1、影响四川省房地产业发展的因素分析影响四川省房地产业发展的因素分析“民以食 为天，以住为地”，房地产在国民经济生活中起着重要的作用.房地产业的发展对国民经济发展有着巨大而深刻的影响.因此，审慎分析房地产市场的现状及其基本走势,深入探讨影响房地产业发展的主要因素，对促进房地产业事业持续重要. 一、对房地产市场的基本判断 年间，我国房地产投资连续四年增幅在以上，年房地产投资增长率高达 ，投资规模首次突破万 亿元，占全社会固定资产投资的 占的。 在投资规模不断增长的同时，商品房的销售率也在稳步上涨。然而，在总体看好的情况下，局部地区投资增长过快、供求结构失衡的问题仍十分突出。除区域性问题外，投资结构也
2、不容忽视。近几年，全国普通住宅销售率连续四年在以上，且不断走高，但高档住宅、经济适用房、办公楼宇和商厦的销售率却都比较低，导致办公楼和商厦的大量闲置。简言之，此轮房地产投资热，有合理的因素，如中国经济的高速增长、城市化进程的加快、居民收入水平的提高、消费的升级换代、大量外资的进入和劳动力的流动形成的市场有效需求等，但是，政府人为造市、无视地区市场环境的盲目大干快上，导致奢靡之风、非理性投资泛滥，由此引发的投资热和结构失衡，对房地产业来说则是贻害长远。二、选定变量进行计量经济学的分析我们选定四川省房地产开发总值(单位:亿元)作为被解释变量,同时一共选定了八个解释变量:1四川省每年税收总额(单位:
3、万元)2四川省每年 储蓄存款总额(单位:亿元)3四川省建筑材料工业品出厂价格指数( 单位: )4四川省原材料燃料和动力购进价格指数( 建筑材料类)( 单位:)5四川省每人每年可支配收入(单位:元)6四川省( 单位:亿元)7货币 供应量( 单位:亿元)8贷款利率 (3 年-5 年期)(单位:)年份 YX1X2X3X4X5X6X7X819005.18 256513 698.15101.21001862.36890.9515293.411.7%19916.09 323350 813.4796.4118.22132.561016.3119349.911.7%199214.19486213904.689
4、2.4139.32988.211177.2725402.211.7%199334.87658821988.1787.2168.53422.171486.0834879.811.7%199450.578694511024.2198.5215.13846.222001.4146923.511.7%199579.9610520131148.59103.7273.34002.912504.9560750.511.7%199690.6212984071521.97140.5294.44406.092985.1576094.911.7%1997100.2615257921841.2297.0354.647
5、63.263320.1190995.39.9%1998120.6016602861984.5495.3417.25127.083580.26104498.57.65%1999142.5017789792351.2194.0417.25477.893711.61119897.96.03%2000195.9719463192693.1794.7444.75894.274010.25134610.36.03%2001268.1521814663123.3997.3470.76360.474421.76158301.96.03%构造模型: （数据见下页）0+11+2 2+33+44+55+66+77+
6、88对数据进行分析:Dependent Variable: YMethod: Least SquaresDate: 12/22/04 Time: 16:40Sample: 1990 2002Included observations: 13VariableCoefficientStd. Errort-StatisticProb. C9.50619061.473400.1546390.8846X1-0.0001080.000146-0.7371200.5019X2-1.32E-061.34E-06-0.9817610.3818X30.5152790.8940200.5763620.5952X4-
7、1.0198000.958646-1.0637930.3474X5-0.0194620.037130-0.5241620.6279X6-0.1540270.062463-2.4659140.0692X70.0079940.0011546.9251130.0023X89.8308403.2403013.0339290.0386R-squared0.995226 Mean dependent var115.8538Adjusted R-squared0.985677 S.D. dependent var114.4145S.E. of regression13.69309 Akaike info c
8、riterion8.277619Sum squared resid750.0024 Schwarz criterion8.668738Log likelihood-44.80453 F-statistic104.2251Durbin-Watson stat3.388729 Prob(F-statistic)0.000226对数据的多重共线性进行分析:X1X2X3X4X5X6X7X8X1 1.000000 0.992449-0.490698-0.579667 0.926422 0.942814 0.985242-0.808622X2 0.992449 1.000000-0.482882-0.56
9、8568 0.920679 0.932850 0.975851-0.819597X3-0.490698-0.482882 1.000000 0.900281-0.486587-0.511065-0.489324 0.424923X4-0.579667-0.568568 0.900281 1.000000-0.638015-0.648189-0.604672 0.455466X5 0.926422 0.920679-0.486587-0.638015 1.000000 0.995972 0.969636-0.620918X6 0.942814 0.932850-0.511065-0.648189
10、 0.995972 1.000000 0.981303-0.655603X7 0.985242 0.975851-0.489324-0.604672 0.969636 0.981303 1.000000-0.749205X8-0.808622-0.819597 0.424923 0.455466-0.620918-0.655603-0.749205 1.000000可见解释变量间存在多重共线性,对此我们进行修正,采用逐步回归：Y 与 X7 的拟合效果最好Dependent Variable: YMethod: Least SquaresDate: 12/22/04 Time: 16:54Sam
11、ple: 1990 2002Included observations: 13VariableCoefficientStd. Errort-StatisticProb. C-47.5999617.42865-2.7311330.0195X70.0019820.00017811.158270.0000R-squared0.918823 Mean dependent var115.8538Adjusted R-squared0.911444 S.D. dependent var114.4145S.E. of regression34.04795 Akaike info criterion10.03
12、406Sum squared resid12751.89 Schwarz criterion10.12097Log likelihood-63.22136 F-statistic124.5069Durbin-Watson stat0.631549 Prob(F-statistic)0.000000将其余变量逐一引入的如下几个模型:引入6:Dependent Variable: YMethod: Least SquaresDate: 12/22/04 Time: 17:01Sample: 1990 2002Included observations: 13VariableCoefficientS
13、td. Errort-StatisticProb. C24.2790625.195100.9636420.3579X6-0.0907270.027503-3.2987890.0080X70.0041510.0006706.1961140.0001R-squared0.961126 Mean dependent var115.8538Adjusted R-squared0.953351 S.D. dependent var114.4145S.E. of regression24.71162 Akaike info criterion9.451598Sum squared resid6106.64
14、1 Schwarz criterion9.581971Log likelihood-58.43539 F-statistic123.6208Durbin-Watson stat0.917644 Prob(F-statistic)0.0000006 通过检验 ,引入：Dependent Variable: YMethod: Least SquaresDate: 12/22/04 Time: 17:05Sample: 1990 2002Included observations: 13VariableCoefficientStd. Errort-StatisticProb. C-80.263953
15、4.93053-2.2978160.0472X6-0.1432060.024342-5.8829980.0002X70.0058510.0006768.6556460.0000X811.172183.2356693.4528190.0072R-squared0.983278 Mean dependent var115.8538Adjusted R-squared0.977703 S.D. dependent var114.4145S.E. of regression17.08441 Akaike info criterion8.761870Sum squared resid2626.894 S
16、chwarz criterion8.935700Log likelihood-52.95215 F-statistic176.3999Durbin-Watson stat1.513796 Prob(F-statistic)0.0000008 通过检验 ,引入2:Dependent Variable: YMethod: Least SquaresDate: 12/22/04 Time: 17:07Sample: 1990 2002Included observations: 13VariableCoefficientStd. Errort-StatisticProb. C-50.8081526.
17、01905-1.9527290.0866X2-2.24E-066.92E-07-3.2373780.0119X6-0.1627520.018028-9.0275450.0000X70.0074020.00067311.007200.0000X88.1099102.4481403.3126830.0107R-squared0.992761 Mean dependent var115.8538Adjusted R-squared0.989142 S.D. dependent var114.4145S.E. of regression11.92238 Akaike info criterion8.0
18、78435Sum squared resid1137.146 Schwarz criterion8.295723Log likelihood-47.50983 F-statistic274.2850Durbin-Watson stat2.434411 Prob(F-statistic)0.0000002 通过各 项检验,而其他变量未能通过各项检验,以上变量模型拟合度良好:利用 ARCH 检验是否存在异方差ARCH Test:F-statistic0.444405 Probability0.730040Obs*R-squared1.818050 Probability0.611015Test E
19、quation:Dependent Variable: RESID Method: Least SquaresDate: 12/22/04 Time: 17:18Sample(adjusted): 1993 2002Included observations: 10 after adjusting endpointsVariableCoefficientStd. Errort-StatisticProb. C197.6702116.59791.6953150.1409RESID (-1)-0.3897540.411792-0.9464810.3804RESID (-2)-0.4030480.4
20、44113-0.9075350.3991RESID (-3)-0.0912190.421512-0.2164100.8358R-squared0.181805 Mean dependent var103.1534Adjusted R-squared-0.227292 S.D. dependent var136.2191S.E. of regression150.9079 Akaike info criterion13.16039Sum squared resid136639.2 Schwarz criterion13.28143Log likelihood-61.80196 F-statist
21、ic0.444405Durbin-Watson stat1.941317 Prob(F-statistic)0.730040Obs*R-squared=1.8180507.81=x 0.05(3)所以不存在异方差利用 Cochrane-Orcutt 迭代法检验Dependent Variable: YMethod: Least SquaresDate: 12/22/04 Time: 17:36Sample(adjusted): 1991 2002Included observations: 12 after adjusting endpointsConvergence achieved aft
22、er 10 iterationsVariableCoefficientStd. Errort-StatisticProb. C-46.5320127.88659-1.6686160.1462X2-2.54E-069.14E-07-2.7756200.0322X6-0.1678690.018630-9.0107310.0001X70.0076740.0007899.7199520.0001X87.8254202.7018152.8963570.0275AR(1)-0.3078890.472444-0.6516950.5387R-squared0.992656 Mean dependent var
23、125.0767Adjusted R-squared0.986536 S.D. dependent var114.3435S.E. of regression13.26767 Akaike info criterion8.315390Sum squared resid1056.186 Schwarz criterion8.557844Log likelihood-43.89234 F-statistic162.2015Durbin-Watson stat2.025073 Prob(F-statistic)0.000003Inverted AR Roots -.31此时 D1DW=2.025073 4-D2 所以模型的自相关性得到修正,不存在自相关最终我们得到的模型为：-50.808152.24E-0620.162752 60.007402 78.109910 8三、经济意义分析从上述分析可知, 2四川省每年储蓄存款总额,6四川省,7货币供应量,8贷款利率(3 年-5 年期) 是影响四川省房地产开发总值的最主要的因素。 五.2005 年四川房地产业状况分析1、土地拍卖继续使房地产开发中土地成本上涨，利润降低。2、建材上涨房产运作压力加大3、配套设施提高成本上升4、央行加息房价上升5、土地增值税的严格征管6、信贷政策制约房产发展
