Forecasting feeder cattle basis using Bayesian model averaging
Payne, Nicholas D.
MetadataShow full item record
Accurate basis forecasting is an important tool for producers and consumers of agricultural commodities in their price risk management decisions. However, the best performing forecasting model in previous studies varies substantially. The best forecast also differs with respect to commodity and forecast horizon. Given this inconsistency, a Bayesian approach which addresses model uncertainty by combining forecasts from different models is taken. We find that model performance differs by location and forecast horizon, but the combined forecast often performs favorably compared to the best regression models. However, except for very short-horizon forecasts, the simple moving averages have lower out-of-sample forecast errors than the regression models. We also examine using a basis series created using a specific month’s futures contract as opposed to the nearby contract and find that regression models outperform naïve forecasts and the average forecast for longer horizons.