Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
In this paper, we evaluate the usefulness of GARCH-class models in forecasting densities of crude oil futures from an investor perspective. Volatility forecasts are taken as the key inputs in calculating predictive densities. We find that FIEGARCH accommodating both long memory and asymmetric effect provides more accurate density forecasts than the other GARCH-class models most of the time. GARCH-based dynamic trading strategies perform significantly better than the benchmark of the static strategy even after accounting for the transaction cost. The gains of utility of GARCH-based strategies over the benchmark strategy are as high as 18%–20% p.a.