计量经济学论文(eviews分析) 房价的计量经济分析

房价的计量经济分析引言：近改革开放20多年来，从来没有哪一个行业像房地产业这样盛产亿万富翁，各种富豪排行榜上，房地产富豪连年占据半壁江山；“中国十大暴利行业”中，房地产业每年都是“第一名”。是什么造就了这样的状况。房地产的问题，在开发商，政府，购房者三者来看，就是一场完完全全的博弈。而这场博弈的焦点则是房价问题。如果说开发商与政府之间的博弈是围绕“土地”这个关键词，那么整个房地产市场则在价格上开展了新一轮的对峙。先是开发商与购房者在房价涨跌上僵持不下；再有开发商与政府之间的土地成本论；最后则是关于房地产是否归为暴利行业的争执，“价格”成了市场关注的焦点。而对于房价的构成因素，至今仍然是不透明的。公布房价成本成为另政府极为头疼的一件事。房价成本是一个非常复杂的集合体，并且项目间差异性较大，同时还有软资产、品牌等组成部分，特别是现在的商品房，追求品质、功能完善以及个性化成本构成越来越难衡量。写作目的：通过对一系列影响房价的基本因素的分析，了解对其主要因素和次要因素。并对这些因素进行统计推断和经济意义上的检验。选择拟和效果最好的最为结论。在一定层面上分析房地产如此暴利的因素。当然笔者的能力有限，并不能全面的分析这一问题。仅仅就几个因素进行分析。写作方法：理论分析及计量分析方法，将会用到Eviews软件进行帮助分析。关键词：房价成本计量假设检验最小二乘法拟合优度现在我们以2003年的数据，选取30个省市的数据为例进行分析。在Eviews软件中选择建立截面数据。现在我们以2003年的数据，选取31个省市的数据为例进行分析。令Y=各地区建筑业总产值。（万元）X1=各地区房屋竣工面积。（万平方米）X2=各地区建筑业企业从业人员。（人）X3=各地区建筑业劳动生产率。（元/人）X4=各地区人均住宅面积。（平方米）X5=各地区人均可支配收入。（元）数据如下：YX1X3X2X4X5126985214254.800569767.0129961.024.7714013882.625208402.1465.800238957.0147063.023.0957010312.917799313.4748.300989317.070048.0023.167107239.0605401279.1313.300591276.089151.0022.996807005.0302576575.1450.700265953.061074.0020.053107012.900101707943957.100966790.082496.0020.235107240.5803469281.1626.800303837.077486.0020.705907005.1704401878.2181.300441518.068033.0020.492006678.900119580343609.200505185.0153910.029.3453014867.492794935417730.002727006.100569.024.435309262.4603127277916183.902429352.127430.031.0233013179.536227073.4017.600910691.066407.0020.754806778.0305493441.2952.100553611.0108288.030.298709999.5403593356.2750.900574705.070826.0022.619806901.420148136189139.8002072530.60728.0024.480808399.9106345217.3433.600932901.066056.0020.200906926.1208729958.4840.8001048763.81761.0022.902807321.9808188402.4969.7001119106.74553.0024.425807674.200151632428105.0001492820.101932.024.9328012380.432818466.1721.600353700.077472.0024.173207785.040394053.0121.500061210.0055361.0023.432007259.2505862095.4939.600817997.069432.0025.724408093.670122533748784.6002070534.59748.0026.358507041.8702122907.980.3000293310.072152.0018.194306569.2303967957.2248.700522470.069238.0024.929407643.570293427.0121.300036593.0073205.0019.929908765.4504404362.1580.000410311.093212.0021.750506806.3502236860.1327.200449409.046857.0021.113806657.240747325.0242.9000101501.061046.0019.105506745.3201080546.578.700088225.0061459.0022.255006530.4803196774.1450.800203375.095835.0020.781107173.540做多重共线性检验：引入的变量太多，可能存在变量间的共线性，影响方程的估计。首先进行做多重共线性检验可以减少变量使后面的分析变得简洁。X1X2X3X4X5YX110.9608709909074460.2713751927607750.5386972790690410.4183068002953290.961473842608042X20.96087099090744610.1250293750973190.477885891518730.2798506233443580.898672551511606X30.2713751927607750.12502937509731910.5408809599699260.836240848942410.467710383760092X40.5386972790690410.477885891518730.54088095996992610.686512808507740.589777148826127X50.4183068002953290.2798506233443580.836240848942410.6865128085077410.58982338526214Y0.9614738426080420.8986725515116060.4677103837600920.5897771488261270.589823385262141可以看出有多重共线性。采取逐步回归法：第一次回归，我们可以根据T检验值和可决系数看出：X1的效果最好：DependentVariable:YMethod:LeastSquaresDate:12/06/10Time:17:37Sample(adjusted):131Includedobservations:31afteradjustmentsVariableCoefficientStd.Errort-StatisticProb.X11651.40387.6770318.835080.0000C903234.0502408.21.7978090.0826R-squared0.924432Meandependentvar7446408.AdjustedR-squared0.921826S.D.dependentvar7227629.S.E.ofregression2020815.Akaikeinfocriterion31.93824Sumsquaredresid1.18E+14Schwarzcriterion32.03076Loglikelihood-493.0427F-statistic354.7601Durbin-Watsonstat1.930762Prob(F-statistic)0.000000而X1于X2存在严重自相关，所以引入第二个变量时将X2排除。通过比较发现引入X3时，拟合优度最大，所以加入X3DependentVariable:YMethod:LeastSquaresDate:12/06/10Time:17:40Sample(adjusted):131Includedobservations:31afteradjustmentsVariableCoefficientStd.Errort-StatisticProb.X11547.35457.8319726.756040.0000X360.575779.1368996.6297950.0000C-3711880.765709.2-4.8476370.0000R-squared0.970594Meandependentvar7446408.AdjustedR-squared0.968493S.D.dependentvar7227629.S.E.ofregression1282914.Akaikeinfocriterion31.05893Sumsquaredresid4.61E+13Schwarzcriterion31.19771Loglikelihood-478.4134F-statistic462.0886Durbin-Watsonstat2.098685Prob(F-statistic)0.000000X3与X5也存在严重共线性，在引入第三个变量时同时排除X5，那只能引入X4了DependentVariable:YMethod:LeastSquaresDate:12/06/10Time:17:47Sample(adjusted):131Includedobservations:31afteradjustmentsVariableCoefficientStd.Errort-StatisticProb.X11569.18666.7446723.510290.0000X364.0494510.562586.0638100.0000X4-69455.16102797.7-0.6756490.5050C-2476469.1985261.-1.2474280.2230R-squared0.971083Meandependentvar7446408.AdjustedR-squared0.967870S.D.dependentvar7227629.S.E.ofregression1295550.Akaikeinfocriterion31.10668Sumsquaredresid4.53E+13Schwarzcriterion31.29171Loglikelihood-478.1536F-statistic302.2316Durbin-Watsonstat2.298423Prob(F-statistic)0.000000但是引入后通过T检验X4不显著，同时常数项C也变得不显著，且拟合度没有显著提高。所以剔除X4。通过该检验最终模型为：Y=1547.354325*X1+60.57576644*X3-3711880.158T=26.756046.629795-4.847637R-squared0.970594Durbin-Watsonstat2.098685以上指标都显示拟合得很好。F-statistic354.7601异方差检验WhiteHeteroskedasticityTest:F-statistic1.742532Probability0.161697Obs*R-squared8.011602Probability0.155597TestEquation:DependentVariable:RESID^2Method:L