How to account for correlation in price models for commodities?
On this fifth webinar of the Forecasts with Uncertainty Webinar Series we will present how to estimate and forecast prices of groups of correlated commodities - for example, petroleum and its refined products or soy bean, oil and meal.
Correlation adds an extra layer of complexity when creating realistic forecast models, especially for commodities. We will use historical data to fit, calibrate and simulate correlated time series models for commodities, and discuss some pitfalls and practical insights on modeling.