Outsourcing is a common strategy for original equipment manufacturers (OEMs) to reduce costs and enhance service performance, yet it exposes firms to supplier-related risks. A key challenge is how manufacturers can allocate limited resources across multiple suppliers to balance the risk of underperformance against expected returns. While prior studies have applied portfolio theory to supplier development programs, they typically assume linear investment–return relationships. This study introduces a discrete, investment–dependent return function, commonly used in practice, to capture these dynamics and extends the framework to incorporate cooperative strategies among manufacturers. To address computational complexity, we develop a column-generation-based approach that significantly improves solution efficiency. The results provide practical insights for OEMs on optimizing investment decisions under uncertainty in complex outsourcing environments, highlighting trade-offs between engaging in new supplier development programs and maintaining existing contracts. A comparison with the classical Markowitz portfolio model further demonstrates the advantages of the proposed approach.