺘﻲ‪:‬‬
‫ﺍﻟﺠﺩﻭل )‪ (4‬ﻤﻌﺎﻴﻴﺭ ﺘﺤﺩﻴﺩ ﻋﺩﺩ ﻤﺩﺩ ﺍﻟﺘﺒﺎﻁﺅ ﺍﻟﺯﻤﻨﻲ ﺍ‬
‫‪VAR Lag Order Selection Criteria‬‬
‫)‪Endogenous variables: D(GDP) D(GFCF‬‬
‫‪Exogenous variables:‬‬
‫‪Date: 05/09/11 Time: 05:55‬‬
‫‪Sample: 1983 2009‬‬
‫‪Included observations: 18‬‬

‫‪Lag‬‬ ‫‪LogL‬‬ ‫‪LR‬‬ ‫‪FPE‬‬ ‫‪AIC‬‬ ‫‪SC‬‬ ‫‪HQ‬‬

‫‪1‬‬ ‫‪-417.0975‬‬ ‫‪NA‬‬ ‫*‪7.18e+17‬‬ ‫‪46.78861‬‬ ‫*‪46.98647‬‬ ‫*‪46.81589‬‬

‫‪2‬‬ ‫‪-415.6685‬‬ ‫‪2.222931‬‬ ‫‪9.68e+17‬‬ ‫‪47.07427‬‬ ‫‪47.46999‬‬ ‫‪47.12884‬‬
‫‪3‬‬ ‫‪-414.2780‬‬ ‫‪1.853944‬‬ ‫‪1.34e+18‬‬ ‫‪47.36422‬‬ ‫‪47.95780‬‬ ‫‪47.44607‬‬
‫‪4‬‬ ‫‪-409.4230‬‬ ‫‪5.394456‬‬ ‫‪1.32e+18‬‬ ‫‪47.26922‬‬ ‫‪48.06066‬‬ ‫‪47.37835‬‬
‫‪5‬‬ ‫‪-400.6541‬‬ ‫‪7.794592‬‬ ‫‪9.05e+17‬‬ ‫*‪46.73934‬‬ ‫‪47.72864‬‬ ‫‪46.87575‬‬
‫‪6‬‬ ‫‪-400.1221‬‬ ‫‪0.354640‬‬ ‫‪1.74e+18‬‬ ‫‪47.12468‬‬ ‫‪48.31184‬‬ ‫‪47.28837‬‬
‫‪7‬‬ ‫‪-398.3717‬‬ ‫‪0.777959‬‬ ‫‪3.67e+18‬‬ ‫‪47.37463‬‬ ‫‪48.75966‬‬ ‫‪47.56561‬‬
‫‪8‬‬ ‫‪-390.4710‬‬ ‫‪1.755717‬‬ ‫‪6.89e+18‬‬ ‫‪46.94122‬‬ ‫‪48.52410‬‬ ‫‪47.15948‬‬

‫‪* indicates lag order selected by the criterion‬‬

‫)‪LR: sequential modified LR test statistic (each test at 5% level‬‬
‫‪FPE: Final prediction error‬‬
‫‪AIC: Akaike information criterion‬‬
‫‪SC: Schwarz information criterion‬‬
‫‪HQ: Hannan-Quinn information criterion‬‬

‫‪EViews 6‬‬ ‫ﺍﻟﻤﺼﺩﺭ‪:‬ﻨﺘﺎﺌﺞ ﺤﺴﺎﺒﺎﺕ ﺒﺭﻨﺎﻤﺞ‬

‫‪350‬‬
‫ﻋﺜﻤﺎﻥ ﻨﻘﺎﺭ – ﻤﻨﺫﺭ ﺍﻟﻌﻭﺍﺩ‬ ‫ﻤﺠﻠﺔ ﺠﺎﻤﻌﺔ ﺩﻤﺸﻕ ﻟﻠﻌﻠﻭﻡ ﺍﻻﻗﺘﺼﺎﺩﻴﺔ ﻭﺍﻟﻘﺎﻨﻭﻨﻴﺔ – ﺍﻟﻤﺠﻠﺩ ‪ - 28‬ﺍﻟﻌﺩﺩ ﺍﻟﺜﺎﻨﻲ ‪2012-‬‬

‫ﺘﺸﻴﺭ ﺇﻟﻰ ﻀﺭﻭﺭﺓ ﺃﺨﺫ ﻓﺠﻭﺓ ﺯﻤﻨﻴﺔ ﻭﺍﺤﺩﺓ‪،‬‬ ‫‪SC‬‬ ‫ﻭ‬ ‫‪HQ‬‬ ‫ﻭ‬ ‫‪FPE‬‬ ‫ﻨﺠﺩ ﻤﻥ ﺍﻟﺠﺩﻭل )‪ (4‬ﺃﻥ ﺍﻟﻤﻌﺎﻴﻴﺭ ﺍﻟﺜﻼﺜﺔ‬
‫ﺇﻟﻰ ﻀﺭﻭﺭﺓ ﺃﺨﺫ ﺨﻤﺱ ﻓﺠﻭﺍﺕ ﺯﻤﻨﻴﺔ‪.‬‬ ‫‪AIC‬‬ ‫ﻓﻲ ﺤﻴﻥ ﻴﺸﻴﺭ ﺍﻟﻤﺅﺸﺭ‬
‫‪ -4-3‬ﺍﺨﺘﺒﺎﺭ ﺍﻟﺴﺒﺒﻴﺔ‪:‬‬
‫ﻤﻊ ﺜﻼﺙ ﻓﺠﻭﺍﺕ ﺯﻤﻨﻴﺔ ﻋﻨﺩ ﻤﺴﺘﻭﻯ‬ ‫)‪D(GDP‬‬ ‫ﻴﺴﺒﺏ ﺍﻟﻤﺘﻐﻴ‪‬ﺭ‬ ‫)‪D(GFCF‬‬ ‫ﻴﻅﻬﺭ ﺍﻟﺠﺩﻭل )‪ (5‬ﺃﻥ ﺍﻟﻤﺘﻐﻴ‪‬ﺭ‬
‫ﻻ ﻴﺴﺒﺏ ﺍﻟﻤﺘﻐﻴ‪‬ﺭ )‪.D(GDP‬‬ ‫)‪D(GFCF‬‬ ‫ﺩﻻﻟﺔ ‪ ،%5‬ﺇﺫﹾ ﺇﻨﱠﻨﺎ ﻨﺭﻓﺽ ﻓﺭﻀﻴﺔ ﺍﻟﻌﺩﻡ ﺍﻟﺘﻲ ﺘﻘﻭل‪ :‬ﺇﻥ‪ ‬ﺍﻟﻤﺘﻐﻴ‪‬ﺭ‬
‫ﻋﻨﺩ ﻤﺴﺘﻭﻯ ﺩﻻﻟـﺔ ‪ ،%5‬ﺇﺫﹾ ﺇﻨﱠﻨـﺎ ﻻ ﻨـﺴﺘﻁﻴﻊ‬ ‫)‪D(GFCF‬‬ ‫ﻻ ﻴﺴﺒﺏ ﺍﻟﻤﺘﻐﻴ‪‬ﺭ‬ ‫)‪D(GDP‬‬ ‫ﻓﻲ ﺤﻴﻥ ﺍﻟﻤﺘﻐﻴ‪‬ﺭ‬
‫‪.‬‬ ‫)‪D(GFCF‬‬ ‫ﺭﻓﺽ ﻓﺭﻀﻴﺔ ﺍﻟﻌﺩﻡ ﺍﻟﺘﻲ ﺘﻘﻭل‪ :‬ﺇﻥ‪ ‬ﺍﻟﻤﺘﻐﻴ‪‬ﺭ )‪ D(GDP‬ﻻ ﻴﺴﺒﺏ ﺍﻟﻤﺘﻐﻴ‪‬ﺭ‬
‫‪Granger‬‬ ‫ﺍﻟﺠﺩﻭل )‪ (5‬ﺍﺨﺘﺒﺎﺭ ﺴﺒﺒﻴﺔ‬
‫‪Pairwise Granger Causality Tests‬‬
‫‪Date: 05/01/11 Time: 12:17‬‬
‫‪Sample: 1983 2009‬‬
‫‪Lags: 3‬‬

‫‪Null Hypothesis:‬‬ ‫‪Obs‬‬ ‫‪F-Statistic‬‬ ‫‪Prob.‬‬

‫)‪D(GFCF) does not Granger Cause D(GDP‬‬ ‫‪23‬‬ ‫‪3.25907‬‬ ‫‪0.0491‬‬

‫)‪D(GDP) does not Granger Cause D(GFCF‬‬ ‫‪0.28233‬‬ ‫‪0.8374‬‬

‫‪EViews 6‬‬ ‫ﺍﻟﻤﺼﺩﺭ‪:‬ﻨﺘﺎﺌﺞ ﺤﺴﺎﺒﺎﺕ ﺒﺭﻨﺎﻤﺞ‬

‫‪ -5-3‬ﺘﻘﺩﻴﺭ ﻨﻤﻭﺫﺝ ‪: VAR‬‬

‫ﺒﺎﻟﻨﻅﺭ ﺇﻟﻰ ﻨﺘﺎﺌﺞ ﺍﺨﺘﺒﺎﺭ ﺍﻟﺴﺒﺒﻴﺔ ﺠﺩﻭل )‪ ، (5‬ﻭﻗﻴﻡ ﻤﻌﺎﻴﻴﺭ ﺘﺤﺩﻴﺩ ﻋﺩﺩ ﻤﺩﺩ ﺍﻟﺘﺒﺎﻁﺅ ﺍﻟﺯﻤﻨﻲ ﺠـﺩﻭل )‪(4‬‬
‫ﻜﻤﺎ ﻴﺸﻴﺭ ﺇﻟﻰ ﺫﻟﻙ‬ ‫‪VAR‬‬ ‫ﻭ ﻤﻥ ﺃﺠل ﺍﻟﺘﻭﻓﻴﻕ ﻓﻴﻤﺎ ﺒﻴﻨﻬﻤﺎ ﻨﺨﺘﺎﺭ ﺨﻤﺱ ﻓﺠﻭﺍﺕ ﺯﻤﻨﻴﺔ ﻋﻨﺩ ﺘﻘﺩﻴﺭ ﻨﻤﻭﺫﺝ‬
‫‪:‬‬ ‫‪VAR‬‬ ‫‪ .‬ﻴﻌﻁﻲ ﺍﻟﺠﺩﻭل )‪ (6‬ﻨﺘﺎﺌﺞ ﺘﻘﺩﻴﺭ ﺍﻟﻨﻤﻭﺫﺝ‬ ‫‪AIC‬‬ ‫ﺍﻟﻤﻌﻴﺎﺭ‬
‫‪VAR‬‬ ‫ﺍﻟﺠﺩﻭل)‪ :(6‬ﺍﻟﻤﻌﺎﻤﻼﺕ ﺍﻟﻤﻘﺩﺭﺓ ﻟﻠﻨﻤﻭﺫﺝ‬
‫‪Vector Autoregression Estimates‬‬
‫‪Date: 05/09/11 Time: 06:42‬‬
‫‪Sample (adjusted): 1989 2009‬‬
‫‪Included observations: 21 after adjustments‬‬
‫] [ ‪Standard errors in ( ) & t-statistics in‬‬

‫)‪D(GDP‬‬ ‫)‪D(GFCF‬‬

‫))‪D(GDP(-1‬‬ ‫‪0.102214‬‬ ‫‪0.061234‬‬

‫)‪(0.22582‬‬ ‫)‪(0.14932‬‬

‫‪351‬‬
‫ﺍﺴﺘﺨﺩﺍﻡ ﻨﻤﺎﺫﺝ ‪ VAR‬ﻓﻲ ﺍﻟﺘﻨﺒﺅ ﻭﺩﺭﺍﺴﺔ ﺍﻟﻌﻼﻗﺔ ﺍﻟﺴﺒﺒﻴﺔ ﺒﻴﻥ ﺇﺠﻤﺎﻟﻲ ﺍﻟﻨﺎﺘﺞ ﺍﻟﻤﺤﻠﻲ ﻭﺇﺠﻤﺎﻟﻲ ﺍﻟﺘﻜﻭﻴﻥ ﺍﻟﺭﺃﺴﻤﺎﻟﻲ ﻓﻲ ﺴﻭﺭﻴﺔ‬

‫]‪[ 0.45265‬‬ ‫]‪[ 0.41009‬‬

‫))‪D(GDP(-2‬‬ ‫‪0.120189‬‬ ‫‪0.159979‬‬

‫)‪(0.23185‬‬ ‫)‪(0.15331‬‬
‫]‪[ 0.51839‬‬ ‫]‪[ 1.04351‬‬

‫))‪D(GDP(-3‬‬ ‫‪0.395350‬‬ ‫‪0.083647‬‬

‫)‪(0.22487‬‬ ‫)‪(0.14870‬‬
‫]‪[ 1.75811‬‬ ‫]‪[ 0.56254‬‬

‫))‪D(GDP(-4‬‬ ‫‪0.246807‬‬ ‫‪-0.030421‬‬

‫)‪(0.19939‬‬ ‫)‪(0.13185‬‬
‫]‪[ 1.23779‬‬ ‫]‪[-0.23073‬‬

‫))‪D(GDP(-5‬‬ ‫‪0.339129‬‬ ‫‪0.034068‬‬

‫)‪(0.20571‬‬ ‫)‪(0.13602‬‬
‫]‪[ 1.64860‬‬ ‫]‪[ 0.25045‬‬

‫))‪D(GFCF(-1‬‬ ‫‪0.039536‬‬ ‫‪0.215690‬‬

‫)‪(0.49315‬‬ ‫)‪(0.32609‬‬
‫]‪[ 0.08017‬‬ ‫]‪[ 0.66143‬‬

‫))‪D(GFCF(-2‬‬ ‫‪0.477633‬‬ ‫‪-0.042968‬‬

‫)‪(0.42584‬‬ ‫)‪(0.28159‬‬
‫]‪[ 1.12162‬‬ ‫]‪[-0.15259‬‬

‫))‪D(GFCF(-3‬‬ ‫‪0.416738‬‬ ‫‪-0.114465‬‬

‫)‪(0.51093‬‬ ‫)‪(0.33785‬‬
‫]‪[ 0.81565‬‬ ‫]‪[-0.33881‬‬

‫))‪D(GFCF(-4‬‬ ‫‪0.049016‬‬ ‫‪-0.103498‬‬

‫)‪(0.51655‬‬ ‫)‪(0.34156‬‬
‫]‪[ 0.09489‬‬ ‫]‪[-0.30301‬‬

‫))‪D(GFCF(-5‬‬ ‫‪-1.373851‬‬ ‫‪-0.383464‬‬

‫)‪(0.52134‬‬ ‫)‪(0.34473‬‬
‫]‪[-2.63524‬‬ ‫]‪[-1.11235‬‬

‫‪R-squared‬‬ ‫‪0.576775‬‬ ‫‪0.193337‬‬

‫‪Adj. R-squared‬‬ ‫‪0.230500‬‬ ‫‪-0.466659‬‬

‫‪352‬‬
‫ﻋﺜﻤﺎﻥ ﻨﻘﺎﺭ – ﻤﻨﺫﺭ ﺍﻟﻌﻭﺍﺩ‬ 2012- ‫ ﺍﻟﻌﺩﺩ ﺍﻟﺜﺎﻨﻲ‬- 28 ‫ﻤﺠﻠﺔ ﺠﺎﻤﻌﺔ ﺩﻤﺸﻕ ﻟﻠﻌﻠﻭﻡ ﺍﻻﻗﺘﺼﺎﺩﻴﺔ ﻭﺍﻟﻘﺎﻨﻭﻨﻴﺔ – ﺍﻟﻤﺠﻠﺩ‬

Sum sq. resids 1.23E+10 5.39E+09

S.E. equation 33472.63 22133.60
F-statistic 1.665657 0.292937
Log likelihood -241.7963 -233.1100
Akaike AIC 23.98060 23.15334
Schwarz SC 24.47799 23.65073
Mean dependent 40763.69 9505.405
S.D. dependent 38158.00 18276.27

Determinant resid covariance (dof adj.) 5.38E+17

Determinant resid covariance 1.48E+17
Log likelihood -474.6946
Akaike information criterion 47.11377
Schwarz criterion 48.10855

EViews 6 ‫ﻨﺘﺎﺌﺞ ﺤﺴﺎﺒﺎﺕ ﺒﺭﻨﺎﻤﺞ‬:‫ﺍﻟﻤﺼﺩﺭ‬

:‫ﺍﻟﻤﻘﺩﺭ ﻓﻬﻭ‬ VAR(5) ‫ﺎ ﻨﻤﻭﺫﺝ‬‫ﺃﻤ‬

D(GDP) = 0.102214435006*D(GDP(-1)) + 0.120188644544*D(GDP(-2)) +
0.395350280292*D(GDP(-3)) + 0.246807044358*D(GDP(-4)) + 0.339128590429*D(GDP(-5)) +
0.039536102212*D(GFCF(-1)) + 0.477633382176*D(GFCF(-2)) + 0.41673827985*D(GFCF(-3)) +
0.0490161964479*D(GFCF(-4)) - 1.37385079754*D(GFCF(-5))
D(GFCF) = 0.0612337559439*D(GDP(-1)) + 0.159979309989*D(GDP(-2)) +
0.0836474864066*D(GDP(-3)) - 0.0304206982274*D(GDP(-4)) + 0.0340676203833*D(GDP(-5)) +
0.215690201792*D(GFCF(-1)) - 0.042968175574*D(GFCF(-2)) - 0.114465407213*D(GFCF(-3)) -
0.103498269684*D(GFCF(-4)) - 0.383463505351*D(GFCF(-5))
: ‫ﺃﻭ ﻴﻤﻜﻥ ﻜﺘﺎﺒﺘﻪ ﺒﺎﻟﺸﻜل ﺍﻟﻤﺼﻔﻭﻓﻲ ﺍﻵﺘﻲ‬
 D( gdp) t  0.1022 0.0395  D( gdp) t −1  0.1202 0.4776   D( gdp) t − 2 
 D( gfcf )  = 0.0612 0.2157. D( gfcf )  + 0.1599 - 0.0429. D( gfcf )  +
 t   t −1    t −2 

0.3954 0.4167   D( gdp)t − 3   0.2468 0.0490   D( gdp)t − 4 

0.0836 - 0.1145. D( gfcf )  + - 0.0304 - 0.1035. D( gfcf )  +
  t −3    t −4 

0.3391 - 1.374   D( gdp)t − 5 

0.0341 - 0.3835. D( gfcf ) 
  t −5 

:‫ ﺍﺨﺘﺒﺎﺭﺍﺕ ﺍﻟﺒﻭﺍﻗﻲ‬-6-3
‫ﻤﻥ ﺃﺠل ﺍﻟﺘﺤﻘﻕ ﻤﻥ ﺼﺤﺔ ﺍﻟﻨﻤﻭﺫﺝ ﺍﻟﻤﻘﺩﺭ ﻴﺠﺏ ﺃﻥ ﻨﺘﺄﻜﺩ ﻤﻥ ﺨﻀﻭﻉ ﺍﻟﺒﻭﺍﻗﻲ ﻟﻠﺘﻭﺯﻴﻊ ﺍﻟﻁﺒﻴﻌﻲ ﻭﺃﻨﻬـﺎ‬
.‫ﻏﻴﺭ ﻤﺭﺘﺒﻁﺔ ﺫﺍﺘﻴﺎﹰ‬

353
 ‫ ﻓﻲ ﺍﻟﺘﻨﺒﺅ ﻭﺩﺭﺍﺴﺔ ﺍﻟﻌﻼﻗﺔ ﺍﻟﺴﺒﺒﻴﺔ ﺒﻴﻥ ﺇﺠﻤﺎﻟﻲ ﺍﻟﻨﺎﺘﺞ ﺍﻟﻤﺤﻠﻲ ﻭﺇﺠﻤﺎﻟﻲ ﺍﻟﺘﻜﻭﻴﻥ ﺍﻟﺭﺃﺴﻤﺎﻟﻲ ﻓﻲ ﺴﻭﺭﻴﺔ‬VAR ‫ﺍﺴﺘﺨﺩﺍﻡ ﻨﻤﺎﺫﺝ‬

(7) ‫ ﻜﻤﺎ ﻴﻅﻬﺭ ﻓﻲ ﺍﻟﺠﺩﻭل‬14Jarque-Bera ‫ ﻨﺴﺘﺨﺩﻡ ﺍﺨﺘﺒﺎﺭ‬:‫ ﺍﻟﺘﻭﺯﻴﻊ ﺍﻻﺤﺘﻤﺎﻟﻲ ﻟﻠﺒﻭﺍﻗﻲ‬-1

‫ ﺍﺨﺘﺒﺎﺭ ﺍﻟﺘﻭﺯﻴﻊ ﺍﻟﻁﺒﻴﻌﻲ ﻟﻠﺒﻭﺍﻗﻲ‬:(7)‫ﺍﻟﺠﺩﻭل‬
VAR Residual Normality Tests
Orthogonalization: Cholesky (Lutkepohl)
Null Hypothesis: residuals are multivariate normal
Date: 05/09/11 Time: 23:53
Sample: 1983 2009
Included observations: 21

Component Jarque-Bera df Prob.

1 5.308622 2 0.0703
2 4.573779 2 0.1016

Joint 9.882402 4 0.0425

EViews 6 ‫ﻨﺘﺎﺌﺞ ﺤﺴﺎﺒﺎﺕ ﺒﺭﻨﺎﻤﺞ‬:‫ﺍﻟﻤﺼﺩﺭ‬

‫ﻴﺸﻴﺭ ﺍﻻﺨﺘﺒﺎﺭ ﺇﻟﻰ ﻋﺩﻡ ﺭﻓﺽ ﻓﺭﻀﻴﺔ ﺍﻟﻌﺩﻡ ﻟﻜل ﻤﻥ ﺍﻟﺒﻭﺍﻗﻲ ﺍﻟﺨﺎﺼﺔ ﺒﺎﻟﻤﻌﺎﺩﻟـﺔ ﺍﻷﻭﻟـﻰ ﻭ ﺍﻟﻤﻌﺎﺩﻟـﺔ‬
.%5 ‫ ﺃﻱ ﻋﺩﻡ ﺭﻓﺽ ﻓﺭﻀﻴﺔ ﺍﻟﺘﻭﺯﻴﻊ ﺍﻟﻁﺒﻴﻌﻲ ﻟﻠﺒﻭﺍﻗﻲ ﻋﻨﺩ ﻤﺴﺘﻭﻯ ﺩﻻﻟﺔ‬.‫ﺍﻟﺜﺎﻨﻴﺔ‬
.(8) ‫ﻥ ﺍﻟﺠﺩﻭل‬‫ ﻜﻤﺎ ﻴﺒﻴ‬،15Ljung-Box ‫ ﻨﺴﺘﺨﺩﻡ ﺍﺨﺘﺒﺎﺭ‬:‫ ﺍﺨﺘﺒﺎﺭ ﺍﻻﺭﺘﺒﺎﻁ ﺍﻟﺫﺍﺘﻲ ﻟﻠﺒﻭﺍﻗﻲ‬-2
‫ ﺍﺨﺘﺒﺎﺭ ﺍﻻﺭﺘﺒﺎﻁ ﺍﻟﺫﺍﺘﻲ ﻟﻠﺒﻭﺍﻗﻲ‬:(8)‫ﺍﻟﺠﺩﻭل‬
VAR Residual Portmanteau Tests for Autocorrelations
Null Hypothesis: no residual autocorrelations up to lag h
Date: 05/09/11 Time: 23:50
Sample: 1983 2009
Included observations: 21

Lags Q-Stat Prob. Adj Q-Stat Prob. df

1 0.063567 NA* 0.066745 NA* NA*

2 0.818964 NA* 0.901658 NA* NA*

14
BERA, A.K. and Jarque .C.M.(1981), "An efficient large Sample test for normality of observations
and regression residuals ", Working paper in Econometrics No 40,Australion National university,
Canberra.
15
Ljung, G.M., and Box G.E.P. (1978) "on a measure of the lack of fit in time Series models".
Biometrika, n65:PP.297-303

354
‫ﻋﺜﻤﺎﻥ ﻨﻘﺎﺭ – ﻤﻨﺫﺭ ﺍﻟﻌﻭﺍﺩ‬ ‫ﻤﺠﻠﺔ ﺠﺎﻤﻌﺔ ﺩﻤﺸﻕ ﻟﻠﻌﻠﻭﻡ ﺍﻻﻗﺘﺼﺎﺩﻴﺔ ﻭﺍﻟﻘﺎﻨﻭﻨﻴﺔ – ﺍﻟﻤﺠﻠﺩ ‪ - 28‬ﺍﻟﻌﺩﺩ ﺍﻟﺜﺎﻨﻲ ‪2012-‬‬

‫‪3‬‬ ‫‪1.724568‬‬ ‫*‪NA‬‬ ‫‪1.958196‬‬ ‫*‪NA‬‬ ‫*‪NA‬‬

‫‪4‬‬ ‫‪3.179978‬‬ ‫*‪NA‬‬ ‫‪3.756055‬‬ ‫*‪NA‬‬ ‫*‪NA‬‬
‫‪5‬‬ ‫‪6.353596‬‬ ‫*‪NA‬‬ ‫‪7.921429‬‬ ‫*‪NA‬‬ ‫*‪NA‬‬
‫‪6‬‬ ‫‪8.376506‬‬ ‫‪0.0787‬‬ ‫‪10.75350‬‬ ‫‪0.0952‬‬ ‫‪4‬‬
‫‪7‬‬ ‫‪9.362148‬‬ ‫‪0.3127‬‬ ‫‪12.23197‬‬ ‫‪0.1412‬‬ ‫‪8‬‬
‫‪8‬‬ ‫‪12.45684‬‬ ‫‪0.4097‬‬ ‫‪17.23108‬‬ ‫‪0.1411‬‬ ‫‪12‬‬
‫‪9‬‬ ‫‪19.12419‬‬ ‫‪0.2623‬‬ ‫‪28.89894‬‬ ‫‪0.0246‬‬ ‫‪16‬‬
‫‪10‬‬ ‫‪20.05877‬‬ ‫‪0.4543‬‬ ‫‪30.68315‬‬ ‫‪0.0595‬‬ ‫‪20‬‬
‫‪11‬‬ ‫‪22.17966‬‬ ‫‪0.5685‬‬ ‫‪35.13701‬‬ ‫‪0.0664‬‬ ‫‪24‬‬
‫‪12‬‬ ‫‪22.38578‬‬ ‫‪0.7631‬‬ ‫‪35.61796‬‬ ‫‪0.1526‬‬ ‫‪28‬‬

‫‪*The test is valid only for lags larger than the VAR lag order.‬‬
‫‪df is degrees of freedom for (approximate) chi-square distribution‬‬

‫ﺍﻟﻤﺼﺩﺭ‪:‬ﻨﺘﺎﺌﺞ ﺤﺴﺎﺒﺎﺕ ﺒﺭﻨﺎﻤﺞ ‪EViews 6‬‬

‫ﻴﺸﻴﺭ ﺍﺨﺘﺒﺎﺭ ﺍﻻﺭﺘﺒﺎﻁ ﺍﻟﺫﺍﺘﻲ ﺇﻟﻰ ﻋﺩﻡ ﺭﻓﺽ ﻓﺭﻀﻴﺔ ﺍﻟﻌﺩﻡ ﺍﻟﺘﻲ ﺘﻌﻨﻲ ﻋﺩﻡ ﻭﺠﻭﺩ ﺍﺭﺘﺒـﺎﻁ ﺫﺍﺘـﻲ ﻋﻨـﺩ‬
‫ﻤﺴﺘﻭﻯ ﺩﻻﻟﺔ ‪.%5‬‬

‫‪ -7-3‬ﺍﻟﺘﻨﺒﺅ‪:‬‬
‫ﺒﺎﺴﺘﺨﺩﺍﻡ ﺍﻟﻨﻤﻭﺫﺝ ﺍﻟﻤﻘﺩﺭ ﻨﻘﻭﻡ ﺒﺎﻟﺘﻨﺒﺅ ﺤﺘﻰ ﻋﺎﻡ ‪ 2015‬ﻜﻤﺎ ﻴﻅﻬﺭ ﻓﻲ ﺍﻟﺠﺩﻭل )‪.(9‬‬

‫ﺤﺘﻰ ﻋﺎﻡ ‪ 2015‬ﻤﻊ ﺤﺩﻱ ﺍﻟﺜﻘﺔ ﺒﺩﺭﺠﺔ ‪%95‬‬ ‫‪GFCF‬‬ ‫ﻭ‬ ‫‪GDP‬‬ ‫ﺍﻟﺠﺩﻭل )‪:(9‬ﺍﻟﺘﻨﺒﺅﺍﺕ ﺒﻘﻴﻡ‬
‫‪16‬‬ ‫‪17‬‬
‫‪Year‬‬ ‫‪LGDPf‬‬ ‫‪GDPf‬‬ ‫‪UGDPf‬‬ ‫‪LGFCFf‬‬ ‫‪GFCFf‬‬ ‫‪UGFCFf‬‬
‫‪2010‬‬ ‫‪1374082‬‬ ‫‪1441028‬‬ ‫‪1507973‬‬ ‫‪288955.8‬‬ ‫‪313223‬‬ ‫‪337490.2‬‬
‫‪2011‬‬ ‫‪1424193‬‬ ‫‪1491139‬‬ ‫‪1558084‬‬ ‫‪304571.6‬‬ ‫‪330838.8‬‬ ‫‪355106‬‬
‫‪2012‬‬ ‫‪1556511‬‬ ‫‪1623456‬‬ ‫‪1690401‬‬ ‫‪331130.4‬‬ ‫‪355397.6‬‬ ‫‪379664.8‬‬
‫‪2013‬‬ ‫‪1663338‬‬ ‫‪1730283‬‬ ‫‪1797228‬‬ ‫‪354226.6‬‬ ‫‪378493.8‬‬ ‫‪402761‬‬
‫‪2014‬‬ ‫‪1720697‬‬ ‫‪1787642‬‬ ‫‪1854587‬‬ ‫‪376805‬‬ ‫‪401072.2‬‬ ‫‪425339.4‬‬
‫‪2015‬‬ ‫‪1811342‬‬ ‫‪1878288‬‬ ‫‪1945233‬‬ ‫‪399653.7‬‬ ‫‪424920.9‬‬ ‫‪449188.1‬‬

‫‪EViews 6‬‬ ‫ﺍﻟﻤﺼﺩﺭ ‪ :‬ﻨﺘﺎﺌﺞ ﺤﺴﺎﺒﺎﺕ ﺒﺭﻨﺎﻤﺞ‬

‫‪ 16‬اﻟﺤﺪ اﻷدﻧﻰ ﻟﻠﺘﻮﻗﻊ ﺑﺎﺣﺘﻤﺎل ‪. %95‬‬

‫‪ 17‬اﻟﺤﺪ اﻷﻋﻠﻰ ﻟﻠﺘﻮﻗﻊ ﺑﺎﺣﺘﻤﺎل ﺛﻘﺔ ‪. %95‬‬

‫‪355‬‬
 ‫ﺍﺴﺘﺨﺩﺍﻡ ﻨﻤﺎﺫﺝ ‪ VAR‬ﻓﻲ ﺍﻟﺘﻨﺒﺅ ﻭﺩﺭﺍﺴﺔ ﺍﻟﻌﻼﻗﺔ ﺍﻟﺴﺒﺒﻴﺔ ﺒﻴﻥ ﺇﺠﻤﺎﻟﻲ ﺍﻟﻨﺎﺘﺞ ﺍﻟﻤﺤﻠﻲ ﻭﺇﺠﻤﺎﻟﻲ ﺍﻟﺘﻜﻭﻴﻥ ﺍﻟﺭﺃﺴﻤﺎﻟﻲ ﻓﻲ ﺴﻭﺭﻴﺔ‬

‫ﻨﺒﻴﻥ ﻓﻲ ﺍﻟﺸﻜﻠﻴﻥ )‪ (5‬ﻭ )‪ (6‬ﺍﻟﻘﻴﻡ ﺍﻟﻔﻌﻠﻴﺔ ﻭﺍﻟﻘﻴﻡ ﺍﻟﻤﻘﺩﺭﺓ )ﺍﻟﻨﻅﺭﻴﺔ ( ﻭﺍﻟﻘﻴﻡ ﺍﻟﻤﺘﻨﺒﺄ ﺒﻬﺎ ﻟﻜل ﻤﻥ ﺇﺠﻤﺎﻟﻲ‬
‫ﺍﻟﻨﺎﺘﺞ ﺍﻟﻤﺤﻠﻲ ﻭﺇﺠﻤﺎﻟﻲ ﺘﻜﻭﻴﻥ ﺭﺃﺱ ﺍﻟﻤﺎل ﺍﻟﺜﺎﺒﺕ‪.‬‬
‫‪2,000,000‬‬

‫‪1,800,000‬‬

‫‪1,600,000‬‬

‫‪1,400,000‬‬

‫‪1,200,000‬‬

‫‪1,000,000‬‬

‫‪800,000‬‬

‫‪600,000‬‬

‫‪400,000‬‬
‫‪1985‬‬ ‫‪1990‬‬ ‫‪1995‬‬ ‫‪2000‬‬ ‫‪2005‬‬ ‫‪2010‬‬ ‫‪2015‬‬

‫‪GDP‬‬ ‫‪GDPF‬‬
‫‪18‬‬
‫ﺍﻟﺸﻜل )‪ :(5‬ﺍﻟﻘﻴﻡ ﺍﻟﻔﻌﻠﻴﺔ ﻹﺠﻤﺎﻟﻲ ﺍﻟﻨﺎﺘﺞ ﺍﻟﻤﺤﻠﻲ ﻭ ﺍﻟﻘﻴﻡ ﺍﻟﻨﻅﺭﻴﺔ ﻭﺍﻟﻤﺘﻨﺒﺄ ﺒﻬﺎ‬
‫‪440,000‬‬

‫‪400,000‬‬

‫‪360,000‬‬

‫‪320,000‬‬

‫‪280,000‬‬

‫‪240,000‬‬

‫‪200,000‬‬

‫‪160,000‬‬

‫‪120,000‬‬

‫‪80,000‬‬
‫‪1985‬‬ ‫‪1990‬‬ ‫‪1995‬‬ ‫‪2000‬‬ ‫‪2005‬‬ ‫‪2010‬‬ ‫‪2015‬‬

‫‪GFCF‬‬ ‫‪GFCFF‬‬

‫ﺍﻟﺸﻜل )‪ :(6‬ﺍﻟﻘﻴﻡ ﺍﻟﻔﻌﻠﻴﺔ ﻹﺠﻤﺎﻟﻲ ﺘﻜﻭﻴﻥ ﺭﺃﺱ ﺍﻟﻤﺎل ﺍﻟﺜﺎﺒﺕ ﻭﺍﻟﻘﻴﻡ ﺍﻟﻨﻅﺭﻴﺔ ﻭﺍﻟﻤﺘﻨﺒﺄ ﺒﻬﺎ‬

‫‪ 18‬اﻟﻘﯿﻢ اﻟﻮاﻗﻌﺔ ﻋﻠﻰ ﻣﻨﺤﻨﻰ ‪ GDP‬ھﻲ اﻟﻔﻌﻠﯿﺔ ‪،‬واﻟﻮاﻗﻌﺔ ﻋﻠﻰ ﻣﻨﺤﻨﻰ ‪ GDPF‬ﺣﺘﻰ ﻋ ﺎم ‪ 2009‬ﻧﻈﺮﯾ ﺔ وﺑﻌ ﺪ‬
‫ﻋﺎم ‪ 2009‬اﻟﻤﺘﻨﺒﺄ ﺑﮭﺎ ‪.‬‬

‫‪356‬‬
‫ﻋﺜﻤﺎﻥ ﻨﻘﺎﺭ – ﻤﻨﺫﺭ ﺍﻟﻌﻭﺍﺩ‬ ‫ﻤﺠﻠﺔ ﺠﺎﻤﻌﺔ ﺩﻤﺸﻕ ﻟﻠﻌﻠﻭﻡ ﺍﻻﻗﺘﺼﺎﺩﻴﺔ ﻭﺍﻟﻘﺎﻨﻭﻨﻴﺔ – ﺍﻟﻤﺠﻠﺩ ‪ - 28‬ﺍﻟﻌﺩﺩ ﺍﻟﺜﺎﻨﻲ ‪2012-‬‬

‫‪ -8-3‬ﺍﻟﻤﻘﺎﺭﻨﺔ ﺒﻴﻥ ﺍﺴﺘﺨﺩﺍﻡ ﻤﻌﺎﺩﻟﺔ ﺍﻻﺘﺠﺎﻩ ﺍﻟﻌﺎﻡ ﻭﻨﻤﺎﺫﺝ ‪: VAR‬‬

‫ﺇﻥ ﺍﻟﻨﻤﺎﺫﺝ ﺍﻷﻜﺜﺭ ﺍﺴﺘﺨﺩﺍﻤﺎﹰ ﻓﻲ ﺍﻟﺘﻁﺒﻴﻘﺎﺕ ﺍﻟﻌﻤﻠﻴﺔ ﻟﻠﺘﻨﺒﺅ‪ ،‬ﻫﻲ ﻨﻤﺎﺫﺝ ﺍﻻﺘﺠﺎﻩ ﺍﻟﻌﺎﻡ ﺍﻟﺘﻲ ﺘﺩﺭﺱ ﺘﻁـﻭﺭ‬
‫ﻅﺎﻫﺭﺓ ﻤﺎ‪ ،‬ﺒﺩﺍﺭﺴﺔ ﻋﻼﻗﺘﻬﺎ ﺒﺎﻟﺯﻤﻥ ﺒﻭﺼﻔﻪ ﻴﻠﺨﺹ ﺍﻟﻌﻭﺍﻤل ﺍﻟﻤﺅﺜﺭﺓ ﺒﺎﻟﻅـﺎﻫﺭﺓ ﻤﻭﻀـﻭﻉ ﺍﻟﺩﺭﺍﺴـﺔ‬
‫ﺠﻤﻴﻌﻬﺎ‪ .‬ﻟﻘﺩ ﻗﻤﻨﺎ ﺒﺈﻴﺠﺎﺩ ﻤﻌﺎﺩﻟﺔ ﺍﻻﺘﺠﺎﻩ ﺍﻟﻌﺎﻡ ﻟﺘﻁﻭﺭ ﺇﺠﻤـﺎﻟﻲ ﺍﻟﻨـﺎﺘﺞ ﺍﻟﻤﺤﻠـﻲ ﻟﻠﺴﻠـﺴﻠﺔ ﻤﻭﻀـﻭﻉ‬
‫ﺍﻟﺩﺭﺍﺴﺔ‪ ،19‬ﻭﻗﻤﻨﺎ ﺒﺎﻟﺘﻨﺒﺅ ﺒﻬﺎ ﺒﻘﻴﻡ ﺇﺠﻤﺎﻟﻲ ﺍﻟﻨﺎﺘﺞ ﺍﻟﻤﺤﻠﻲ ﺤﺘﻰ ﻋﺎﻡ ‪ 2015‬ﻜﻤﺎ ﻓﻲ ﺍﻟـﺸﻜل ﺭﻗـﻡ )‪،(7‬‬
‫ﻭﺍﻟﺠﺩﻭل )‪ .(10‬ﻤﻥ ﻤﻘﺎﺭﻨﺔ ﺍﻟﻨﺘﺎﺌﺞ ﺍﻟﺘﻲ ﺃﻋﻁﺘﻬﺎ ﻤﻌﺎﺩﻟﺔ ﺍﻻﺘﺠﺎﻩ ﺍﻟﻌﺎﻡ ﻤﻊ ﺍﻟﻨﺘﺎﺌﺞ ﺍﻟﺘـﻲ ﺘـﻡ ﺇﻴﺠﺎﺩﻫـﺎ‬
‫ﻫﻲ ﺃﻋﻠﻰ ﻤﻥ ﺍﻟﻘﻴﻡ ﺍﻟﻤﺘﻨﺒـﺄ‬ ‫‪VAR‬‬ ‫ﺍﻋﺘﻤﺎﺩﺍﹰ ﻋﻠﻰ ﻨﻤﺎﺫﺝ‪ ،VAR‬ﻨﺭﻯ ﺃﻥ ﺍﻟﻘﻴﻡ ﺍﻟﻤﺘﻨﺒﺄ ﺒﻬﺎ ﺒﺎﺴﺘﺨﺩﺍﻡ ﻨﻤﺎﺫﺝ‬
‫ﺒﻬﺎ ﺒﺎﺴﺘﺨﺩﺍﻡ ﻤﻌﺎﺩﻟﺔ ﺍﻻﺘﺠﺎﻩ ﺍﻟﻌﺎﻡ‪ ،‬ﻭﻫﺫﺍ ﻫﻭ ﺍﻷﻜﺜﺭ ﻤﻨﻁﻘﻴﺔ‪ .‬ﻓﻤﻥ ﺍﻟﻤﻌﻠﻭﻡ ﺃﻥ‪ ‬ﺃﺤﺩ ﺍﻟﻌﻭﺍﻤل ﺍﻟﻤﺅﺩﻴﺔ ﺇﻟﻰ‬
‫ﺯﻴﺎﺩﺓ ﺍﻟﻨﺎﺘﺞ ﺍﻟﻤﺤﻠﻲ ﻫﻭ ﺍﻟﻘﻴﺎﻡ ﺒﺎﺴﺘﺜﻤﺎﺭﺍﺕ ﺠﺩﻴﺩﺓ ‪،‬ﻭﻫﺫﻩ ﺍﻻﺴﺘﺜﻤﺎﺭﺍﺕ ﺍﻟﺠﺩﻴﺩﺓ ﺍﻟﺘـﻲ ﺃﺩﺕ ﺇﻟـﻰ ﻫـﺫﺍ‬
‫ﺍﻻﺭﺘﻔﺎﻉ ﻓﻲ ﺍﻟﻘﻴﻡ ﺍﻟﻤﺘﻨﺒﺄ ﺒﻬﺎ ﺒﺎﺴﺘﺨﺩﺍﻡ ﻨﻤﺎﺫﺝ ‪ ،VAR‬ﻫﻲ ﺍﺴﺘﺜﻤﺎﺭﺍﺕ ﺍﻟﺴﻨﻭﺍﺕ ﺍﻟـﺴﺎﺒﻘﺔ ﺍﻟﺘـﻲ ﺘـﺄﺨﺭ‬
‫ﻋﻠـﻰ‬ ‫‪VAR‬‬ ‫ﺘﺄﺜﻴﺭﻫﺎ ﻓﻲ ﺍﻟﻨﺎﺘﺞ ﻭﻟﻡ ﻨﺄﺨﺫﻫﺎ ﺒﺎﻟﺤﺴﺒﺎﻥ ﻤﻌﺎﺩﻟﺔ ﺍﻻﺘﺠﺎﻩ ﺍﻟﻌﺎﻡ ‪.‬ﻤﻥ ﻫﻨﺎ ﻨﺭﻯ ﺃﻓﻀﻠﻴﺔ ﻨﻤﺎﺫﺝ‬
‫ﻤﻌﺎﺩﻟﺔ ﺍﻻﺘﺠﺎﻩ ﺍﻟﻌﺎﻡ ﻓﻲ ﻤﺜل ﻫﺫﻩ ﺍﻟﺤﺎﻻﺕ ‪.‬‬

‫ﺍﻟﺸﻜل )‪ :(7‬ﺍﻟﻘﻴﻡ ﺍﻟﻔﻌﻠﻴﺔ ﻹﺠﻤﺎﻟﻲ ﺘﻜﻭﻴﻥ ﺭﺃﺱ ﺍﻟﻤﺎل ﺍﻟﺜﺎﺒﺕ ﻭ ﺍﻟﻘﻴﻡ ﺍﻟﻨﻅﺭﻴﺔ ﻭ ﺍﻟﻤﺘﻨﺒﺄ ﺒﻬﺎ‬

‫ﺍﺨﺘﻴﺭﺕ ﻤﻌﺎﺩﻟﺔ ﺍﻟﻤﺴﺘﻘﻴﻡ ﻟﻠﻤﻘﺎﺭﻨﺔ ﻷﻨﻬﺎ ﺍﻷﻜﺜﺭ ﺍﺴﺘﺨﺩﺍﻤﺎ ‪.‬ﺤﺘﻰ ﻓﻲ ﺤﺎل ﺍﺴﺘﺨﺩﺍﻡ ﻤﻌﺎﺩﻟﺔ ﺃﻜﺜﺭ ﺘﻤﺜﻴﻼ‪ ،‬ﻤﻌﺎﺩﻟﺔ ﻤـﻥ ﺩﺭﺠـﺔ ﺍﻟﺜﺎﻨﻴـﺔ‬ ‫‪19‬‬

‫ﻤﺜﻼﹰ ﺤﺼﻠﻨﺎ ﻋﻠﻰ ﻨﺘﺎﺌﺞ ﻤﺸﺎﺒﻬﺔ‪ ،‬ﻭﺫﻟﻙ ﻷﻨﻪ ﻋﻨﺩ ﺍﺴﺘﺨﺩﺍﻡ ﻫﺫﻩ ﺍﻟﻤﻌﺎﺩﻟﺔ ﻓﻲ ﺍﻟﺘﻨﺒﺅ ﻻ ﺘﺄﺨﺫ ﺘﺄﺜﻴﺭ ﺍﻟﻘﻴﻡ ﺍﻟﻤﺎﻀﻴﺔ ﻟﻠﻤﺘﻐﻴ‪‬ﺭ ﺍﻟﻤﺴﺘﻘل ‪.‬‬

‫‪357‬‬
 ‫ﺍﺴﺘﺨﺩﺍﻡ ﻨﻤﺎﺫﺝ ‪ VAR‬ﻓﻲ ﺍﻟﺘﻨﺒﺅ ﻭﺩﺭﺍﺴﺔ ﺍﻟﻌﻼﻗﺔ ﺍﻟﺴﺒﺒﻴﺔ ﺒﻴﻥ ﺇﺠﻤﺎﻟﻲ ﺍﻟﻨﺎﺘﺞ ﺍﻟﻤﺤﻠﻲ ﻭﺇﺠﻤﺎﻟﻲ ﺍﻟﺘﻜﻭﻴﻥ ﺍﻟﺭﺃﺴﻤﺎﻟﻲ ﻓﻲ ﺴﻭﺭﻴﺔ‬

‫‪VAR‬‬ ‫ﻭﻤﺎ ﻗﻠﻨﺎﻩ ﻜﻠﹼﻪ ﺒﺎﻟﻨﺴﺒﺔ ﺇﻟﻰ ﻤﻘﺎﺭﻨﺔ ﻨﺘﺎﺌﺞ ﺍﻟﺘﻨﺒﺅ ﺒﺎﺴﺘﺨﺩﺍﻡ ﻤﻌﺎﺩﻟﺔ ﺍﻻﺘﺠﺎﻩ ﺍﻟﻌﺎﻡ ﻤﻊ ﻨﺘﺎﺌﺞ ﻨﻤﺎﺫﺝ‬
‫ﻟﻠﺘﻨﺒﺅ ﺒﺈﺠﻤﺎﻟﻲ ﺍﻟﻨﺎﺘﺞ ﺍﻟﻤﺤﻠﻲ ‪،‬ﻴﻨﺩﺭﺝ ﻋﻠﻰ ﻨﺘﺎﺌﺞ ﺍﻟﺘﻨﺒﺅ ﺒﻘﻴﻡ ﺇﺠﻤﺎﻟﻲ ﺍﻟﺘﻜﻭﻴﻥ ﺍﻟﺭﺃﺴﻤﺎﻟﻲ ﺍﻟﺜﺎﺒﺕ‪ ،‬ﺍﻟﺸﻜل‬
‫)‪ (8‬ﻭﺍﻟﺠﺩﻭل )‪.(10‬‬

‫ﺍﻟﺸﻜل )‪ :(8‬ﺍﻟﻘﻴﻡ ﺍﻟﻔﻌﻠﻴﺔ ﻹﺠﻤﺎﻟﻲ ﺘﻜﻭﻴﻥ ﺭﺃﺱ ﺍﻟﻤﺎل ﺍﻟﺜﺎﺒﺕ ﻭ ﺍﻟﻘﻴﻡ ﺍﻟﻨﻅﺭﻴﺔ ﻭ ﺍﻟﻤﺘﻨﺒﺄ ﺒﻬﺎ‬

‫ﺤﺘﻰ ﻋﺎﻡ ‪2015‬‬ ‫‪GFCF‬‬ ‫ﻭ‬ ‫‪GDP‬‬ ‫ﺍﻟﺠﺩﻭل )‪:(10‬ﺍﻟﺘﻨﺒﺅﺍﺕ ﺒﻘﻴﻡ‬

‫‪GDP‬‬ ‫‪GFCF‬‬
‫‪Year‬‬ ‫‪VAR‬‬ ‫‪LINER‬‬ ‫‪VAR‬‬ ‫‪LINER‬‬
‫‪2010‬‬ ‫‪1441028‬‬ ‫‪1329774‬‬ ‫‪313223‬‬ ‫‪273236.2‬‬
‫‪2011‬‬ ‫‪1491139‬‬ ‫‪1365466‬‬ ‫‪330838.8‬‬ ‫‪280037.6‬‬
‫‪2012‬‬ ‫‪1623456‬‬ ‫‪1401158‬‬ ‫‪355397.6‬‬ ‫‪286839‬‬
‫‪2013‬‬ ‫‪1730283‬‬ ‫‪1436850‬‬ ‫‪378493.8‬‬ ‫‪293640.4‬‬
‫‪2014‬‬ ‫‪1787642‬‬ ‫‪1472542‬‬ ‫‪401072.2‬‬ ‫‪300441.8‬‬
‫‪2015‬‬ ‫‪1878288‬‬ ‫‪1508234‬‬ ‫‪424920.9‬‬ ‫‪307243.2‬‬

‫‪358‬‬
‫ﻋﺜﻤﺎﻥ ﻨﻘﺎﺭ – ﻤﻨﺫﺭ ﺍﻟﻌﻭﺍﺩ‬ ‫ﻤﺠﻠﺔ ﺠﺎﻤﻌﺔ ﺩﻤﺸﻕ ﻟﻠﻌﻠﻭﻡ ﺍﻻﻗﺘﺼﺎﺩﻴﺔ ﻭﺍﻟﻘﺎﻨﻭﻨﻴﺔ – ﺍﻟﻤﺠﻠﺩ ‪ - 28‬ﺍﻟﻌﺩﺩ ﺍﻟﺜﺎﻨﻲ ‪2012-‬‬

‫ﺭﺍﺒﻌﺎﹰ‪ -‬ﺍﻟﻨﺘﺎﺌﺞ ﻭﺍﻟﺘﻭﺼﻴﺎﺕ‪:‬‬

‫‪ -1‬ﺘﺸﻜل ﻜﻼ ﻤﻥ ﺍﻟﺴﻠﺴﻠﺘﻴﻥ ﺍﻟﺯﻤﻨﻴﺘﻴﻥ ﻟﻠﻨﺎﺘﺞ ﺍﻟﻤﺤﻠﻲ ﺍﻹﺠﻤﺎﻟﻲ ﻭﺇﺠﻤﺎﻟﻲ ﺘﻜﻭﻴﻥ ﺭﺃﺱ ﺍﻟﻤﺎل ﺍﻟﺜﺎﺒـﺕ‬
‫ﻭﺠﻭﺩ ﺠﺫﺭ ﺍﻟﻭﺤـﺩﺓ‪ ،‬ﻭﻗـﺩ ﺃُﺨِـﺫﹶ‬ ‫‪Dickey and Fuller‬‬ ‫ﺴﻴﺎﻗﺎﹰ ﻋﺸﻭﺍﺌﻴﺎﹰ ﻏﻴﺭ ﻤﺴﺘﻘﺭ‪ ،‬ﻭﺃﻅﻬﺭ ﺍﺨﺘﺒﺎﺭ‬
‫ﻤﺭﺸﺢ ﺍﻟﻔﺭﻭﻕ ﺍﻷﻭﻟﻲ ﻟﺠﻌﻠﻬﺎ ﻤﺴﺘﻘﺭﺓ ‪.‬‬

‫)‪(Vectorial AutoRegressive‬ﻋﻠﻰ ﺍﻟﻜﺸﻑ ﻋﻥ ﺍﻟﻌﻼﻗﺔ ﺍﻟـﺴﺒﺒﻴﺔ‬ ‫ﻨﻤﺎﺫﺝ ‪VAR‬‬ ‫‪ -2‬ﺘﺒﻴﻥ ﻤﻥ ﺍﻟﺒﺤﺙ ﻗﺩﺭﺓ‬
‫ﺍﻟﻤﺘﺒﺎﻁﺌﺔ ﺒﻴﻥ ﺍﻻﺴﺘﺜﻤﺎﺭﺍﺕ ﻭﺇﺠﻤﺎﻟﻲ ﺍﻟﻨﺎﺘﺞ ﺍﻟﻤﺤﻠﻲ‪ ،‬ﻭﺃُﺨﺫ ﺒﺎﻟﺤﺴﺒﺎﻥ‪.‬‬

‫‪ -3‬ﻤﻥ ﺨﻼل ﺘﺤﺩﻴﺩ ﻋﺩﺩ ﻓﺘﺭﺍﺕ ﺍﻟﺘﺒﺎﻁﺅ ﺍﻟﺯﻤﻨﻲ ﻓﻲ ﺘﺄﺜﻴﺭ ﺍﻻﺴﺘﺜﻤﺎﺭﺍﺕ ﻓﻲ ﺍﻟﻨﺎﺘﺞ ﺍﻟﻤﺤﻠﻲ ﺍﻹﺠﻤـﺎﻟﻲ‬
‫ﻴﻤﻜﻥ ﺍﻻﺴﺘﺩﻻل ﻋﻠﻰ ﻤﺩﻯ ﺍﻟﺘﺄﺨﻴﺭ ﻓﻲ ﺇﺩﺨﺎل ﺍﻟﻤﺸﺭﻭﻋﺎﺕ ﺍﻟﺠﺩﻴﺩﺓ ﻓﻲ ﺍﻻﺴﺘﺜﻤﺎﺭ‪.‬‬

‫ﻟﻠﺘﻨﺒﺅ ﺒﺈﺠﻤﺎﻟﻲ ﺍﻟﻨﺎﺘﺞ ﺍﻟﻤﺤﻠﻲ‪ ،‬ﻭﻜﺫﻟﻙ ﺒﺤﺠـﻡ ﺍﻻﺴـﺘﺜﻤﺎﺭﺍﺕ ﻓـﻲ‬ ‫‪VAR‬‬ ‫‪ -4‬ﺘﻡ ﺍﻟﺘﻭﺼل ﺇﻟﻰ ﻨﻤﻭﺫﺝ‬
‫ﺴﻭﺭﻴﺔ ‪.‬‬

‫‪ -5‬ﻨﻭﺼﻲ ﺒﺎﺴﺘﺨﺩﺍﻡ ﺍﻟﻨﻤﻭﺫﺝ ﺍﻟﺫﻱ ﺘﻡ ﺍﻟﺘﻭﺼل ﺇﻟﻴﻪ ﻓﻲ ﺍﻟﺘﻨﺒﺅ ﺒﺤﺠﻡ ﺍﻟﻨﺎﺘﺞ ﺍﻟﻤﺤﻠﻲ ﺍﻹﺠﻤﺎﻟﻲ ﻭﻜـﺫﻟﻙ‬
‫ﺒﺤﺠﻡ ﺍﻻﺴﺘﺜﻤﺎﺭﺍﺕ‪ ،‬ﻭﺍﻋﺘﻤﺎﺩ ﺍﻟﺘﻨﺒﺅﺍﺕ ﺍﻟﺘﻲ ﺃﻋﻁﺎﻫﺎ ﺒﻭﻀﻊ ﺨﻁﻁ ﺍﻟﺘﻨﻤﻴﺔ ﺍﻻﻗﺘﺼﺎﺩﻴﺔ ﻭﺍﻻﺠﺘﻤﺎﻋﻴﺔ‬
‫ﺍﻟﻤﺴﺘﻘﺒﻠﻴﺔ ‪.‬‬

‫‪359‬‬
 ‫ ﻓﻲ ﺍﻟﺘﻨﺒﺅ ﻭﺩﺭﺍﺴﺔ ﺍﻟﻌﻼﻗﺔ ﺍﻟﺴﺒﺒﻴﺔ ﺒﻴﻥ ﺇﺠﻤﺎﻟﻲ ﺍﻟﻨﺎﺘﺞ ﺍﻟﻤﺤﻠﻲ ﻭﺇﺠﻤﺎﻟﻲ ﺍﻟﺘﻜﻭﻴﻥ ﺍﻟﺭﺃﺴﻤﺎﻟﻲ ﻓﻲ ﺴﻭﺭﻴﺔ‬VAR ‫ﺍﺴﺘﺨﺩﺍﻡ ﻨﻤﺎﺫﺝ‬

‫ﺍﻟﻤﺭﺍﺠﻊ‬
:‫ﺍﻟﻤﺭﺍﺠﻊ ﺒﺎﻟﻠﻐﺔ ﺍﻻﻨﻜﻠﻴﺯﻴﺔ‬
1- BERA, A.K. and Jarque .C.M.(1981), "An efficient large Sample test for normality of
observations and regression residuals ", Working paper in Econometrics No 40,Australion
National university, Canberra.
2- Cromwell, J. B., Hannan M. J., Labys W. C. and Terraza M. (1994), "Multivariate tests for
Time Series Models", SAGE publications, Inc. California.
3- Dickey D. and Fuller W.(1979), " Distribution of the estimators for Autoregressive Time Series
With a unit Root ", Journal of the American Statistical Association, n74: pp .427-431.
4- Dickey D. and Fuller W.(1981) 'The likelihood Ratio Statistics for Autoregressive Time Series
With a unit Root", Econometrica ,n49: pp .1057-1072.
5- Kirchgässner G. and Wolters J. (2007) "Introduction to Modern Time Series Analysis",
SPRINGER-Verlag, Berlin Heidelberg.
6- Ljung, G.M., and Box G.E.P. (1978) "on a measure of the lack of fit in time Series models".
Biometrika, n65:PP.297-303.
7- SHUMWAY R.H. and STOFFER D.S. (2006) "Time Series Analysis and Its Applications".
SPRINGER, New York.
8- SIMS C. A. (1981) "Macroeconomics and Reality". Econometrica, n48: pp 1-48.
9- Wei, W. W. S. (1990). "Time Series Analysis Univariate and Multivariate Methods", Addison
Wesley.
10- Zellner, A. and Palm, F. C. (2004). "The Structural Econometric Time Series Analysis
Approach" . Cambridge University Press, New York
:‫ﺍﻟﻤﺭﺍﺠﻊ ﺒﺎﻟﻠﻐﺔ ﺍﻟﻔﺭﻨﺴﻴﺔ‬
1- GOURIEROUX C. et MONFORT A., (1990) "Séries Temporelles et Modèles Dynamiques "
Ed. Economica-Paris.
2- HENIN P.Y. (1989), "Bilans et essais sur la non-Stationnarité des séries Macroéconomiques"
révue d' économie politique – n5-pp 661-691.
3- LARDIC S. et MIGNON V. ,(2002) "Econométrie des séries temporelles macroéconomiques
et financières" Ed. Economica-Paris.
:‫ﺍﻟﻤﺭﺍﺠﻊ ﺒﺎﻟﻠﻐﺔ ﺍﻟﻌﺭﺒﻴﺔ‬
. 2008- 1990 ‫ ﺍﻟﻤﺠﻤﻭﻋﺔ ﺍﻹﺤﺼﺎﺌﻴﺔ ﺍﻟﺴﻭﺭﻴﺔ‬-1
‫ﺠﺎﻤﻌـﺔ ﺍﻟﻤﻠـﻙ‬.(‫ ﻁـﺭﻕ ﺍﻟﺘﻨﺒـﺅ ﺍﻹﺤـﺼﺎﺌﻲ ) ﺍﻟﺠـﺯﺀ ﺍﻷﻭل‬،‫ ﻋﺩﻨﺎﻥ ﻤﺎﺠﺩ ﻋﺒﺩ ﺍﻟﺭﺤﻤﻥ‬،‫ ﺒﺭﻱ‬-2
. ‫ ﻡ‬2002،‫ﺴﻌﻭﺩ‬
‫ﺠﺎﻤﻌـﺔ ﺍﻟﻤﻠـﻙ ﻋﺒـﺩ‬. ‫ﻤﻘﺩﻤﺔ ﻓﻲ ﺍﻟﺘﺤﻠﻴل ﺍﻟﺤﺩﻴﺙ ﻟﻠﺴﻼﺴل ﺍﻟﺯﻤﻨﻴﺔ‬، ‫ﺴﻤﻴﺭ ﻤﺼﻁﻔﻰ‬، ‫ﺸﻌﺭﺍﻭﻱ‬ -3
. ( ‫ ﻡ‬2005) ‫ ﻫـ‬1426، ‫ﺍﻟﻌﺯﻴﺯ‬
.

.2011/6/7 ‫ﺘﺎﺭﻴﺦ ﻭﺭﻭﺩ ﺍﻟﺒﺤﺙ ﺇﻟﻰ ﻤﺠﻠﺔ ﺠﺎﻤﻌﺔ ﺩﻤﺸﻕ‬

360