The Paris agreement is a game-changer. Now that world leaders have agreed – finally – to do something about climate change at the end of 2015, everything changes in terms of the way we exploit natural resources. To manage that agreement, we need an inventory of the world and everything in it – in open data. And, like any open data ecosystem, it will need millions of eyes on it. Why do we need it? What will it look like? What’s your role in it? Johnny West, founder of OpenOil, explains at Re:publica 2016.
For more about the talk, visit the Re:Publica Website here.
Oil geek question: what do ConocoPhillips, Halliburton and Schlumberger have in common? Well… they all share the fact that they have, or at least at one point had, Panama incorporated subsidiaries. So why is that, if Panama isn’t actually producing any oil, as the U.S. Energy Information Administration states on their website?
With the Panama Papers big in the news, this will probably not come as a surprise. And to be clear, setting up local subsidiaries in Panama is not illegal, nor is it unusual for extractive companies to open up incorporations across the globe – even where they are not operating directly. That becomes obvious through the thousands of documents available in public domain, in which oil, gas or mining companies disclose their vast networks of affiliates, including those in Panama, but also those based in one of the various other financial centers, like the Caymans or the British Virgin Islands (BVI). So unlike all of the stories coming out of the Mossack Fonseca leak, the aforementioned offshore links can be found in the companies’ own filings, which means that you can find them on Aleph, OpenOil’s corporate filings database. In fact, a quick search on Aleph will bring up 500 similar references to Panama incorporations of extractives companies, close to 1,500 references to those in the BVI, or 2,000+ on the Cayman Islands.
Establishing such links between companies, people, and jurisdictions is essential to anyone wanting to map out corporate networks – may it be investigative journalists, risk analysts, tax authorities, or anyone else interested in finding out about who multinationals are dealing with. In particular, however, they help to address two major issues in the extractive industries as we have outlined in our case-study on BP.
1) The first issue is the “Bad Guy Issue”.
Corrupt access to natural resources, when dodgy companies are granted licenses for lucrative oil and mining concessions because they are politically connected. Such connections could conceivably be demonstrated using public documents, but it is obviously challenging to do so.
Looking at some of the Mossack Fonseca stories, for example, you will see that corporate filings will help you to lay the groundwork. Such as for this one on how the Beny Steinmetz Group Resources (BSGR) acquired rights to a mining license in Guinea for $165 million, but soon after sold 51% of the rights for more than ten times the price. Ultimately it was through a number of leaks and a series of criminal investigations that the details of the deal could be reconstructed. But Aleph’s 2,000,000 documents already provide you with some relevant leads. Such as this document, that suggests a connection between BSGR and Onyx – a company that in return represented a crucial link between BSGR’s acquisition of the mining site and a $2.4 million payment to Mamadie Touré, one of the wives of former Guinean president Lansana Conté. Then there is also Vale’s announcement on having acquired a 51% stake in the same assets for 2.5$ billion, which should have already alarmed anyone who was aware of the original asset transfer to BGSR for $165 million.
Search for the name of a jurisdiction and “incorporated” in proximity to find out which companies are registered in the country.
Proximity searches for two companies, here BSGR and Onxy, can bring up leads to whether there is a connection between them.
2) Then there is a second issue, the “Sharp Guy Issue”.
Whether complex corporate structures allow multinationals to engage in “aggressive” tax planning. By “sharp guys” we mean the small armies of lawyers and accountants who are engaged in handling the billions that pass through the accounts of oil, gas and mining companies. And as has been well documented elsewhere, these guys are sharp, so that – in a world full of complex tax treaties – it is extremely difficult to prove any so-called “transfer mispricing”, or other means to shift profits out of resource producing countries into low- or no- tax jurisdictions, such as the above mentioned Caymans or British Virgin Islands… or Panama.
The question here is more whether governments in say, Africa, would act differently if they were able to see the whole corporate chain of the companies operating. Might they adjust their own taxation policies in light of that? Would they subject billions of dollars of tax-deductible costs submitted to their tax authorities to audit, or at least to more rigorous examination?
So while establishing such corporate chains can be powerful, it can also require you to spend a lot of time browsing through filings, only to find the one odd mention of a particular affiliate of interest. Companies operate globally, but they tend to report locally, meaning that there are many different places where you might have to search – the Canadian SEDAR database or the Australian stock exchange ASX to name just a couple. Fortunately, this is where two of Aleph’s strong points come in. First, Aleph helps you to search across various document bases simultaneously. Second, that you can fully text-search all of the two million available filings, allowing you to find text snippets deep-down in a 100-page long PDF – a service that many of the existing public databases do not allow you to do.
So keeping in mind a few general tips and tricks, there are many different searching techniques through which Aleph will help to map out company networks. To name one, you can choose to search “top-down” for subsidiary lists of a company group, such as listed companies are required to file in many jurisdictions, e.g. in the form of “Exibit-21” on SEC’s EDGAR database. Such filings will then provide you with the names of subsidiaries, their place of incorporation, and the equity held by the parent company – allowing you to follow a corporate chain from a holding company all the way down to an operating entity.
Alternatively, you might choose to search “bottom-up” by entering the name of a single affiliate of your interest and work your way up to it’s ultimate holding company. This works best if you type in the exact legal name of a company in quotes. And if your search brings up too many results for you to process, try to narrow down your search results by adding terms such as “subsidiary”, “owned” or “interest” via a proximity search – all this of course depending on your particular use-case.
While we are working hard to improve Aleph’s functionality and scope – so that you will soon be able to filter results by date, filing type, or company – you should of course also consider other information sources, such as OpenCorporates, or even the companies’ own websites. For all the oil geeks, however, it might be interesting to know that we are also establishing Aleph’s API – which for example will allow you to not just search for one entity, but many simultaneously – such as the 40,000 odd companies mentioned in the Panama Papers. How exactly you will be able to do so, will be described on our github page – so watch this space!
Search for subsidiary lists, such as the SEC required “Exhibit 21”, to research a company’s subsidiaries and their place of incorporation
Search for the name of a company and “subsidiary”, “owned” or “interest” in proximity to find out about the company’s affiliates.
This year’s introduction of mandatory disclosures in France and the UK will bring about a considerable amount of reports listing extractive companies’ payments to governments.
The mandatory disclosures promise increased transparency, however, we are only at the beginning of a debate on how to best make use of the new data. One idea has already become apparent: comparing the mandatory disclosure data to EITI figures in order to find irregularities.
In the context of our involvement in Publish What You Pays “Data Extractors” programme, we have simulated how such a comparison could look like and formulated a few first thoughts, as detailed in this document. In this, we try to assess how these different sources referring to the same project relate to each other, in fact, how comparable they are after all. In the following, we would like to highlight a few aspects one has to take into account when comparing the two datasets.
Since the most actual EITI reports date from 2014, we needed to make sure the company reports were also covering that same year. This limited our comparison to the four companies in the Oil & Gas sector, that had both published payments to the government reports and that are operating in one of the few countries for which there already is a 2014 EITI report available. In total, we had six cases. The graph above represents the comparison between the EITI data (in blue) and the figures put forward by the companies (in red) on payments to government. In all cases, we found that the two reports had diverging figures. Deviations range from 0.84% (Mnazi Bay) up to almost 200% (Tullow in Rovuma Area 2&5). This begs the question as to why both reports fail to show the same results:
- Staggered Reporting Cycles The reporting cycles may diverge. We can see this, for instance, when we want to compare Wentworth operations with the matching EITI reports. These are Mozambique and Ghana. In Mozambique, both reports cover the same timeframe, January 1st, 2014 to december 31st. However, in Tanzania, the EITI report refers to a cycle from June 31, 2013 to June 31, 2014, whereas Wentworth covers January 1, 2014 to December 31, 2014.
- Project / Company Conflation Some reports relate to projects, other to companies. EITI reports often list payments by companies rather than by projects. However, there are companies that have only one project in the country, which they operate as part of a group. Statoil, for instance, is listed as a company in the EITI reports, and the figure for its payments in the Mozambique is higher than that in Statoil’s report. One reason being that Statoil has only 65% interest in the project. Without the other shareholding companies’ reports, we cannot have a full comparison with the EITI report.
- What is a ‘payment to government’? There are differences in what counts as ‘payments to government’. We went through the reports and looked at the items listed as payments. Tullow’s report in the Deepwater Tano project in Ghana for instance includes a range of items that are not included in Ghana’s EITI report, such as VAT, local payroll taxes, withholding taxes or infrastructure improvement. It seems that these items are at the discretion of each EITI member state.
This exercise has shown one methodology to analyse the data around EITI reports and the new incoming data from the EU mandatory disclosures. Although such a comparison proves to be challenging, it might help to prepare a thorough analysis that could lead to a more transparent and standardised reporting, as well as, in the best case scenario, helping to find the missing millions.
We invite you to join in on two workshops organized by the OpenOil team where we will invite participants to learn more about open data on the extractive industries and how to search and dig out information from the Aleph search engine.
“Panama Papers: Using Open Data to track Big Oil”- Berlin Workshop, Tues. April 26th from 7pm to 9pm.
With the Panama Papers big in the news, we want to invite you to join us to explore, learn about and investigate into the global offshore tax industries. Using open data to track Big Oil, we will investigate how companies use tax havens to shift profits away from resource producing countries, use nominees directors to hide their real owners and obtain corrupt access to lucrative oil, gas and mining licenses.
We would like to showcase the uses of open data in this context using the open tool Aleph – OpenOil’s corporate filings database.
The workshop will take place from 7pm to 9pm at the Open Knowledge Foundation offices, Berlin, and will feature speaker finance activist Brett Scott, author of “The Heretic’s Guide to Global Finance: Hacking the Future of Money”
We invite data driven journalists, transparency and open data activists to join us!
When: 7pm-9pm, April 26th
Where: Open Knowledge Foundation, map
“Finding the Data Hiding in Plain Sight” -London, Fri. April 29th from 1pm to 5pm.
We invite you to join us to learn and explore how open data can track Big Oil. Using investigative tools and sector specific language we will showcase how researchers, analysts, journalists and anybody interested in the industries can quickly find the relevant from the trivial in big data.
Aleph by OpenOil will be featured. It offers open data from the extractive industries and has over 2 million corporate filings that exist in the public domain. The database not only automatically updates itself, providing an up-to-date record of corporate filings across the globe, but it is also indexed and fully searchable.
The workshop will take place from 1pm to 5pm at the Publish What You Pay offices, in Lambeth, London, and will feature will feature speaker finance activist Brett Scott, author of “The Heretic’s Guide to Global Finance: Hacking the Future of Money”
When: 1pm-5pm, April 29th
Where: Publish What You Pay (TBC), map
7-14 Great Dover Street,
London, SE1 4YR,
Looking forward to seeing you!
Every week thousands of pages are filed by extractive companies to stock exchanges across the globe, containing valuable information on oil, gas and mining operations worldwide. Through such disclosures, investors, regulators and researchers can update themselves on any listed company’s financial situation, the economics of particular oil or mining projects, newly signed host-government contracts, changes in directors or shareholders, and lots more. We at OpenOil for example have already successfully searched through them to find precious documents on multiple occasions, such as the hundreds of host-government contracts available on our contract repository.
But so can you… by using our search-tool Aleph. Here we would like to share a few tips and tricks on how to best make use of Aleph. In fact, we are only discovering the power of the search tool ourselves and will put out a series of blogs in the coming weeks to describe particular use-cases, such as comparing interest rates of intra-group loans. In the meantime, we encourage you to explore the thousands of documents already. And to best do so, you should keep the following tips and tricks in mind, so as to find the information that is relevant to you.
1) Be exact when selecting search terms
When using Aleph, the selection of the right search terms is crucial. By typing in any series of words into the search bar, you will tell Aleph to list all the documents in which there is an exact match of your search terms. It therefore makes a difference whether you type in the singular or plural of a word, use the British or the American spelling, search for a noun or an adjective, and so on… For instance, the search term “confidentiality” is going to list all corporate filings that contain exactly that term, but not those documents that contain variations of the word, such as “confidential” or “confidentially”. The * symbol can help here: it will find all words beginning with a certain prefix, so that confidentia* will match confidential, confidentially, or confidentiality.
Listing all documents that contain words starting with the prefix “confidentia” and those with alternative endings, such as “confidentiality” or “confidentially”
2) Narrow down search results
If a search term results in too many documents for you to read through, try to narrow down Aleph’s search results. There are three ways to do so: first, by specifying just one of the document bases. Let’s say you are only interested in EITI reports, you can select the document base on the right side menu and it will only list those documents within that document source, that contain your search term.
The second way to narrow down search results is by adjusting the search terms. Try for example to search for “confidential*” AND “agreement” (make sure you write AND in capital letters), which is going to list all documents that contain the two words. Or even better, search for an exact term in quotes, such as “confidential” AND “production sharing agreement”.
If you still didn’t find what you are looking for, try a proximity search: the third way to narrow down results. Such a search will list documents that contain multiple search terms, but only where your search terms appear in proximity of each other. You can do so by adding ~10 or any other number at the end of quote, e.g. “confidential agreement”~10, which will tell Aleph to list those documents that contain both “confidential” and “agreement” in proximity of 10 words of each other.
Listing all documents that contain variations of the term “confidential*” AND the exact quote “production sharing agreement”
Listing all documents that have all words inside the quote in proximity of max. 10 words
3) Use Aleph to find the particular, and not the common
While Aleph’s API allows for statistical-level analysis, we believe one of the strengths of Aleph’s frontend lies in finding the hidden – a mention somewhere deep down in a PDF. It is therefore important that you test out search terms that are unique to the topic you want to research. For example, search for the name of a particular oil block or mining site, rather than for a country, e.g. “Bulyanhulu” (the name of a gold mine in Tanzania) as opposed to “Tanzania gold”. Another way to find the particular, is by searching for a particular technical term. Let’s say you are interested in commodity trading, then you might want to search for terms such as “cargo” and “API” (a term to specify the quality of oil cargos), – and not for generic terms such as “commodity” or “trading”. The same principle also applies to company names: rather search for the full legal name of the subsidiary that is running the operations in a particular country, as opposed to the names of a group, e.g. “Bulyanhulu Gold Mine Limited” (the subsidiary operating the gold mine), and not “Acacia” (it’s ultimate parent company).
Search for the name of a particular oil, gas or mining project, such as the “Bulyanhulu” gold mine in Tanzania
Search for the full legal name of a subsidiary running the operations in a particular country, rather for the name of a company group
4) Use the language of companies
Corporate filings are full of legal jargon. It is therefore helpful to pay close attention to the language companies use in such documents, so as to abstract out relevant search terms.
Let’s say you are following a company-government dispute in a particular country, and you want to see whether there have been any precedents that could indicate the outcome of the dispute. A first search of the term “court settlement” will already lead you to a few relevant company announcements, such as this one. Carefully reading through these first leads however, you will come across a series of terms that are frequently used in the context, such as “out of court settlement”, “cease all legal action” or “compensations for an amount of…” – all of which will allow you to expand the scope of relevant documents, in this case in regards to court settlements.
Once you found a document relevant to your research, try to adopt the language to find similar documents
5) Be creative and playful
Last but not least, be creative. Play around with different search-terms, methods and use-cases. And keep in mind, many of the biggest data success stories have been born by coincidences…