Client’s name / industry

Minderoo Foundation

Products / Services

Working closely with a number of global partners including Systemiq, Planet Tracker and others Neural Alpha lead the development of a novel plastics supply chain mapping exercise to help produce the inaugural Plastic Waste Makers index.

Result statistics

Working with large volumes of supply chain data, physical asset and ownership data Neural Alpha developed a unique supply chain model to aggregate several years of supply chain data into total production volumes and export flows for global plastics producers.

Key findings of the 2021 report included:

-      It’s a sovereign problem with more than 30% of the sector’s shareholdings by value held by governments

-      In 2019 just 20 polymer producers accounted for more than half of all single-use plastic waste globally and thetop 100 accounted for 90%

-      Twenty Institutional AssetManagers hold over $300bn worth of shares in polymer producers 

“That’s why the Plastic Waste Makers Index, prepared by the Minderoo Foundation and its partners, comes at such a critical time. As awareness of the toll of plastic pollution has grown, the petrochemical industry has told us it’s our own fault and has directed attention toward behaviour change from end-users of these products - rather than addressing the problem at its source!” Al Gore, commenting in the foreword of the2021 Plastic Waste Makers Index report

Objectives/Goals/Challenges

To turn raw, noisy and incomplete global plastics supply chain data into a robust model ready for data analysis and data science techniques to reveal key trends, risks and opportunities for improving sustainability outcomes.

Solutions

Neural Alpha employed our sophisticated data engineering and matching capabilities to take millions of raw supply chain records and transform the data into a robust dataset to link with waste and other datasets in order to produce a defensible and reconciled database for analysis.

Raw export and import volumes data was reconciled against UN Comtrade data to establish an accurate baseline prior to matching.  Using a novel constrained matching technique, we were then able to match and resolve 10s of thousands of different global company names to dozens of national companies’ registries’ legal entities. The legal entities were then extended into corporate groups to link to ultimate parents and Special Purpose Vehicles (SPVs) using ownership data and financial exposures then calculated using equity and debt holdings data.

Benefits/Results

Neural Alpha were able to deliver a MinimumViable Product (MVP) dataset for analysis and external audit and assurance in a matter of weeks enabling a reasonable period of data analysis and research to follow leading to the key findings of the report.