About OpenCorporates network data

We are used to thinking of a company as a single entity. This may be true for small businesses, but in today's global economy, a large corporation usually exists as large network of interconnected companies spread around the world.

Control of one company by another is usually exercised by share ownership: this page shows that Goldman Sachs New Zealand Holdings Limited owns 100% of the shares of Goldman Sachs New Zealand Management Limited.

Sometimes these corporate networks are very big. For example, we currently know of over 3,000 companies in the Goldman Sachs network. There are certainly many more companies than this which we don't know about.

Although it is possible to buy corporate network data from specialists, very little of this information is available as Open Data. Thanks to a grant by Alfred P Sloan Foundation, we are beginning to build a completely free and open database of network information. This part of our database is very small at the moment, but it will be growing very quickly over the next year.

At the moment, our network data is limited to three different sources.:

• Filings made by large domestic and foreign companies to the U.S. Securities and Exchange Commission (example)
• Banking data held by the National Information Center of the Federal Reserve System in the U.S. (example)
• Information about individual shareholders published by the official New Zealand corporate registry (example)

The best quality data is that held by individual corporate registries around the world. We believe that it is in society's interests for shareholding data to be available as open data, everywhere. That's why we are working with the World Bank Institute to monitor the transparency of corporate registries around the world.

We do not claim our data is free from errors: sometimes a company makes a mistake in its filing, and sometimes our automatic data importers have bugs. However, what makes us unique is that we aim to show a full provenance for every piece of data on which we base our inferences about corporate networks. For example, here is a list of all the sources for all the data we have about one company in the Goldman Sachs corporate network.

Our corporate network research is still at a very early phase. Over the next few months, we intend to:

• Vastly improve the quantity of network data that we import. This will include crowdsourcing tools to give general users the opportunity to contribute more, and better, data.
• Make our network graphs more beautiful, more informative, and easier to use
• Improve our network inferences - at the moment, for example, if we have one source saying Company A owns Company C, and other source saying Company A owns Company C via Company B, we will show all these relationships, when in fact we could just show the latter.
• Improve the performance and scale of the networks. At the moment we only show a network up to three levels deep at a time. You can "recentre" any graph to move through the network, but this doesn't give you a full idea of the extent of any single network.

Are you an academic researcher or journalist who is interested in developing data and visualisations about specific corporate networks?

If so, get in touch. We are looking for partners to help explore this new, exciting, and vitally important set of open data.

Email: investigators@opencorporates.com

Twitter: OpenCorporates