Using WRDS

What do you need to use WRDS

You need the following to use WRDS:
- a username/password to login (Account requests through the website will be forwarded to your institute. Some institutions distribute accounts through their library.)
- not strictly necessary, but highly recommended: a copy of SAS (for example, SAS 9.1.3)

How to use WRDS

There are three ways ('interfaces') to use WRDS:
- Basic: Using the web interface of WRDS. The web interface works well with straightforward requests.
- Advanced: Use SAS to retrieve data from WRDS to your computer and manage it locally (on your computer).
- Expert: Work remotely on the WRDS-server using SSH. (SAS is executed remotely on the WRDS server; this is particularly helpful for running scripts that require substantial computer capacity, for example managing the Trade and Quote (TAQ) database)

These tutorials are geared towards 'advanced' usage of WRDS, i.e., users that download the data from WRDS with SAS and manage this data on their computer (also with SAS). By using SAS scripts, every step with respect to data retrieval and handling can be automated. For basic usage with the web interface, see the manuals on WRDS.

If your institute does not have computers with SAS available and you are not able/willing to obtain a (student) copy, you can revert to 'expert' usage by running the scripts in these tutorials on the WRDS server. WRDS has manuals on using WRDS with shell scripting (SSH), see: Home -> Support -> Remote Access Using SSH, Unix.

Most important databases on WRDS


Compustat consists of information taken from the financial statement of the annual/quarterly report ('10-K/Q'). Sales, net income, total assets, equity, cash flow from operations, etcetera.


CRSP contains information on stock prices, bonds and indices. Returns are available on a daily and monthly basis.


IBES is the database that contains analyst forecasts. Forecasts of individual analysts are available, as well as median/mean expectations.

Other databases

Some (but not all) other databases available for purchase on WRDS include Execucomp (on executive compensation), Company Issued Guidance (CIG, on management forecasts), RiskMetrics (Corporate governance), Audi Analytics (on audit fees), TAQ (intraday trades).

Firm identifiers (TIC, CIK, GVKEY, CUSIP, etc)

Firm identifiers are used to identify the firm by a key. Different organizations use different keys, and as a result, the main databases use different ways to identify a firm. In addition, some identifiers do not uniquely identify a single firm. For example, sometime after a delisting a firm's ticker symbol may be reissued to another firm. And, some keys may change as a result of merger/acquisitions.

Company name Ticker CIK CUSIP GVKEY
Google Inc. GOOG 1288776 38259P508 160329 Inc. AMZN 1018724 23135106 64768
Apple Inc. AAPL 320193 37833100 1690

Compustat (financial statement info) uses GVKEY, while CRSP (stock prices) uses PERMNO. IBES (analyst forecasts) uses CUSIP. SEC filings are based on CIK (Central Index Key). Press releases and articles in news-feeds usually include the firm's ticker symbol (TIC).

Things you need to know about SAS

Installing SAS on Windows

Installation of SAS on Windows XP is straightforward. Installing SAS 9.1.3. on Windows Vista requires special instructions. SAS 9.2 supports Windows 7.

Scripts and datasets

SAS programming code is stored in text format in '.sas' files. SAS datasets (only readable by SAS) have the '.sas7bdat' extension.

Using libraries

SAS uses libraries to organize data. A library refers to a directory on the hard disk; the default directory is 'work'. However, to organize your data, you probably will prefer to include your own libraries.

Assigning a library is straightforward. If your files are on drive D, directory "myFiles", then after assigning a library name (libname myLib "D:/myFiles") you can refer to datasets in this directory with the library name, followed by a period and the dataset name (for example: myLib.myData).

Data management

Managing data is extremely straightforward with SAS. There is virtually no restriction to the size of the datasets, while large collections of datasets are easily organized in libraries. Visual inspection is always a few mouse clicks away. While software like Excel certainly has its benefits, data management in SAS is typically programmed in a SAS script. This means that re-running analyses does not take any time (other than CPU processing time). Functionality that you need more than once can be rewritten in macro format. Macro's are invoked by the script, which makes programming more efficient (reusable) and the code easier to read.

Regression analysis

Performing regression analysis is also supported by SAS. In particular, when the research project requires many regressions, SAS is well equipped to handle this. For example in an event study using the market model, each observation requires an estimate of 'beta'. You may use SAS to download daily or monthly returns of the firms as well as an index. For each firm you then estimate the beta, which is used to compute the 'abnormal' return at the event date. Finally, a 'simple' regression is performed to test whether or not some variable is associated with the abnormal return.

Using SAS together with other packages like STATA or SPSS

Even though data management and regression can be performed in SAS, some users prefer to use another package to do the 'final' steps. For example, SAS can be used to retrieve and manage the data. The final dataset created in SAS can then be converted for example to STATA format (using StataTrans). STATA can then be used to create the tables with descriptive statistics, correlation tables, and perform (final) regressions and other statistical tests. However, in these tutorials only SAS is covered.

All rights reserved. © 2010-2014 [Copyright] [Privacy Statement] [Disclaimer] [About]