This article develops a model for the calculation of the price of options in commodities using a parallel computing platform based on the CUDA architecture, with the aim of significantly reducing the computational costs associated with the simulation and financial valuation processes. The proposed model is applied to the analysis of hydropower risk under conditions of uncertainty, taking the Chili River basin as a case study.

The methodology used integrates Monte Carlo simulation techniques and valuation of real options using trinomial trees, allowing estimating the prices of the main commodity variables linked to water resources, specifically electricity generation, human consumption and agricultural use. The approach considers the dynamic and stochastic behavior of these variables, as well as the influence of climatic factors generated from a hybrid stochastic model.

Deploying the model in a GPU environment demonstrates a significant acceleration in compute times versus traditional CPU deployments, especially as the complexity of the analysis and the number of nodes in the trinomial tree increases. The results show that the use of parallel computing CUDA is an efficient tool for risk analysis and optimal decision-making in the management of water resources under uncertainty.