2089 Macroscopic-financial (project value): This approach uses aggregate variables (metal price, recovery, mine and plant cash costs, CAPEX) to construct simulated cash flows and derive the NPV or IRR distribution. This is the approach your paper adopts; recent literature shows specific applications in gold mining projects and continuous flow models, confirming that MCS improves the robustness of valuation and risk assessment compared to DCF. (Sarsoruo et al., 2019) 4. Modeling practices: distributions, correlations and tools Selection of distributions. In finance and mining costs, Normal, Lognormal, and Triangular distributions are common, depending on the nature of the data (e.g., prices and multiplicative factors are usually modeled lognormally; costs/times with limited information are modeled triangularly). There are guides and educational resources that show how to simulate lognormal and triangular distributions, and when to prefer one over the other based on skewness and tailings. (Lampe, 2015) Correlation between inputs. The appropriate correlation of variables (e.g., price-exchange rate, grades, recovery, interrelated costs) is critical; treating inputs as independent can over or underestimate the variance of the NPV. In the literature on multivariate correlation matrices applied to finance and infrastructure, correlation matrices, multivariate matrices, or copulas are recommended to preserve dependencies and improve the consistency of scenarios. (Zavala et al., 2025) Tools. In professional practice, the use of @simulacion 5.0 integrated with Excel stands out for defining distributions, correlations, and running thousands of iterations, with outputs such as histograms, cumulative curves, sensitivity tornadoes , and dashboards for stakeholders. (Varela, 2025) 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 risk-at-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 .
RkJQdWJsaXNoZXIy MTM0Mzk2