Data Driven Decision Making for Commodity Management
Although the share of primary commodities in global output and trade has declined over the past century, fluctuations in commodity prices continue to affect global economic activity. For many countries, especially developing countries, primary commodities remain an important source of export earnings, and commodity price movements have a major impact on overall macroeconomic performance.
In general, the agreed consensus based on models is that growth in overall global demand for commodities will probably reduce as population growth slows and developing economies mature, although demand for some commodities is likely to rise as they tend to be used as inputs for new technologies as well as the challenge to the world in general to meet the transition to cleaner energy.
These shifts in turn will have major implications at a macro economic level for growth and poverty reduction especially in developing economies, almost two-thirds of which are commodity exporters.
In order to predict the price of agricultural commodities — we looked at building a commodity price prediction tool at — a country specific level — for tea auction prices in Sri Lanka (which can be extended to other types of commodities) which makes use of sentiment data by social listening along dimensions ranging from substitutes, technology and process improvements to economic sentiment of key countries exported to, along with a combination of weather and macroeconomic data to predict the weekly tea auction price and have been able to provide a very high degree of accuracy after being back tested on years of historical data.
The commodity price management tool can be used to anticipate future tea prices and take interventions as necessary by tea exporters in order to protect their revenue streams as well as identify disruptions in the top export markets through social listening.
A snapshot from the dashboard is given below.
