Carryover activities in dynamic DEA play a crucial role in transmitting information between periods and maintaining the continuity of resources from one period to the next within organizational processes. These variables reflect the impact of past decisions on current and future performance. In the literature of Dynamic DEA, carryovers typically connect two consecutive periods. However, in real-world scenarios, certain carryover variables in the production process may persist beyond the immediate subsequent period. The allocation of these carryovers is discretionary and lies within the control of the Decision-Maker (DM). This paper introduces a novel dynamic network DEA (DNDEA) model within the Slack-Based Measure (SBM) framework, with the aim of optimizing the allocation of carryovers to the period under evaluation. Our model not only enhances overall efficiency evaluation but also identifies maximum inefficiency within a network system across multiple evaluation periods. To validate the proposed model, we perform a numerical example focused on the performance assessment of Iranian bank branches using Dynamic DEA techniques. Through comparative analysis against the DNSBM model, we demonstrate that our proposed model exhibits greater discriminatory power. Additionally, it provides more comprehensive insights into resource allocation strategies.