A Study On Price Making Of Retirement With House Based On The Combination Forecasting

The problem of China's aging population is becoming more and more serious.With the increasing of the family endowment pressure,the way of government pension can't meet the needs of the old man.As a new type of the pension mode,the advance of the retirement with house in China is not entirely as desired.An important reason is that the valuation of the house is unreasonable.So choosing an appropriate method to assess the value of building in the future is the key to carry out the policy better.At present,there are three methods to valuate the building's future price: Time series model,Random-walking model and Multi-factor equation model.Most people use single model to forecast the house price of the future.Even some just use the means of the history.Using a single method to evaluate can't foster strengths and circumvent weaknesses.So it's important to select appropriate methods.On the basis of relevant theories and practice,this paper compares the advantages and disadvantages of the three typical methods.Then select appropriate models of each method: ARMA model,Geometric brownian motion model,the BP neural network model to value the future house price.Due to the special requirements of the housing pension: long-term randomness,short-term self-correlation,taking accounting into the impact of house prices and higher accuracy,while the ARMA model is not reasonable to forecast the long-term volatility and the Geometric brownian motion model can't be used to the short-term forecast.So it combines the short-term result with the long-term result.To take the influence of various factors of the price of the house into consideration,it combines the results of the BP neural network model with the former combined result.Then,it takes Hangzhou as an example,using the data of Hangzhou to predict the price fluctuation.It finds that the result of the combined model is the most accurate way.Next,it puts other parameters and the future house price into the pricing model to get the pension amount including the one-time payment and the annuity.The result is a little higher than the price in the practice.But it's reasonable when compared with the history study,and only a higher price will attract more people to join in.At last,because of the uncertainty in the long term,it makes a dynamic adjustment with the relative error to get a more accurate result.Besides that,this paper makes a better depreciation rate of the second-hand house.It draws on the previous point of view that the depreciation decreases with the increasing age of the house while not use the 70 years as the depreciation period.The improvement in this article can make the accuracy of the forecast,so that it can have a more appropriate price of the product.The authority of the forecasting will mobilize financial institutions and the elderly to participate in the housing endowment,then expand the business of the house valuation in the valuation institutions.