In this paper, analysts Emily Sundqvist, Doris Rivadeneira, and Lazarin Lashkov provide an extended analysis of skewness strategy, applied to commodity future prices. Considered commodities include the soft commodities Corn, Cotton, and Soybeans, and hard commodities Gold, Silver, and Crude Oil, while two strategies are reviewed.
The first strategy is based on the level of skewness, whereas the second depends on stop-loss and take-profit levels. A number of parameter values are tested with each strategy in order to find the most profitable combinations for two different time periods: 2000-2005 and 2010-2015. These best-performing combinations are later applied to the entire dataset, 2000-2021, and the two strategies are compared to each other and to the most standardized trade which is used as a benchmark. This trade means opening a position in the beginning of the dataset and closing it in its end.
The results lead to the conclusion that the first strategy performs better than the second one, but both of them perform worse than the benchmark. This implies that their usage in the future is unlikely to provide promising results. Furthermore, this study will serve as a basis for future research, where Machine Learning models will be applied to develop an increased understanding of patterns found in the trading of commodity futures.