Using PLS Methodology For Understanding Commodity Market Behaviour And Improving Decision-Making Process
After a long and unexpected financial crisis, a period of rapid growth in China and emerging markets, unstable commodity
prices behavior, the commodity trading sector confronts a very different landscape, now. In this context, our paper aims to
better anticipate commodity market behavior in order to help the commodity investors to improve their decision-making
process. Using the PLS methodology (partial least square) and fundamental drivers of commodity market price behavior
(macroeconomic and specific indicators), our research proposes to elaborate a rational model to better understand and predict
the commodity market movements. Our selecting research methodologies enables to extract the most relevant factors, key
drivers, of commodity market behavior and seem to be able to capture a substantial part of systemic market risk.
Furthermore, we use bootstrap techniques to identify the optimal model for market behavior. Our results are validated using
widely measures used in PLS literature such as: AVE and composite reliability for the outer model validation and R-square
and redundancy index for the inner model validation. The empirical results, the path coefficients and the high reliability
score, come to confirm the validity of the proposed PLS model and its contribution in assisting the governments and
investors in their decision-making process to improve commodity market stability and efficiency.
Key Words- commodity market behavior, key drivers, SEM, PLS model, crisis, decision making