For fund managers, early and large redemption may bring liquidity risk, so it is necessary to finance in advance. In order to reduce the cost of financing and meet the demand for redemption, a mixed integer quadratic constraint programming model is established, which subjects to the actual business constraints, and the uncertainty of financing cost fluctuations is quantified as the variance of the cost. The model minimizes both financing costs and risks. When the risk is not considered, the model degenerates into a deterministic optimization proposition. A simulation example is provided by financial institutions. The optimization results can take into account both the optimality of decision-making and the solving speed, which can meet the needs of the actual business. By comparing the decision schemes obtained by the deterministic and uncertain models, the methods of reducing financing risk are analyzed and summarized. A sensitivity analysis is carried out on the upper limit of variance, and the results further support the existing opinions.