This thesis studies an economy (without growth) populated by overlapping generations of rich and poor agents, with the difference between them being that the rich are endowed with larger amounts of wealth in both period. Both types of agents want to transfer wealth over time and smooth out their consumptions. Agents could borrow from or lend to each other in credit market, when borrowing they are subject to a collateral constraint. Agents could also buy house(s), which are indivisible and pay no dividend whatsoever. Because of the lack of investment instruments and dynamic inefficiency in bubble-free steady states, a rational housing bubble could arise. Due to its high prices and indivisibility, houses are not affordable to the poor. So in the bubble steady states only the rich get to buy houses and enjoy the high rate of return caused by the housing bubble, while the poor are left out, implying potential distributional effect. With a relaxed collateral constraint, the market interest rate is relatively high and the rich could borrow a lot from the poor through the credit market, thus providing them (the poor) with a mean to transfer wealth across periods and benefit from the housing bubble indirectly. A tight borrowing constraint, on the other hand, limits the rich’s borrowing capacity and hurt the poor’s lifetime utility indirectly through lower market interest rate (and the resulting inefficient wealth transfer). This model could be applied to China, where we have witnessed drastic increase in housing prices during the last 15 years. Several researches and empirical evidences point towards the existence of a rational housing bubble in China. Upon a closer look at survey data from selected years we can see a rise in household consumption inequality from 1999 to 2002, and then to 2008. Other research has also found that increasing housing prices have caused much of the increase in wealth inequality in China.