Changes

Jump to navigation Jump to search
Note see also: [[Governance Measures]]
====Institutional InvestorsNew Measures==== From Thomson-Reuters 13F filings (joined using CUSIP8 and year, where available):*iisharestot -- Total number of shares held by institutional investors*iisharestop5 -- Number of shares held by the five largest institutional investors*iitop5prop -- Proportion of shares held by the five largest institutional investors compared to all institutional investors*iicount -- Number institutional investors*blocktot -- Number of blockholders (institutional investors with >= 5% of shares) From COMPUSTAT Fundamentals Annual (joined using CUSIP8 and year, where available):*gp -- gross profit*ni -- net income*gpr -- gross profit ratio*npr -- net profit ratio*csho -- number of shares outstanding (in millions)*revt -- revenue (note that this replaces the old revt variable)*roa -- return on assets (ebitda/totalassets) *leverage -- total liabilities/total assets*totalassets -- same as the old at variable but from a new pull*interlock -- 1/0 CEO is on compensation committee*mktcap -- Market Capitalization calculated as closing price (prcc_c) x common shares outstanding (csho) Thomson-Reuters 13F filings in conjunction with COMPUSTAT Fundamentals Annual (joined using CUSIP8 and year, where both available):*iifractot -- fraction of shares held by institutional investors*iifractop5 -- fraction of shares held by top 5 institutional investors ====Old Governance Measures==== The following governance measures are listed under firm variables below:  agovgoldenparachute - 0/1 for the CEO has a golden parachute (many missing) agovlimitabilityactbywritconsent - 0/1 whether governance has limited shareholders ability to act by written consent (many missing) agovmajvoterequirement - 0/1 for whether board members are elected on majority or plurality, which matters for uncontested seats (many missing) agovresignrequiredonmajorityvote - 0/1 "Indicates that a director is required to submit his/her resignation upon failing to receive support from a majority of votes cast (which, typically, the board may chose to accept or reject)." (many missing)  I tried each of them in a FE panel regression of the type found in table 3, like: areg tdc116l car5totpub car5totpriv agovresignrequiredonmajorityvote i.year, absorb(ceofirmid) cluster(ceofirmid) In each case, the sample dropped to ~2k (from 41k), and there was no significance for ANY variable.
=====Reference Papers and their data=====
*Blockholders' data is reported by firm for the period 1996-2001: https://wrds-www.wharton.upenn.edu/pages/get-data/blockholders/
I pulled everything (8 vars) used the first one. See [[wrds]] for the full date range (1980-1 to 2020-9) as tdtpreprocessing. The file is:
===Restoring the old data===
acqi - current income from acquisitions in CSTAT
acq - cost of acquisitions from CSTAT
mktvalue mkvalt- market value
agovgoldenparachute - 0/1 for the CEO has a golden parachute (many missing)
agovlimitabilityactbywritconsent - 0/1 whether governance has limited shareholders ability to act by written consent (many missing)

Navigation menu