5. Risk metrics derived from MCS: VaR and sensitivity Beyond the probability P(NPV>0), the quantile Value at Risk (VaR) of the loss distribution has been transferred from finance to project evaluation to report tail risk in terms understandable to boards of directors and investors. Finance texts and academic notes describe its definition, computational methods (historical, var-cov, MCS), and limitations (it does not report severity beyond quantile, it requires back testing), offering criteria for its responsible use in capital budgeting. In mining, project VaR and bottom-up approaches have been proposed to quantify riskat-source and integrate cascading risks. (Malz, n.d.) 6. Recent trends: data integration, ML, and advanced analytics Studies from 2022–2024 show a convergence between MCS and data-driven decision making: • ML + DES models to predict production from IoT data, with DES fed by learned distributions (e.g., PSO-SVM for cycle times), validating P10/P50/P90 predictions against actual logs. (Park et al., 2023) . • Continuous cash flow with MCS for gold projects, compared against DCF, showing changes in NPV distribution and driver sensitivity (10,000 iterations with @similacion 5.0) (Varela, 2025) • In operations, adoption of binomials for availability and productivity (CM + trucks), with implementation in Excel + @Risk and analysis of cost-productivity trade-offs. (Mathey, 2022) 7. Synthesis of gaps and opportunities for contribution Even with the widespread use of MCS, the literature identifies areas for improvement: (i) explicit documentation of correlations and their impact on NPV/VaR; (ii) traceability of distribution assumptions (selection, adjustment, back testing) with historical data; (iii) articulation between variance contribution analysis and control/hedging policies ; and (iv) reproducible guides in standard tools (Excel/@simulacion 5.0) to scale from cases to project portfolios . IMPROVEMENT Transforming the Monte Carlo model into an intelligent decision support system, integrated with real data, AI techniques and an optimization component, capable of guiding strategic financial decisions in real time.
RkJQdWJsaXNoZXIy MTM0Mzk2