Open AmritSd opened 1 year ago
Data was pulled from Securities Daily in Compustat. It's 8gigs!
We've switched to the CRSP price data because it has end of period returns (daily adjusted) and delisting variables.
In CRSP: identifying companies by permno seems to be the better option because it identifies individual share classes whereas permco identifies companies.
Source: https://libguides.stanford.edu/c.php?g=559845&p=6686228
(removed comment)
Permno: Individual company share class identifier ticker: Ticker exchcd: Exchange code
siccd: SIC code (To take financial companies out) trdstat: Trading stat A: Active, H: halted, S: sus, X: unkown prc: Closing price (-ve if bid-ask spread used to calculate prices) If zero then data is missing (so forward fill) ret: Closing period return
dlstcd: Delisting code
dlret: return after delisting (A if still active, S if suspended)
We need cusip as well, since we need merge of cusip for institutional data
Pulled data from CRSP>Annual Update>Stock/ Security Files>Daily stock File The date range was 01-2005 to 12-2021 The following columns were requested
Number of shares outstanding (Shrout) was also requested since it is required for calculation ownership %.
We would need daily price OHLC or returns. Indication of splits, delistings, and exchange changes
Edit : 2/3/23:
Taking the following columns from compustat Date range (Jan 2005 - Dec 2021)
To account for splits do prccd / ajexdi
To calculate cum. returns do
p1 = (prcdd / ajexdi) trfd p2 = (prccdd (t-1) / ajexdi (t-1)) trfd(t-1)
r = p1/p2 -1