April 8, 2016 Hatem

Understanding #PanamaPapers with #OpenData

Who did not talk about #PanamaPapers yet ? Nobody, right ?! Now since only few medias companies have access to the leak, I was wondering how important is this leak ? To what can we compare all this huge amount of data ?

Let’s go for a short visit to Panama, a virtual visit obviously, and see what can we find out there ! First, the Company register is publicly available, there is a website for tenders purchase, trademarks, and even the controller documents. That’s already great for a tax haven ! The leak include tons of documents including email, communications between only one agent which is MOSSACK FONSECA, and many officials among others.

Even if the previously mentioned websites do not offer data in open format, they do offer all the data we need! The good news is that you can find all these data downloadable in bulk and in CSV format at PANADATA.


Download the data in CSV format, see the export link in the top menu. Now let’s load this data into R and see what we can learn from it :


The database include more than 600K records, and you can see already from the summary that there are just 11773 records related to MOSSACK FONSECA. So we are talking about less than 2% of company register records which include not only creation but also records for companies dissolution and others. I wanted first to make sure that there are no others “Mossack Fonseca” inside the register, then what’s the value of these companies compared all others companies values.

First, looking for the agent name without case sensitive search give exactly 15319, which is about 2.5%. We are still talking about just 2.5% of total records in the country, while MORGAN Y MORGAN hold more than 8%. Even if we isolate companies created by MF agent, we’ll find about 2.4% of companies created in the whole database.

We can isolate MF transactions this way :

>mossack <- sociedades[grep('mossack fonseca',sociedades$agente,ignore.case = TRUE),]

In term of capital which is not really a significant information, companies created by MF represent only 0.28% of total investment in other companies in the whole country. So if we look at the summary of the isolated MF dataset we can see this :


We can see that most companies are created with the minimum capital of $10K and they are located mostly in the Panama province.

So finally, #PanamaPapers could be important and include tons of data, but if we compare it to the whole country database we can see how irrelevant it is ! We are missing only 97.6% of companies created in Panama.

If a leak can lead to such amount discoveries, you can imagine the huge impact of #OpenData on transparency in the whole world. Now what do you recommend as tax haven ?

, , , ,

Leave a Reply

Your email address will not be published. Required fields are marked *

(HBY) Consultancy