The current rate structures for the electricity retailing exposes utility providers to the full wholesale market risks, and fail to incentivize end-use customers to better estimate and track their loads. In this paper, we propose a Cost-for-Deviation (CfD) retail-pricing scheme, which is designed to minimize the demand uncertainty of individual customers or communities. We formulate day-ahead planning and real-time tracking optimization problems for individual buildings. We also formulate CfD pricing scheme for community of two buildings and devise a collaboration scheme by which the two buildings negotiate. Both centralized and distributed negotiation mechanisms are presented, and the significance of adopting a transaction cost for fair-trading is discussed. A series of experiments demonstrate that CfD pricing is able to reduce demand uncertainty in a building or a community. Hence, a community's cost of hedging quantity risk in the real-time market also reduces. Our conjecture is that by the virtue of end users being in a position to closely monitor their daily loads and by paying fines for not adhering to their plans would ultimately benefit energy efficiency and will reduce infrastructure costs. (C) 2016 Elsevier B.V. All rights reserved.