Commodities appear to be the most opaque among the asset classes of foreign exchange, fixed income, and equities. A few distinguishing factors in the commodity markets’ data sets cause them to stand out – namely, volume, variety, and velocity of data available. Data visualisation helps in transforming opaque data sets to meaningful decision-making tools. This article will focus on how visualisation can help us to see different and often striking aspects of the same data.

To illustrate effects of visualisation, let us consider a data set relating to the net long positions of non-commercials in the CFTC (Commodities Futures Trading Commission). To set the context, it is useful to know first that this data lags by a week, e.g., the positions that are reported in the current week contain data from the prior week. Second, non-commercials strictly refer to positions taken by any party other than merchandisers. Thus, these positions do not represent commodity hedges in production or consumption.

The charts below that show the same data in three different forms. The first chart (not including the 2013 data) tries to capture the relationship between the Z score of the long position and the long position of non-commercials. To arrive at the Z score, existing data is transformed in such a way to examine how elastic it is from the mean of the dataset. One can see that as the Z score exceeds 2, the long positions increase but the relationship diverges. I leave it to the reader to arrive at a corresponding interpretation, but the overarching idea is that transformation and visualisation of variables is a powerful tool. The second chart shows that over the years, 2013 is an anomaly. This chart shows us the reason why the second chart did not include the 2013 data. (Note: Z score is standardised variable; w1 to w52 are weeks from January to December).

Then there are charts that can show long position of non-commercials over time, and in doing so, the seasonality of the positions. Long positions drop in the summer months and witness an increase in the winter months.

Data analysis in the commodity markets has unduly received considerable flak from professionals dedicated to this business. The most often cited reason is a lack of clear evidence that data analysis is a credible decision making tool. On the contrary, visualisation holds the key to creating a stronger acceptance of data analysis as a critical tool. To do so, visualisation needs to appeal to its consumers by illustrating new insights afforded by rigorous analytics. While the math behind number crunching is important, it’s just as important to be able to see the results in an illuminating manner.

The take away here is that as much as analysis is a science, the ability to bring ultimate meaning to data is an art.

Going back to our charts, the conclusion from a parallel visualiation of all three is that if Z score is > 2.0, and it is winter, non-commercial longs tend to increase That being said, I would hope the reader would healthily debate this interpretation!

(The writer is based in London and is the founder and Managing Director of OpalCrest (www.opalcrest.com). )

Related Topics
comment COMMENT NOW