metadata info for edgelist11.csv

download a compressed file here

csv structure: 16685 rows / 9 columns

columns

  1. Bank_code : 5-digit string, kind of code name for the bank, but specific for this 2011 dataset
  2. Bank_country : name for the country of the bank
  3. Bank_name
  4. Country.label : full name of the counterparty country. 'Correction' stands for total exposure correction (see below)
  5. Exposure integer: code for exposure type (as reported in 'code_information.csv'). 7000 stands for a correction term, inserted to match Declared total exposures with the sum of single Declared exposures (over countries and exposure types)
  6. Exposure.label : full name of the exposure
  7. Country_Asset : full name country _ full name exposure. Redundant with respect to col 4(5) and 6(7), but necessary to build graph BANK --> COUNTRY_ASSET. This is done ONLY for Exposure in Sovereign and Institutions. For Corporates, Retail and so on, country division is discarded. Of course, one can reconstruct it via columns 4(5) and 6(7)
  8. mlnEuro float: exposure value in millions of Euro
  9. tot.exp float: total exposure as Declared (and also as summed, once introduced the Exposure=7000 / Country=1000 correction. see above)