This paper presents a novel market clearing mechanism for manual Frequency Restoration Reserve (mFRR) capacity markets, focusing on the Nordic market setup. This market enhances the reliability of the Nordic system by ensuring availability of mFRR. In contrast to today’s market setup which only accounts for capacity costs, the proposed market clearing mechanism accounts for both expected energy activation costs and mFRR capacity costs. The market clearing problem is formulated as a two-stage stochastic Mixed Integer Linear Program (MILP). To efficiently solve the resulting optimization problem, we introduce a nested decomposition algorithm that combines the Benders Decomposition (BD) and Surrogate Absolute Value Lagrangian Relaxation (SAVLR) methods. For practical implementation, input data scenarios are generated using day-ahead (DA) price data modeled by a Bayesian Neural Network (BNN). A real-world case study in Sweden demonstrates that the proposed mechanism can reduce daily mFRR costs by 600-14,500 €, due to reduced energy activation costs. Simulation results further show that the nested decomposition algorithm converges to a near-optimal solution more quickly than standard BD.