Study On The Impact Of COVID-19 On Second-hand Housing Prices In Wuhan

With the rapid development of China’s national economy and the increasing improvement of people’s income level,the real estate industry has developed rapidly.Since 2003,rising housing price around,and the land available for development in cities has become less and less,so second-hand housing has become more and more popular.In2020,the outbreak of COVID-19 dealt a major blow to the economy and housing prices were also affected.Wuhan was severely affected by the COVID-19.It is of great significance to study the impact of COVID-19 on second-hand housing prices to promote the healthy development of the second-hand housing market in Wuhan.This article uses web crawler by Python to obtain the transaction data of all second-hand houses in Wuhan from January 2016 to March 2021 on the Lianjia website,and the Baidu map API is used to query the infrastructure construction around the corresponding community.The transaction data of all second-hand houses including housing information,surrounding facilities and location characteristics of 18characteristic variables in 3 dimensions are obtained.At present,many scholars focus on the average price and listing price of second-hand houses.This paper uses the total transaction price of second-hand houses as the response variable,which can better reflect the real value of houses.In order to have a more comprehensive understanding of the impact of COVID-19 on the price of second-hand houses in Wuhan,this paper also selects Changsha and Nanjing,which are both new first-tier cities,as control cities.By comparing the fluctuations of second-hand housing prices and influencing factors of second-hand houses in the three cities during COVID-19,the impact of COVID-19 on second-hand housing prices in Wuhan is studied from multiple perspectives.After pre-processing the original data,the transaction data of 162,628 ordinary residential second-hand houses in 22 administrative regions of Wuhan,Changsha and Nanjing are finally obtained.During the descriptive statistical analysis of the data,it is found that the distribution of response variables and correlation between indicators in each administrative region are significantly different.Therefore,this paper establish models according to administrative regions.First of all,we use three different feature selection method for dimension reduction of pre-COVID-19 data,and the methods are Lasso,mutual information and stepwise regression.Indexes that have little influence on response variables are removed to reduce the redundancy of the model.Then four machine learning models,random forest,support vector machine,XGBoost and neural network,are used to build a regression model for the data after dimension reduction.Combining three feature selection methods with four machine learning methods,12 different regression models are obtained.Compared with the four regression models established by using machine learning methods alone,the combined model is generally better than the single regression model.The evaluation indexes R~2of each model are optimized by cross validation and grid search method.In each administrative region,the optimal model among 12 models is selected to predict the second-hand housing price during COVID-19 and make a comparative analysis with the actual transaction price.In selecting the optimal model,it is found that R~2of multiple models is very close.In order to select the optimal model more accurately,we use the data from 2016 to 2018 as the training set and the data from2019 as the validation set,selecting the best model to forecast the data during COVID-19.The residual value is used in this paper to measure the impact of COVID-19 on housing price.The larger the residual value,the more serious the suppression of COVID-19 on housing price.Results show that there were no transactions of second-hand houses in Wuhan from January 23,2020 to April 3,2020,indicating that COVID-19 had a severe impact on the second-hand house market in Wuhan at that time.By comparing the impact of COVID-19on the transaction prices of second-hand houses in Wuhan,Changsha and Nanjing,it is found that the transaction prices of second-hand houses in Wuhan generally declined from January to September in 2020,with the peak in April.Due to the proper containment of COVID-19,housing prices began to recover steadily in October 2020.Housing prices in Nanjing and Changsha are affected by COVID-19 for a relatively short time.It can be seen that the transaction price of second-hand housing in Wuhan has been affected by COVID-19 for a long time and to a serious degree,but the affected period is concentrated.Further studied the changes of influencing factors of second-hand houses before and after COVID-19 based on the situation of second-hand house transaction prices in Wuhan affected by COVID-19,it is found that the distance to the nearest general hospital is one of the main factors influencing the transaction price of second-hand houses in Wuhan during COVID-19.It shows that the sudden outbreak of COVID-19 makes people tend to choose second-hand houses closer to hospitals.According to the above research results,we put forward the following policy recommendations:(1)At present,COVID-19 has eased,and the relevant government departments should focus on the overall housing price in the secondary housing market,so as to ensure the stable and orderly operation of the real estate market;(2)People reconsider the allocation of medical facilities around the house during the period of COVID-19,and the government should regulate it reasonably to meet the needs of residents;(3)Before and after COVID-19,the building area is the key factor affecting the second-hand housing prices.Therefore,the government plans to build more than 70-100 square meters of houses to meet the needs of most people;(4)Consumers should have confidence in second-hand houses and purchase houses rationally according to their own needs and consumption ability.