Determinants of Commodity Futures Prices: Decomposition Approach

Abstract

Developing models to analyze time series is a very sophisticated, time-consuming, but interesting experience for researchers. Commodity price component determination is challenging due to remarkable price volatility, uncertainty, and complexity in the futures market. This study aims to determine the components that drive the market price of commodity futures. This study utilized the decomposition methods, empirical mode decomposition (EMD), and variational mode decomposition (VMD), to analyze three commodity futures prices data: corn from agricultural products, crude oil from energy, and gold from industrial metal. We applied these techniques to decompose the daily data of each commodity price from different periods and frequencies into individual intrinsic mode functions for EMD and modes for VMD. We used the hierarchical clustering method and Euclidean distance approach to classify the IMFs and modes into high-frequency, low-frequency, and trend. Next, applying statistical measures, particularly, the Pearson product-moment correlation coefficient, Kendall rank correlation, and Spearman rank correlation coefficient, we observed that the trend and low-frequency parts of the market price are the main drivers of commodity futures markets’ price fluctuations. The low-frequencies are caused by special events. In a nutshell, commodity futures prices are affected by economic development rather than short-lived market variations caused by ordinary disequilibrium of supply-demand.

Situated Information Flow between Food Commodity and Regional Equity Markets: An EEMD‐Based Transfer Entropy Analysis

“…Notwithstanding, evidence from the empirical literature [8,9,[20][21][22][23][24][25]27] suggest that the flow of information and investor response to such information is not only timedependent but also varies across frequencies. It is instructive to note that frequency-domain analysis is a significant component for investors who operate at different time horizons [28][29][30][31][32][33][34][35][36][37]. Moreover, the stylised facts of fat tails and volatility clustering of financial time series introduce the complexity, asymmetry, and nonlinearity in the behaviour of market participants [20,34,36,38,39].…”

ARCH-GARCH Analysis: An Approach to Determine The Price Volatility of Red Chili

“…The existence of a commodity becomes a necessity that must be available every day to fulfill society's food consumption. In its development, agricultural commodities often experience price fluctuations (Antwi, Gyamfi, Kyei, Gill, & Adam, 2021;Kumari, Venkatesh, Ramakrishna, & Sreenivas, 2019;Nigatu & Adjemian, 2020;Nugroho, Prasada, Putri, Anggrasari, & Sari, 2018;Sativa, Harianto, & Suryana, 2017). Von Braun and Tadesse (2012) stated that natural factors often fluctuate agricultural commodities.…”

NFVLearn: A multi‐resource, long short‐term memory‐based virtual network function resource usage prediction architecture

“…One of those is applying a GNN to a directed graph 2,13,19,20 and the other is feature pruning using correlation coefficients. [21][22][23][24] GNN theory 20 is a popular choice among teams designing ML-based solutions in which directed graphs are involved. 19 This is especially true for the NFV research area, since concepts of nodes (VNFs), features (resource load), and links between those nodes through a directed graph (SFC) are a direct translation of the underlying concepts of the NFVI.…”

“…Some useful techniques to select input data features prior to ML model training to remove uncorrelated data while avoiding reducing the models' prediction accuracy can be found in the literature. One of those is applying a GNN to a directed graph 2,13,19,20 and the other is feature pruning using correlation coefficients 21‐24 …”

“…The other approach of interest involves applying correlation coefficients (Pearson correlation coefficient, Spearman rank correlation and Kendall rank correlation coefficient) to the input features of an ML model, then eliminating the input features with the lowest correlation coefficients since uncorrelated features will not improve the prediction accuracy of said models. Researchers in Reference 23 apply this concept to commodity price component determination, while the solution in Reference 22 applies in the context of a multivariate dependence framework. In an approach closer to our research area, the authors in Reference 24 use correlation coefficients to select highly correlated input features similar to those of the observed feature to train their LSTM mechanism.…”