This paper applied conventional cobweb model, and brought in the home owners' down payment ratio and expected future housing price inflation rate, under the process of price formulation associnted with ”forward learning”, a non-linear dynamic price adjustment model thus developed. Next, by utilizing financial leverage ratio and the reaction coefficient representing home owners' attitude toward the return on housing investment to investigate the characteristics regarding the dynamic adjustment of market price. This paper utilized mathematical economics method to uncover the home owners' behaviors involved in the housing consumption and investment, including the conditions and the economic meanings of market equilibrium tendency, the recurring cycles or chaos. This research concluded that in a housing market that is attributed with consumption and investment characteristics, where the majority of home owners are investors then the dynamic adjustment of house price could very well lead to a non-linear type.