Advanced Positioning, Flow, and Sentiment Analysis in Commodity Markets, 2nd Edition

CHAPTER 7Decomposing Trading Flow and Quantifying Position Dynamics

Chapter objectives

• 7.1 Decomposing Net Positioning – The Methodology
• 7.2 Decomposing Long and Short Positioning – The Methodology
• 7.3 Who Has Been Driving Prices?
• 7.4 Who is Trading With Whom?
• 7.5 The Impact on Prices
• 7.6 Speed of Positioning Changes

To understand the explanatory power that changes in the positioning of individual long and short trader categories have on price, a Shapley–Owen decomposition of the regression of changes in price against changes in positioning can be used. This provides some indication of which trader groups have been determining prices on a commodity-specific basis over time. By extension, Shapley–Owen decomposition can also be used to understand the explanatory power that changes in the positioning of individual trader categories have on other trader categories. This provides an indication of the trading flows between groups.

A Shapley–Owen decomposition is needed to derive the explanatory power, or the contribution to R-squared for each variable in a regression, as this is not otherwise possible in a multiple regression due to the interaction effects between the variables. In a multiple regression, only the R-squared of the overall equation is calculated.

In combination with the betas from the multiple regression, the overall approach also provides meaningful insight into potential price changes, by factoring into account the trading flows and the price sensitivity. It is ...