I 摘要 住房与全民生活息息相关，住宅价格波动起伏受到全社会广泛关注。近年来， 国家层面再三重申“房住不炒”定位，要求地方政府切实落实城 市主体责任，及时采 取针对性措施，稳妥实施房地产长效机制，要求住房市场价格总体平稳运行。对 此，如何更科学的制定住宅价格成为房地产企业和学者们研究的重要课题。 将 Hedonic 特征价格模型作为住宅的定价方法，在国外的长期实践表明了其有 效性。自上世纪九十年代 Hedonic 特征价格模型被引入国内后，此定价方法被国内 学者广泛的研 究和应用，随着住房销售备案数据的不断完善，以及住宅特征变量 研究方法及其量化准确性的不断提高，此法定价精准性也越来越高。 本文基于 Hedonic 特征价格模型进行北城 时代项目住宅的定价研究，选择了 2020 年第一季度仁寿县城区域在售的 22 个住宅小区、成交备案的 1378 套住宅作 为样本数据，通过客户需求问卷调查和专业人士访谈结合国 内外文献资料研究， 从可能影响住宅价格的 66 个特征变量中筛选出 18 个变量，其中建筑特征变量 6 个：建筑面积、住房功能间数、装修水平、电梯梯户比、楼层、建筑类型 ；邻里 特征变量 9 个：容积率、绿地率、小区景观、品牌影响、物管费、车位配比、小 区配套、不利影响、人文自然环境；区位特征变量 3 个：区域板块、城市中心距 离、公 共交通。 本文研究运用 SPSS.23 软件进行回归分析，通过判定系数对比，从 Hedonic 特征价格模型常用的线性函数、线性对数函数、对数函数、对数线性函数中选择 拟合程度 最优的对数模型来构建北城时代项目住宅特征价格模型，模型经过共线 性检验、显著性检验、方差齐性检验、方差分析、自相关检验、稳定性检验等一 系列的分析与检验后获得 通过。最后，采用本文研究得到的特征价格模型计算价 格与选取住宅的备案价格之间的误差率为 0.9%，经过实际效果验证表明模型效果 良好。 在文章的最后，笔者认为定价要 反应开发商利润的最大化与购房者效用最大 化，将别墅、多层、高层等不同住宅建筑类型放在同一维度下进行研究，不同产 品对不同的购房者而言其效用可能存在差异，如果按 照住宅产品的细分市场进行 研究，效果可能更好。 关键词：Hedonic，特征价格，特征价格模型，住宅定价ABSTRACT II ABSTRACT The housing price is closely related to people's life. The fluctuation of housing price will also affect people's life and work. In recent years, the national level has repeatedly reiterated the positioning of "housing without speculation", requiring local governments to earnestly implement the main responsibility of the city, take timely targeted measures, steadily implement the long-term mechanism of real estate, and require the overall stable operation of housing market prices. In this regard, how to make a more scientific housing price has become an important topic for real estate enterprises and scholars. Taking hedonic price model as the pricing method of housing, the long-term practice in foreign countries shows its effectiveness. Since hedonic price model was introduced into China in the 1990s, this pricing method has been widely studied and applied by domestic scholars. With the continuous improvement of housing sales record data, and the continuous improvement of housing characteristic variable research method and its quantitative accuracy, the pricing accuracy of this method is also higher and higher. Based on hedonic price model, this paper studies the pricing of Beicheng era project housing. It selects 1378 residential units of 22 residential districts which are completed in the first quarter of 2020 as the sample data. Through customer demand questionnaire survey and professional interviews, combined with domestic and foreign literature research, 18 variables are selected from 66 characteristic variables that may affect housing price There are 6 building characteristic variables: building area, number of housing function rooms, decoration degree, floor, building type, elevator to household ratio; 9 neighborhood characteristic variables: plot ratio, green space rate, community landscape, brand influence, property management fee, parking space ratio, community supporting, adverse impact, human and natural environment; 3 place characteristic variables: regional plate, city center distance, public transport. In this paper, spss.23 software is used for regression analysis. Through the comparison of decision coefficients, the logarithmic model with the best fitting degree is selected from the linear function, logarithmic function, logarithmic linear function and linear logarithmic function commonly used in hedonic price model. The model isABSTRACT III tested by collinearity test, significance test and variance homogeneity test Analysis of variance, autocorrelation test, stability test and so on. Finally, the error rate between the price calculated by the hedonic price model and the recorded price of the selected housing is 0.9%. The actual effect verification shows that the model is effective. At the end of the article, the author thinks that the pricing should reflect the maximization of the developer's profit and the maximization of the buyer's utility. If the villa, multi-storey, high- rise and other residential building types are studied in the same dimension, the utility of different products may be different for different purchasers. If the research is conducted according to the subdivision market of residential products, the effect may be better. Key words: hedonic price, hedonic price model, housing pricing目 录 IV 目 录 第一章 绪论........................1 1.1 研究背景...............1 1.2 研究的意义及目的..............................1 1.2.1 研究的意义1 1.2.2 研究的目的2 1.3 研究内容及方法...2 1.3.1 研究内容....2 1.3.2 研究方法....3 1.4 技术路线...............4 1.5 可能的创新之处...4 第二章 HEDONIC 特征价格模型....................6 2.1 HEDONIC特征价格模型.......................6 2.1.1 Hedonic 特征价格模型的起源..6 2.1.2 Hedonic 特征价格模型的发 展..6 2.1.3 Hedonic 特征价格模型在国内的研究情况.............8 2.2 HEDONIC特征价格模型理论...............9 2.2.1 特征价格模型的应用前提与假 设..........................9 2.2.2 市场供需与特征市场均衡.....11 2.3 本章小结.............13 第三章 北城时代项目住宅特征与特征变量.14 3.1 北城时代项目住宅情 况....................14 3.2 北城时代项目住宅的主要特征........14 3.3 北城时代项目住宅价格的主要影响因素.......................15 3.3.1 住宅物业自身因素影 响.........16 3.3.2 住宅周边环境因素影响.........17 3.3.3 社会经济因素影响.................18 3.3.4 国家、地方、行业的政策因素影响....................19 3.3.5 市场需求因素.........................20 3.3.6 市场供给因素.........................20 3.3.7 心理因素影响.........................20目 录 V 3.3.8 其他影 响因素.........................20 3.4 北城时代项目住宅特征变量............21 3.4.1 问卷调查..21 3.4.2 专业人士访谈.........................22 3.4.3 现有文献 特征变量选择.........23 3.4.4 特征变量甄选原则.................24 3.4.5 特征变量确定.........................26 3.5 本章小结.............27 第四章 变量数据获 取与量化.........................28 4.1 研究区域.............28 4.2 特征变量数据采集............................28 4.2.1 房管局成交备案数据.............29 4.2.2 百度地图..29 4.2.3 住宅项目调查数据.................29 4.3 北城时代住宅项目特征变量的量化29 4.3.1 定量变量的量化.....................29 4.3.2 定性变量 的量化.....................31 4.4 本章小结.............34 第五章 实证研究..............35 5.1 特征价格模型选择............................35 5.1.1 特征价格模 型变量的设置.....35 5.1.2 特征价格模型的选择.............36 5.2 样本数据的统计性描述....................36 5.3 模型的估计与检验............................37 5.3.1 方差分析..37 5.3.2 显著性检验.............................38 5.3.3 共线性检验.............................38 5.3.4 方差齐性检 验.........................39 5.3.5 自相关检验.............................40 5.3.6 残差的正态性检验.................40 5.3.7 稳健性检 验.............................42 5.4 结果分析.............43 5.4.1 显著性分析及和符号分析.....43目 录 VI 5.4.2 价格弹性分析.........................45 5.4.3 住宅特征边际价格分析.........47 5.4.4 不显著性变量分析.................48 5.5 模型效果检验.....48 5.6 本章小结.............49 第六章 研究总结与展望..51 6.1 总