edwinhu

proc sql;
    create table out.sp500 as
        select distinct a.permno, a.date,
        a.permco,
        c.ncusip as CUSIP,
        c.exchcd,
        c.tsymbol,
        c.shrcls
        from crsp.msf a,

edwinhu

# curl to get all of the zip files of PDFs
curl https://www.sec.gov/foia/docs/adv/formadv_part2_[1-112].zip -o "/data/hue/adv2/formadv_part2_#1.zip"

# list files
unzip -l formadv_part2_1.zip
# awk extract the part I care about
unzip -p formadv_part2_100.zip 103705_325511_1_20200131.pdf | pdftotext - - | awk 'BEGIN{IGNORECASE=1};/^item 11/,/^item 12/'

# build an index
for f in *.zip

edwinhu

extract.bife <- function(model) {
  s <- summary(model)
  names <- rownames(s$coefmat_beta)
  co <- s$coef[, 1]
  se <- s$coef[, 2]
  tstat <- s$coef[, 3]
  pval <- s$coef[, 4]

  aic <- s$AIC
  bic <- s$BIC

edwinhu

/* The goal of this code is to identify all gvkeys in Compustat,
   and to estimate approximate start/end dates for these gvkeys */
proc sql;
    /* All NA Companies */
    create table comp_name_range as
        select *
        from
        /* Company files contains all NA gvkeys */
        (select gvkey, ein, cik,
        UPCASE(COMPBL(PRXCHANGE('s/[\W_]+/ /',-1,conm))) as conm,

edwinhu

/*******************READ ME*********************************************
* - Macro to download and parse the Fama French Factors -
*
* SAS VERSION:    9.4.0
* DATE:           2014-03-06
* AUTHOR:         eddyhu at the gmails
*
****************END OF READ ME******************************************/
%macro GET_FFDATA(file=F-F_Research_Data_Factors,
                  dsetout=ff_factors,

edwinhu

/*******************READ ME*********************************************
* - Macro to Helmert transform (forward mean-difference) panel data -
*
* SAS VERSION:    9.4.0
* DATE:           2013-05-05
* AUTHOR:         eddyhu at the gmails
*
****************END OF READ ME******************************************/

%macro HELMERT(dsetin = &syslast., 

edwinhu