[Objective] To address the dispatching challenges caused by the randomness of renewable energy output and the delay in demand response within photovoltaic-storage direct current flexible （PSDF） microgrids， this paper proposes a stochastic optimal dispatching strategy that explicitly considers demand response delay characteristics. [Methods] First， a two-layer coordinated optimization model covering the energy producer （EP） and the load aggregator （LA） is established， incorporating three types of flexible loads—shiftable， transferable， and reducible—with delay response characteristics to accurately capture realistic user-side control behaviors. Second， an asymmetric bargaining mechanism based on Nash bargaining theory is developed to account for differences in bargaining power between the two parties， thereby achieving fair and efficient electricity pricing. Furthermore， to handle the uncertainties in photovoltaic and wind power output as well as load demand， a Monte Carlo scenario generation combined with K-means clustering reduction is employed to construct typical operating scenarios. Finally， the alternating direction method of multipliers （ADMM） is applied for distributed optimization to ensure convergence and privacy preservation. [Results] Simulation results indicate that the proposed model effectively reduces the cost of the load aggregator by 1.17%， increases the profit of the energy producer by 4.07%， and decreases the overall operating cost of the microgrid by 10.03%. [Conclusions] The proposed strategy balances the economic performance of the load side and the system power equilibrium， achieving a unified improvement in economy， fairness， and privacy. It provides a practical and scalable solution for optimal dispatching of microgrids with high renewable energy penetration.