The rapidly rising prices in China's housing market have attracted a lot of heated discussions . There has been a large volume of researches examining the effects of expectation on housing price,but few studies have been done on how these effects change over time and in different market situations . Using the data on monthly constant quality housing price and rent from January2005 to December2010 in Beijing,this paper analyzes the impact of expectation on housing rent to price ratio .Based on a user cost model,we studied the asymmetric pattern of the expectation and housing price dynamics . Three main results have been obtained from this study . First,from2005 to2010,the rent to price ratio in Beijing increased by175 percent,with housing price rising7 .3 times that of rent .For most of the time during this period,the return rate in the housing market was lower than the risk-free interest . All of these show that the housing price hikes in Beijing are associated with high risks as well . Second,using a panel data to test eight districts of Beijing from January2005 to December 2010,we find that changes of rent to price ratio can mostly be explained by the extraordinary high expectations on the capital gains .Rent to price ratio will change by 20 .50% and 32 .94% respectively in response to a unit standard deviation change in real interest and expectation . Third,there is a positive correlation between the coefficient of the expectation on housing price and the real housing price . During the boom period,expectation plays a significant role in pushing up housing prices .On the contrary,during the relatively stagnant period,the role of expectation for the housing price movement becomes less important . Such expectation-driven market behavior reveals a potential hazard in China's housing market . There are two findings in the current study . First,due to the lack of high quality rental data,most rent to price studies calculate rent and house price from the rent index and the housing price index . A comparison of the two smoothed aggregate time series (the two indices) may only reveal the trend but miss the risks associated with the real housing price movements .Based on the housing transaction data on second-hand housing and rental data with detailed housecharacteristics,we constructed pairs of constant quality housing price and rent data for a more reliable analysis . Second,this paper not only studies how expectation influences housing rent to price ratio,but also how the influence changes in different market situations .By rolling the sample time window,we estimated the coefficients of expectation on housing price in different times,and found an asymmetric pattern of the impacts of expectations on housing price movements .