Hidden Power of the Big Three? Passive Index Funds, Re-Concentration of Corporate Ownership, and New Financial Risk by Jan Fichtner, Eelke M. Heemskerk, Javier Garcia-Bernardo http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2798653
The raw data was downloaded from Orbis, containing the following fields for all US listed companies:
- 'Company name'
- 'Country ISO Code',
- 'NACE Rev. 2 Core code (4 digits)',
- 'Operating revenue (Turnover) th USD Last avail. yr',
- 'Number of employees Last avail. yr',
- 'GUO - Name',
- 'Ticker symbol',
- 'Stock exchange(s) listed',
- 'Shareholder - BvD ID number',
- 'Shareholder - Direct %',
- 'Shareholder - Total %', 'BvD ID number',
- 'Current market capitalisation th USD',
- 'Shareholder - Name',
- 'Total assets (last value) th USD',
- 'Type of entity'
- Non-US exhanges and private US exchanges. The exchanges that were remaining were: "'NYSE MKT','NYSE ARCA','NASDAQ/NMS (Global Market)','NASDAQ National Market', 'New York Stock Exchange (NYSE)'
- Private equity firms and Funds. We added JPMORGAN CHASE & CO (BvD ID US132624428) by hand since Orbis has mistakenly classified this very large US bank as a Private Equity firm.
- US041867445 (State Street Bank and Trust Co) because this subsidiary of State Street acts as a custodian, holding the shares for the ultimate owners.
- Public ownership (many small ownership combines) and owners whose ID start by ZZ (no people or companies).
- The largest ownership percentage between direct and total ownership (null values considered zero).
- The sum of the ownership percentages of the Big Three
- The position of that sum among all the shareholder, with the code 1 = largest, 2 = second largest, 3 = third largest, 4 = rest.
- Big 3.ipynb: ipython notebook containing the code used
- nodes.csv : Files with the companies and the position of the Big Three among the shareholders.