Eventus FAQs

The attached article explains how and includes examples.

The attached article documents the Request statement option Zero, which works with the Short option to produce zero-net-investment portfolios in event time or calendar time.

With some regularity we are asked whether it's possible to obtain an output file of ARs or CARs using the Fama-French three-factor calendar-time portfolio regression approach.

The problem with looking for individual firm ARs or CARs from a calendar-time portfolio regression is that the method is what its name says: a portfolio-level regression. The model is not estimated at the security level. There simply are no individual firm ARs or CARs in this research design. It would not make sense to try to apply the parameter estimates to individual securities, as the model is estimated across time and any one security may not be in the portfolio in a particular period.

The Fama-French three-factor model can also be used in a traditional two-step event study (EvtStudy statement option combination FamaFrench TwoStep). In this case obtaining a file of individual firm results works the same way as with the market model or market-adjusted returns. Please see the Eventus 9 new features documentation for the various options for obtaining this type of data in a SAS, Stata, Excel, CSV or text file.

The attached article was last revised in May 2019.

We noticed this article on the SAS website that could be useful to Eventus users who are familiar with Stata but new to SAS. You don't need to know a lot of SAS to be productive with Eventus, but it's useful to know some basics.

To use CRSP stock data, you must subscribe to it, normally through your institution. If you use CRSP via WRDS, your institution is a CRSP subscriber. If your institution isn't on WRDS, it still may be a CRSP subscriber. Only the Center for Research in Security Prices of The University of Chicago Booth School of Business can provide subscriptions to CRSP data. Their website is www.crsp.com.

Most Eventus for Windows users can see https://eventus1.box.com/v/Install-CRSP-Data-for-Eventus for instructions on downloading CRSP stock data and preparing it for use with Eventus. If Eventus is already installed, the app Install CRSP for Eventus in the Windows Start menu (or screen) folder Eventus will guide you through the process of installing or updating CRSP data.

For older versions of Eventus, or if you have been using CRSP with Eventus for some time and prefer to stay with the same format of CRSP when updating, please see https://eventus1.box.com/v/crspsetup. Please note that CRSP had announced that after 2016, they would no longer be releasing data updates in the CRSPAccess format that has been standard for two decades. More recently, they have extended the end of life date for CRSPAccess format by at least two years.

Only Eventus 9.7 and later can use the new CRSP SAS data set format in Windows. We will continue to support CRSPAccess format in all active versions of Eventus for Windows as long as CRSP continues updating data in the format. New CRSP-specific capabilities added to Eventus may require use of the SAS data set format.

/* A common question is how to compare abnormal returns between two samples, *//* such as a test sample and a control sample. This example shows some ways to *//* do this. This example does not address a control firm approach, where each *//* test firm is matched to a specific control firm as a benchmark for the test *//* firm's normal return. *//* *//* First we create SAS data set request files. Text request files could be used *//* also. */data control_sample;infile datalines truncover;input cusip:$8. EventDat:mmddyy10. EventNum:4.;datalines;01527110 4/24/2006 800101858110 10/15/2003 800209063H10 6/6/2005 800316776010 9/30/2003 800425848T10 8/8/2003 800527805810 12/14/2005 800629717810 8/28/2003 800731561610 9/16/2003 800834988210 3/23/2004 800940414L10 9/11/2006 801042709310 3/27/2006 801144746210 1/11/2006 801247110910 10/30/2006 801348300770 9/27/2006 801450239210 1/13/2004 801554927110 5/20/2004 801659161010 1/21/2003 801762937750 12/21/2005 801863110310 1/30/2006 801963862010 6/3/2003 802063890410 9/2/2003 802168190410 8/17/2005 802268218910 1/2/2004 802376027610 11/17/2005 802477669610 11/14/2003 802582028010 2/19/2004 802682967Y10 5/7/2004 802784546710 7/26/2005 802887154620 6/27/2005 802987959M10 4/9/2004 803090265310 1/25/2006 803192922P10 4/3/2006 803292923C10 11/12/2003 803398939010 6/16/2004 803498975W10 5/18/2006 8035;run;data main_sample;infile datalines truncover;input cusip:$8. EventDat:mmddyy10. EventNum:4.;datalines;02493710 3/23/2004 900102493710 5/26/2005 900204004710 8/31/2006 900304621X10 1/10/2005 900405349910 7/18/2003 900515060220 6/7/2005 900616411R10 12/22/2003 900720576820 12/16/2003 900837940X10 3/17/2004 900939525910 4/15/2005 901040414L10 9/11/2006 901142217K10 5/12/2006 901245662V10 9/18/2003 901346033520 5/7/2003 901446069010 10/2/2003 901547110910 10/30/2006 901648887910 9/29/2006 901750419U20 11/14/2003 901855438210 1/12/2006 901962855J10 4/7/2005 902068190410 6/4/2003 902168668810 1/17/2006 902272607910 12/9/2005 902375281A10 6/21/2006 902475884910 5/20/2003 902581685110 2/4/2003 902684761M10 1/23/2004 902787296010 8/25/2005 902888023U10 11/9/2004 902990323610 1/27/2004 903092240G10 7/31/2003 903192532F10 4/1/2005 903292864510 11/13/2006 903395988E20 9/11/2003 903496638710 8/16/2004 9035;run;/* The VPREFIX option selects a prefix for the variable names under which *//* the window cumulative or buy-and-hold abnormal returns are stored in the *//* OutWin data set. */Eventus Page=Wide;title 'Control Sample';Request insas=work.control_sample CusiPerm AutoDate id=EventNum; Windows (-2,+2); Evtstudy vprefix=CumulativeAR Median Value NoSinglePeriod OutWin=work.controlcars;Eventus Page=Wide;title 'Main Sample';Request insas=work.main_sample CusiPerm AutoDate id=EventNum; Windows (-2,+2); Evtstudy vprefix=CumulativeAR Median Value NoSinglePeriod OutWin=work.maincars;/* This SAS data step combines the results from the two samples and creates a *//* dummy variable Main, equal to 1 when a firm is in the main sample. The SAS *//* data set option in= creates a temporary dummy variable that exists only *//* during data step execution, so we have to use an assignment statement to *//* create a variable that remains in the data set. */data complete;set maincars(in=in_main) controlcars;Main=in_main;run;/* This step tests the unweighted mean difference in abnormal returns and *//* cumulative abnormal returns between the two samples. *//* *//* The null is that (mean1-mean2)=0 where mean1 is the mean of the group having *//* the smaller value of the class variable (main=0 in this example). The t-stat *//* is negative in the output for this example, consistent with the negative *//* point estimate of (mean1-mean2)=(control-main). *//* *//* If the Windows statement lists more than one window, a quick way to include *//* all in the TTest procedure is to change CumulativeAR1 to CumulativeAR:. */proc ttest data=complete;title 'Equality of Means Tests (No Weight Variable Used)';class main;var CumulativeAR1;run;/* Another way to do the same thing is simple regression. The OLS t-stat is *//* equal to the proc ttest t-stat for the equal-variance case, except for the *//* sign (because Proc TTest tests control-main). */SuperReg InSAS=complete id=EventNum model CumulativeAR1=main;/* This step tests the weighted mean difference in abnormal returns and *//* cumulative abnormal returns between the two samples. *//* Consistent with the Patell test, The Weight statement weights each *//* security-event inversely to the variance of the prediction error. */proc ttest data=complete;title 'Equality of Means Tests (Weight Variable Used)';class main;var CumulativeAR1;weight car_weight1;run;/* If the Windows statement lists more than window, a separate TTest step is *//* required for each window when the Weight statement is used. *//* *//* The SAS macro weighted_ttest repeats the PROC step for each of several *//* CumulativeARs computed by a Windows statetement in Eventus. To use this, *//* change the Windows statements to include the desired number of windows, *//* run the two Eventus runs and the data complete; step, and change the 8 in *//* the %do statement below to the acutal number of windows. *//* *//* The macro variable &i resolves to the window number in "CumulativeAR" and *//* "car_weight". */%macro weighted_ttest;%do i=1 %to 8; proc ttest data=complete; class main; var CumulativeAR&i; weight car_weight&i; run;%end;%mend;/* %weighted_ttest;Remove this line and the first two characters of the above line to use the macro */

Attached is an article about how to use the calendar-time portfolio regression method in Eventus with details about how the method works, and a zip file containing sample code and data.

A user writes: "Hi, I am an Eventus user through WRDS. I want to add a new factor to the Fama-French calendar time regression, and not sure how to do it. ... Thank you!"

Response:

I assume you have noticed the documentation or example about how to use custom factors (the MyFactors= and Factors= options) and are only asking how to combine this approach with the FamaFrench and associated options.

It isn't possible to combine these two sets of options. Instead, you need to build one custom factors data set containing all the factors you want to use at once, both standard Fama-French factors and the new factor you want to add.

You will need to code your own Eventus statements to run on WRDS through PC SAS remote submit or Unix.

Kind regards,

Arnie

Users sometimes ask why Eventus seems to expect an estimation with market-adjusted returns (MAR). Sometimes users experiment with different estimation periods with MAR and observe changing results, which can seem unexplainable.

With a MAR event study, Eventus uses the estimation period to calculate standard deviations, ranks and sign percentages, so most test statistics are at least a little sensitive to the estimation period. Sample characteristics can determine that returns differ greatly between two estimation periods, increasing the variation in test statistics in such cases.

Also, the mean or median CAR can change because changes in estimation period settings alter the sample composition. For example, if we set MinEstN=10 and experiment with EstLen=100 versus EstLen=10, we may observe different sample compositions. With EstLen=10, MinEstN=10 excludes any security-event that has a single missing return in this 10-day period, whereas with EstLen=100, such a security-event would not be excluded as long as it had no more than 90 missing return days in the 100-day period.

Changing the Est= option can also affect missing value screens, and even can move the estimation period to a date range during which a particular security is not listed.

Users asking about the latter situation often submit code including the NoNames or NoPrint options. The first sections of the default Eventus output (in the Ouptut window or *.lst file) are designed to allow the user to see which security-events are included or excluded. It makes sense to suppress this output when one is already aware of and satisfied with the sample that survives, but allowing it to appear (by omitting NoNames and NoPrint) can provide useful information when one is still testing and developing the Eventus code for a particular research need.

If you aren't interested in standard deviations or test statistics, you can do the following to minimize the impact of the estimation period: Add EstLen=10 Est=+1 (or -1) to the Request statement and Overlap to the EvtStudy statement. This will overlay the estimation period on the event period.

Some users have commented that they have trouble transferring columns of displayed Eventus results from the SAS output window to a table in the paper being written, especially when only a subset of columns is to be reported. The following procedure may make it easier. This was developed in Microsoft Windows.

1. In Word, insert a table with the appropriate number of rows and columns.

2. In the SAS Output window, or in Notepad if using the *.lst file from a WRDS web or shell run, hold down the Alt key and use the mouse to select the column of results you want to transfer. Holding down the Alt key, to activate selection of a column that is an arbitrary number of characters wide within a page of text, is a standard feature of Windows that works in many applications.

3. Switching back to Word, use the mouse to select the empty cells where the copied results should go.

4. Hit Ctrl-v or use your favorite way of issuing the Paste command.

Repeat steps 2-4 as needed. Good luck with your paper!

If your institution is an Eventus for WRDS subscriber, or you or your institution has an a current Eventus for Windows subscription, or if you are an authorized user on an Eventus for Windows single-version license that has unexpired support rights, you can us a question directly. In the upper right corner of this page, click on the menu icon and select Support. The menu icon looks like the image below, but you have to click on the one above right.

Here are some tips for dealing with an "Out of memory" warning when running Eventus code (on WRDS through SSH or PC SAS, or outside WRDS such as Eventus for Windows).

Add the option BigSample to the Eventus statement, e.g. Eventus BigsSample; This implements some memory-saving techniques behind the scenes and may be enough by itself. If you have other options on the Eventus statement, BigSample can go before or after them.

a. If you don't need to see the report of data availability for each stock, add the option NoNames to the EvtStudy statement. b. If you don't need to see the market model parameter estimates, add the option NoPList to the EvtStudy statement. c. If you don't need to see overall results for individual days, add the option NoSinglePeriod to the EvtStudy statement. d. If you don't need any listing output, add the option NoPrint (which does everything that the options in a-c do and more) to the EvtStudy statement, e.g., EvtStudy OutWin=outputdataset NoPrint;

If you're not sure what all the options you have in your statements do, ask us for advice. Sometimes eliminating some options can reduce memory usage.

If you are using Eventus on WRDS (Solaris or WRDS Cloud) through SSH or through PC SAS/Connect, see WRDS' advice on adjusting available memory at: https://wrds-web.wharton.upenn.edu/wrds/research/notes/index.cfm#requesting

As a last resort, break up the run into more than one.

A user recently inquired whether Eventus would produce different results depending on whether the request file contained 6- versus 8-character CUSIP values. Many users have asked similar and related questions about CUSIP length over the years.

To answer the user's question, the abnormal returns for any one security will not be different, but the securities included in the sample could differ when using 6- versus 8-character CUSIPs.To explain this in detail, it helps to start with a little background, CUSIP, which is controlled by Standard & Poor's, is officially a 9-character string. The first 6 identify the issuer, the 7th and 8th identify the issue and the 9th is a check digit that is fully determined by the first 8 characters. CRSP (the default data source for Eventus on WRDS and, when subscribed and installed, for Eventus for Windows and Linux) omits the check digit and uses an 8-character CUSIP field. Common equity issues always have an 8th character of 0. As CRSP covers only common stocks, all CUSIPs in CRSP end in 0 as the 8th character. CUSIPs from other sources that have a non-zero digit in the 8th position are not common stock CUSIPs.Given the above-described structure, we can see that eight characters are required to uniquely identify a security. Eventus accepts request files containing either 6-character or 8-character CUSIPs (but not a mixture of both in the same request file). In some cases, the same corporation may have only sequentially different issues in CRSP. For example, a corporate name change sometimes triggers a change of CUSIP. A legal reorganization may cause the retirement of the old common stock and the distribution of a new stock issue with a different CUSIP. Because Eventus takes the requested date into account, purely sequential multiple issues in CRSP are not problematic with respect to using 6-character CUSIPs and Eventus will find the issue that is actually trading on the requested date.In other cases, using 6-character CUSIPs can produce matches that you may or may not be expecting. For example, if an issuer has two publicly traded classes of common stock outstanding on the request file date, the 6-character CUSIP, which identifies only the issuer, is likely to match both classes, potentially increasing the sample size. Please check the information at the top of the Eventus output file to see a report of the sample composition relative to the number of request file lines.

CRSP above refers to the most excellent CRSP Stock database offered by the Center for Research in Security Prices of the Booth School of Business, University of Chicago. Eventus, a product of Cowan Research LC, is the world's first and still foremost software for implementing a broad class of academic research methods known as event studies and includes additional tools for working with CRSP Stock data. Eventus can use data from any source.

Goodbye to the Patell Test in Eventus (PDF)

Question: I'm running a volume event study and wondering, how does one get negative volume? Also what is the scale?

Answer: The event study is done on relative volume, sometimes called

turnover, which is raw share volume scaled by shares outstanding. The

default is to use log of relative volume as suggested in the

literature, so negative values are common. If you want to run the event

study on straight relative volume, not logged, use the NoLogTrans

option of the EvtStudy statement.

Please sign in (site accounts are free and we don't send you promotional e-mail or share your information) for full access to the knowledge base. Once you are registered for the site, you can submit questions through the site or by sending e-mail (from your registered e-mail address).

Detailed user support is available to any user with a current support entitlement, which includes faculty, grad students, research assistants and employees at all Eventus for WRDS subscribing institutions, all Eventus for Windows annual-subscriber institutions, and those Eventus for Windows single-version licensee institutions that have kept their support package up to date.

Please see the attached article, "Using Eventus for WRDS through PC SAS/Connect Remote Submission." This October 2011 revision adds a non-CRSP event study example and now has all sample program files and accompanying input files attached to the PDF.

Using Eventus for WRDS through PC SAS/Connect Remote Submission

Question: I am using the OutWin= option to get CARs or BHARs in a SAS data set, but I would like to include in this file some information from my request file (in addition to the information that the outwin file includes such as cusip, permno, the window dates, and the computed returns), because I need that information for uniquely identifying each record for subsequent data merges.

Solution: You can use the ID= option to specify any one variable from the request file that will be included. As of now that's all; a future version of Eventus may offer multiple ID variable capability.

For now, to prepare for merging by multiple variables, you could create a composite variable for merging purposes only. For example, let's say tag1 is a 1-3 digit integer, tag2 is a 5-character text string and tag3 is a floating point non-negative number <100 with 4 decimal places, and we want to merge by tag1 tag2 tag3 in that order. We could do:

data request; set request; composite_id=put(tag1,z3.0) || tag2 || put(tag3,z7.4); run;

Where request is an existing SAS data set that will be overwritten by a new copy of itself with all the old data plus the added variable. Put a new name in the data statement if you don't want the old copy overwritten.

The put functions are used to achieve consistency in positioning the number within the string so that it will sort and merge properly; in particular the zw.d format pads numbers with leading zeros when needed.

Caution: The maximum length of composite_id is set by the first observation. This is not an issue in the above example, because the use of z formats ensures that the numeric variables always enter composite_id as 3 and 7 characters respectively (assuming their numerical properties are as stated in the first paragraph above). Some other SAS numeric formats conserve space when numbers are shorter and would not be recommended for use in this situation. In situations where you want to declare the maximum length of a character variable instead of allowing it to be implied by the first use, see the length statement in SAS documentation.

The user who asked this question was familiar with merging SAS data sets, information about which is available in SAS documentation.

In the default two-step event study method where parameters are estimated over a non-event period (linear model coefficients if any, variance, percent positive, etc.), the timing and length of the estimation period can be controlled by the Est= and EstLen= options of the Request statement. Sometimes having the same number of trading days or months between the event and the estimation period isn't desirable. For example, where two events for the same firm occur close to each other in calendar time, you may want to have the estimation period for the second event to be based on the first event date or vice versa.

The Est=Specific option of the Request statement lets you provide the date on which the estimation period ends in the request file. If the request file is a SAS data set, use column name EstEnd (for calendar dates) or CRSPEst (for CRSP dates). If the request file is a text file, the estimation date column must immediately follow the event date column(s).

If you want to make the estimation period end a specific number of days or months before or after a reference date, such as a nearby other event date, the attached sample code may help.

Users sometimes ask when the stock return, price or volume data in Eventus will be updated. Eventus itself is an empirical analysis package and doesn't include data. The data set that Eventus uses by default is the CRSP U.S. Stock Database to which your institution subscribes.

When the CRSP database is updated and whether Eventus "sees" the new data depend on 1) when CRSP delivers the new data -- to WRDS, or to your institution in the case of Eventus for Windows -- 2) when someone installs the data and 3) whether the pointers Eventus looks for are updated to point to the new edition of CRSP. In the case of WRDS, WRDS staff are responsible for maintaining both the CRSP database and the link between Eventus and CRSP. In the case of locally installed Eventus, information on how Eventus finds CRSP is available in Appendix C of the Eventus User's Guide.

If you need information on the update schedule, please check with CRSP, WRDS or whomever supports locally installed CRSP and Eventus at your institution, as applicable.

Problem:

Here is the error I am receiving.

ERROR: The current date of Thursday, October 15, 2009 is past the final expiration date for your SAS system, which is Monday, September 28, 2009. Please contact your SAS Software Representative to obtain your updated SETINIT information. ERROR: Initialization of SETINIT information from SASHELP failed. FATAL: Unable to initialize the options subsystem. ERROR: (SASXKINI): PHASE 3 KERNEL INITIALIZATION FAILED. UNABLE TO INITIALIZE THE SAS KERNEL

It sounds as if it thinks my SAS SETINIT has expired but it was recently updated and should good for the next year. Any idea what could be the cause of this?

Solution:

This happens on Vista systems (we've had no reports of its happening with Windows 7). To fix the problem, renew the SAS license (which you may know as "applying the SETINIT") again. First go to the Start menu, find the SAS license renewal program, right click on it and select Run as Administrator. Then proceed as usual. If you don't use Run as Administrator to run the SAS license renewer, the SAS license update is not effective for running SAS in batch mode, which the Eventus installation, hotfix, license renewal, factor builder and other routines use.