This paper presents a stochastic framework for optimal scheduling of microgrids (MGs) considering unscheduled islanding events, initiated by disturbances in the main grid. This scheduling approach considers different uncertainties and determines the day-ahead schedule of the resources considering emergency operations. The proposed strategy attempts to effectively manage demand and supply side resources to mitigate the effects of uncertainties in both normal and emergency operations. The prevailing uncertainties associated with renewable power generations, demand and electricity prices as well as uncertainties of islanding duration are addressed in the presented framework. The objective is to maximize the expected profit of the operator over the scheduling horizon, while restricting the risk of mandatory load shedding imposed by uncertain parameters within an acceptable level. According to the proposed strategy, an efficient probabilistic index is obtained from generation reserve margin (GRM) in islanded mode, and applied to create a proper offering price signal to coordinate responsive loads with renewable generations providing more reliable operations. The effectiveness of the proposed strategy in terms of economy and reliability is investigated via a comparison with other methods. Extensive numerical results illustrate that the proposed offering price strategy can improve the MG's operation from both reliability and economic aspects.