What is a Commodity Filter?
A Commodity filter is a filter that allows the user to filter your tier-n data based on some value chain structure. After filtering, only those suppliers that are relevant for the selected value chain are included in the tier-n graph.
For example, here we have the value chain structure of aluminum cans for drinks.
Applying the commodity filter Aluminum Cans to some collection containing relevant suppliers, only such suppliers that supply the cans themselves, the sheet needed for producing the cans, and any material in between all the way up to the bauxite mine will be included in the resulting tier-n graph.
What problem does a Commodity Filter solve?
Supply chains are big and going several tiers down in the supply chain means a lot of data. For example, the company Ball Corporation, manufacturing aluminum cans, has hundreds of thousands of suppliers in their supplier network when extending the analysis to tier 5. Many of them are irrelevant for most use cases because many companies are included that have nothing to do with the can manufacturing. The problem is especially pronounced in case large conglomerates with many business lines are encountered somewhere in the supply chain.
Applying the above-mentioned commodity filter for Aluminum Cans reduces this number to mere thousands. This is essential for example for exposure analyses, because now exposures will only be displayed for companies that actually supply something relevant for the commodity in question.
How does the Commodity Filter work?
The commodity filter can be applied to a collection in the Network page or the Tier-N Explorer. In case a filter is already applied in the Collection Details, it cannot be reapplied elsewhere.
Components of the graph
The filter itself has 3 components
Commodities considered: The user can choose using a tick box one or many commodities that are relevant for the suppliers in their respective collection.
Suspected suppliers: The filter matches relevant suppliers using two criteria
Actual suppliers: Those suppliers who have shipped the relevant HS Codes in this specific supply chain. By default, only actual suppliers are selected. Choosing only actual suppliers is equivalent of excluding other data sources apart from Customs data form the analysis because the other data sources do not contain specific information about the products.
Suspected suppliers: Those suppliers who in general do business in the relevant HS Codes but where there is no evidence of such shipments in this specific supply chain. This needs to be enabled in case the user wants to use the filter and use other data sources apart from Customs data. The suspected suppliers are selected based on the commodities assigned to them that are visible on their profile pages.
Logistics providers: We often encounter logistics providers in the tier-n graphs in Prewave. These, however, often introduce noise to the graphs as they might appear in the data only as logistics intermediaries, therewith introducing all their other customers into the network as well. Logistics providers are be default excluded. The logistics providers are selected based on the industry Transportation, Logistics assigned to them that is visible on their profile pages.
Rules for traversing the graph
Once the filter has been applied the tier-n graph is built following the commodity structure. Here is an example for aluminum cans.
The following rules apply:
Don’t go backwards in the value chain: When we evaluate whether a new supplier should be included in the graph or not, the supplier must always supply either the same or earlier stage raw material as the previous supplier. For example, if we have a tier-1 aluminum cans supplier, we only accept suppliers who either supply the cans themselves (a potential trading company with no own manufacturing) or some raw material from the above graph, such as aluminum sheet. Steps can be skipped because some companies might be vertically integrated, i.e. an aluminum sheet company might not buy aluminum ingots but might smelt them themselves and therefore buy alumina as in input instead.
Respect suspected suppliers and logistics provider selections: Depending on what was defined in the filter, these considered or not. For example, if logistics providers are excluded any supplier with the industry Transportation, Logistics will not be included in the graph.
Summary
A Commodity filter is an essential filter in reducing noise when working with tier-n data. It helps you pinpoint the essential out of a large dataset and makes the vast tier-n data manageable. It is highly customizable for your needs.