A predictive and iterative trade model based on risk cost assignment is proposed to manage dynamic transmission congestion risks in a power market. The instability risk under working conditions resulted from predictive market clearing is defined as the product of fault probability and emergency control cost. The optimal emergency control is determined by search for the minimum value of risks. The risk assignment plan among the critical machines is announced in real time to guide the competing market participants to consider risk cost through rebidding. Bidding iterations give the point where minimum sum of market clearing price（MCP）-based purchasing cost and instability risk is obtained. At this point， the MCP-based purchasing cost reflecting preventive control is optimally coordinated with emergency control. The model can introduce participants to decrease dynamic congestion risks in an open， fair and just way. The influence factors of dynamic congestions and the mechanism of interactions between stability control at physical aspect and congestion management at the economical aspect of a power system are also studied based on a dynamic and interactive simulation tool which runs market competitions and physical system analysis simultaneously.