Answers are grounded in this paper and cite their sources.
Financial forecasts often combine numerical assumptions with qualitative assessments whose influence remains difficult to audit. Evidence-Weighted Monte Carlo (EWMC) links a five-point scorecard to bounded changes in downside and upside scenario probabilities. The original experiment evaluates six synthetic evidence conditions, eight procedures and 100 replications per condition. Under informative evidence, EWMC reduces the continuous ranked probability score (CRPS) from 6.119 to 5.836. Simpler procedures achieve lower errors in several conditions. The revision separates relative source weighting from total support through a strength attenuation rule and evaluates 16 additional stress conditions with 100 replications each. A rolling calibration and suspension procedure reduces reversal CRPS from 6.991 for the original frozen rule to 6.229, compared with 6.156 for the numerical baseline. Under stable informative evidence, the same safeguard increases CRPS from 5.885 to 5.976. An exponential-tilt comparator supplies a restricted relative-entropy benchmark. A retrospective application uses Apple’s dated 2020 revenue guidance, a subsequent management warning and the reported quarterly outcome. The example demonstrates implementation with real disclosures, with no prospective or independent coding validation. The contribution consists of an auditable procedure, explicit safeguards and reproducible failure tests. The results establish conditional computational behaviour. They provide no general evidence of superior corporate cash-flow forecasts.
No takes yet. Share an insight, caveat, or question.
Alessio Faccia (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: