中国商品市场名义价格粘性的测度

An Estimate of the Degree of Nominal Price Stickiness in China

本文利用网络文本提取与挖掘技术,构建超过15亿条观测的产品级高频数据集,对中国商品市场2010年12月至2013年2月间名义价格粘性程度进行估算。结果表明:(1)相对于发达国家,我国的价格粘性程度处于较低水平,总体价格变化的加权中位数频率是每天1.23%,总体价格持续时间中值为2.7个月,剔除促销的影响后,价格持续时间中值增加到3.4个月。(2)划分子样本分析发现,东部地区和中部地区价格粘性程度比较接近且均高于西部地区,进口品和成交量排名前20零售商的商品价格更为灵活。(3)价格调整幅度的分布呈双峰形态,与状态相关定价模型的预测一致,意味着中国商品市场企业定价存在一定程度的"选择效应"。(4)通货膨胀方差分解发现,集约边际解释了通货膨胀方差的64%,表明中国通货膨胀主要来源于商品价格变化幅度的扩张。本文的研究为构建更具微观基础的宏观经济模型提供了新的经验证据。

Abstract：Using webtext extraction and data mining technology, this paper constructed a product-level, high frequency data set of more than 1.5 billion observations from December 2010 to February 2013 to estimate the nominal price stickiness in Chinese commodity market. The results show that ： （ 1 ） Compared with developed countries, the degree of price stickiness in China is at a low level, with a weighted median price change frequency of 1.23% per day. The median implied duration of price is about 2.7 months with sales and 3.4 months without sales. （2） The degree of price stickiness in eastern and central regions is relatively close and was higher than that of the western region. The price of imported goods is more flexible as well as top 20 goods according to the sales ranking. （3） The distribution of the size of price changes is bimodal, consistent with predictions of State-Dependent Pricing model （Golosov and Lucas, 2007）. （4） The intensive margin accounts for 64% of inflation＇s variance, indicating that magnitude of price changes is the main source of inflation in China.