|
开封翻译公司关键字：From Wuhan City, the premium prices and time-series data plot (Figure 1) can be seen: over time, house prices and land prices increase at different rates, with time as the independent variables were the prices, land as the dependent variable to build a regression equation (regression results shown in Figure 1), the rate of change over time prices 50.616, 15.485 more than the marginal value of land. Their statistical significance, as the city's economic growth, housing demand and the contradiction between the increasing land supply, which is the case of Wuhan City and the country basically.
To further describe the Wuhan City housing prices and the magnitude of change over time, land prices, land prices and the author of the time series data for the variance of the hypothesis test:Test results show that the relative land prices, the large variation in prices in Wuhan. Visible, Wuhan land and land market regulation mechanism of the formation mechanism has gradually matured, showing steady upward trend of land prices.Wuhan City, land and housing prices Granger causality testAssumptions and Inspection RulesAssumes that the current land price (LP) and its own in house prices (HP) on past values, and price (HP) and its own past values ??and on land. Inspection requirements estimated the following regression:Results point four cases: if "(1)" in the lag coefficient estimates to meet HP Σαj ≠ 0, and "(2)" the lag in the estimated value of LP coefficients satisfy Σδj = 0, then indicates a one-way from HP to LP causality (unidirectional causality). If Σαi = 0, and Σδj ≠ 0, then there exists a one-way from LP to HP causal relationship. If Σαi ≠ 0, and Σδj ≠ 0, then the said land prices and a two-way (bilateral) causality. If Σδj ≠ 0, and Σi = 0, it means the independence between the respective (independence).Whether the price premium for the reason, assuming H0: Σαi = 0, that is not part of this lag HP key return. With the F test:Where, RSSR for the current LP LP of all items for the lag regression analysis, to be bound by the residual sum of squares; RSSUR expressed "a style" for return, get the residual sum of squares; m is equal to the lag number; k is the "type 1" in the number of parameters to be estimated.the selected level of significance calculated F value exceeds the critical F value, reject the virtual assumption.Similarly land can test whether the reason for the price. Take lag two (m = 2) tested showed between land prices and no significant causal relationship; Granger test due to the application, the distributed lag model with similar problems, the causal direction and contained The number of lags may have an important relationship (Damodar, 2003), to enhance the confidence of the above conclusions were obtained m =
|