Dynamic pricing (a.k.a. real-time pricing) is a method of invoking a response in demand pricing electricity at hourly (or more often) intervals. Several studies have proposed dynamic pricing models that maximize the sum of the welfares of consumers and suppliers under the condition that the supply and demand are equal. They assume that the cost functions of suppliers are convex. In practice, however, they are not convex because of the startup costs of generators. On the other hand, many studies have taken startup costs into consideration for unit commitment problems (UCPs) with a fixed demand. The Lagrange multiplier of the UCP, called convex hull pricing (CHP), minimizes the uplift payment that is disadvantageous to suppliers. However, CHP has not been used in the context of demand response. This paper presents a new dynamic pricing model based on CHP. We apply CHP approach invented for the UCP to a demand response market model, and theoretically show that the CHP is given by the Lagrange multiplier of a social welfare maximization problem whose objective function is represented as the sum of the customer's utility and supplier's profit. In addition, we solve the dual problem by using an iterative algorithm based on the subgradient method. Numerical simulations show that the prices determined by our algorithm give sufficiently small uplift payments in a realistic number of iterations.