Q: What is WRDS?
A: Wharton Research Data Services (WRDS), is the leading comprehensive, web-based, data research service used by academic, government, and non-profit institutions. WRDS handles all data management and provides the user with one location for research to a variety of databases through a common interface. Developed in 1993 to support faculty research at Wharton, the WRDS service has since evolved to become a common tool for research for over 370 institutions around the world. WRDS supports a global research community of over 30,000 users in 33 countries. This hosted data service has become the locus for quantitative data research and is recognized by academic and financial research communities around the world as the leading business intelligence tool. For more information on obtaining a subscription, please contact WRDS at [email protected].
Q: What is the source of WRDS data?
A: WRDS data is compiled from independent sources that specialize in recording specific historical data. These sources include: The Center for Research in Security Prices (CRSP), S&P Capital IQ, Standard & Poor’s DRI, Thomson Reuters (I/B/E/S — the Institutional Brokers Estimates System), Markit, FactSet, HFR, Dow Jones, the Federal Deposit Insurance Corporation (FDIC), the New York Stock Exchange Trade and Quotes (TAQ), and the Philadelphia Stock Exchange (PHLX). The Wharton School maintains distribution agreements for specific data sets with each distributor and stores the accumulated data on the Wharton UNIX and NT networks. The data can be accessed by any user whose school maintains a license agreement with the Wharton School and with the appropriate data sources.
Q: Where do I find the data I need?
A: The databases available on wrds (COMPUSTAT, CRSP, TAQ…) are listed on the left-hand side of the wrds homepage or in the “Data Set” drop-down list that appears at the top left of every page. After selecting a database link, a brief description appears in the main section and in the left column a list of file links is shown. Each link leads to a query page, where a description of variables and a query form are available. In the WRDS UNIX system, a consistent directory structure is used to access the WRDS data. WRDS data is found in the UNIX “/wrds” directory. Data from each distributor is found in the respective directory under “/wrds”; i.e., in the /wrds directory, there are directories for each data distributor (CRSP, COMPUTSTAT, DRI, FDIC, etc.).
Most WRDS datasets are available in both a binary format accessible via a FORTRAN or C/C++ program and in a SAS dataset format.
Q: Where can I go with specific questions regarding wrds?
A: Contact WRDS research and technical support at [email protected] with any problems, questions, suggestions concerning database access, data query and programming in SAS, Fortran or C.
Contact the WRDS marketing team at [email protected] with questions regarding a subscription to WRDS, account authorization, or information on various data sets on WRDS.
Q: What databases are accessible through WRDS?
A: WRDS data includes: stock exchange (market-wide and company specific), balance sheet (company specific), income statement (company specific), banks, bonds, bills, inflation, stocks, macro-economic information and currency options (PHLX).
Q: What is FDIC?
A: “The Federal Deposit Insurance Corporation (FDIC) database contains financial data and history of all entities filing the Report Of Condition and Income (Call Report) and some savings institutions filing the OTS Thrift Financial Report (TFR). These entities include commercial banks, savings banks, or savings and loans (The Federal Deposit Insurance Corporation Database p.1).”
“The Federal Deposit Insurance Corporation (FDIC) is a United States government corporation created by the Glass-Stegall Act of 1933. It provides deposit insurance, which guarantees the safety of deposits at 7,895 institutions for safety and soundness, performs certain consumer-protection functions and manages bank in receiverships (failed banks).”
FDIC files include structure, financial time series, complex derived integers, ratio, and merger history data. For further information, visit the FDIC page.
Q: What is PACAP?
A: The PACAP Research Center creates, maintains, and distributes comprehensive, computerized databases which track capital market data for eight Pacific-Basin countries on a continuous and systematic basis.
Q: What is PHLX?
A: The Philadelphia Stock Exchange (PHLX), founded in 1790, is the oldest organized stock exchange in The United States.
Q: How can we merge CRSP and Compustat if we do not subscribe to CRSP/Compustat merged?
A: In this case you need to extract data from Compustat and from CRSP and cross-merge using a common variable, for instance, the CUSIP identifier. Note that CRSP CUSIP’s identifier has 8 digits because it contains the 2-digit issuer code as well. Compustat CUSIP identifiers have 9 digits.
In the following program we will use PROC SQL to obtain information for three firms both from Compustat and CRSP.
title ‘Merge compustat and CRSP by CUSIP’; option sources; libname comp ‘/wrds/comp/sasdata/na’; libname crsp ‘/wrds/crsp/sasdata/sm’; data compu1 (keep = cusip8 tic conm); set comp.names; where tic in (“DELL” “T” “IBM”); cusip8 = substr(cusip,1,8); /* create 8-digit CUSIP */ run; proc sql; create table total as select compu1.cusip8, compu1.tic, stocknames.* from compu1, crsp.stocknames where compu1.cusip8 = stocknames.cusip; quit;
The SQL statement creates a table named “total” that contains all the variables from msfnames file together with the variables “cusip8” and “tic” from compu1 file. The table will only contain those observations that have the same CUSIP identifier in comp1 and stocknames files.
Q: Where do I find WRDS sample programs?
A: WRDS samples programs can be found at WRDS Support – Sample Programs. They are also located in the samples directory under the directory of each data distributor, e.g., /wrds/crsp/samples contains CRSP programs. Programs ending in “.f” are Fortran programs; programs ending in “.sas” are SAS programs; and programs ending in “.c” are written in C.
Q: How do I compile and run a Fortran program?
A: A Fortran program must be compiled before it can be run. Compilation checks the program for errors and creates an executable file that can be run to obtain the results. Some programs require the inclusion of an object library at compilation.
Q: What is a Fortran object library?
A: An object or archive library is a file that contains data definitions and parameters. This library also contains symbols such as calls to routines that may be needed for correct compilation of a program. Currently, the only wrds programs that require the use of an object library are the CRSP programs. The CRSP object library file for random access is $CRSP_LIB/crsplib.a
Q: How do I extract data for a specific company using Fortran?
A: To extract data for a specific company, the identification code for the company must first be known. Next, find the sample Fortran program that applies to the data set and make the necessary changes to specify the company to be extracted. These changes are as follows and are based on an example of finding IBM’s (PERMNO 12490) returns using the CRSP Monthly Stock file.
Q: How do I extract data for a group of companies using Fortran?
A: This is easily accomplished for CRSP Stock files by using a feature that allows the program to read from a list of PERMNOs. To implement this feature, create a file that contains the list of sorted PERMNOs in a single column aligned one space from the left, and name it permno.dat. Next, make changes to the program (e.g.: msf.f) that include removing the C from the beginning of each of the following lines as instructed by the comments:
Q: How do I extract data for a specific time period using Fortran?
A: Methods of extracting data for a specific time period is dependent on the data being used. Most of the sample programs include comments that instruct the user to make the necessary changes.
Q: How do I extract data for all companies in an industry using Fortran?
A: Extracting data for all companies in an industry involves using the industry identification code in the IF statement.
Q: How do I extract only the variables I need using Fortran?
A: Extracting only specific variables involves making changes to the WRITE statement in the Fortran sample program.
Q: Can I transfer the data to a spreadsheet program from Fortran?
A: When a WRDS sample program is used, the extracted data is printed to a file with the extension .out and is saved in the UNIX directory where the program was executed.
Q: Where can I get more reference with Fortran Programming?
A: The following is a source for additional or supplementary information on Fortran programming:
Q: What is SAS?
A: The Statistical Analysis Software (SAS) System is an integrated applications system that gives you strategic control over all your data processing needs and makes an unlimited variety of applications possible; from general-purpose data processing to highly specialized analyses in diverse applications areas. SAS is also a powerful programming language and has a collection of ready-to-use programs (called procedures) that enables you to access, manage, analyze, and present your data (Getting Started with the SAS System Using SAS/ASSIST Software p.1-3).
Q: How do I use SAS on wrds?
A: To modify and use the sample programs available on WRDS, you can login to the wrds Unix system and run the program from your WRDS account. For information on how to log on to wrds and run SAS programs, see the SAS section of the wrds website’s Support tab where you will find documents that address logging on to WRDS and Using SAS with WRDS. To run a SAS program (for example, msf.sas), type the following command at the prompt: % sas msf.sas
Q: Can I transfer my data to a spreadsheet program from SAS?
A: First save your data to a text (ascii) file and then download the file to the local PC using FTP or some other file transfer utility. The file can then be imported into applications, such as a spreadsheet program by treating the file as a text file.
Q: How do I print the data I’ve extracted from SAS?
A: The best way to print data that has been extracted is to save the data to a text file and then download it to the local PC. Use your favorite editor or word processing application to import the text file, format and print.
Q: Where can I get more help with SAS Programming?
A: Following is a list of sources for additional or supplementary information on SAS:
- Getting Started with SAS at WRDS | https://wrds-web.wharton.upenn.edu/wrds/classroom/
- The Official SAS Technical Support | http://www.sas.com/service/techsup