About mspak

mspak · ‎04-25-2012

Dear all, I have a SAS file with the following columns: ticker = database company identification code years = initial public offering (IPO) year ID_HK = identification code for HK database IPO_Date = IPO Date prospectus_announcement_date = IPO prospectus announcement date Offer amount = IPO offer amount (million units) Offer price = IPO Offer price 1997 - 2011 = are the number of shares outstanding (million units) for each balance sheet dates I wish to calculate a variable, FLOAT = number of shares publicly offered/ number of shares outstanding after IPO = offer amount / shares outstanding after IPO (first Balance sheet after IPO) If I define the "first Balance sheet after IPO" = IPO Year or one year after IPO year depending on the number of shares outstanding shares data availability. If the data on IPO Year is available, then the shares outstanding for the IPO year should be used, otherwise the shares outstanding for the one year after IPO. How should I produce an output (in columns) as follows: ticker = database company identification code years = initial public offering (IPO) year ID_HK = identification code for HK database IPO_Date = IPO Date prospectus_announcement_date = IPO prospectus announcement date Offer amount = IPO offer amount (million units) Offer price = IPO Offer price Float = offer amount / shares outstanding after IPO (first Balance sheet after IPO) Thank you for any suggestion. Regards, mspak

mspak · ‎04-17-2012

Hi Ksharp, I have a macro for the winsorization method (see the following code): %macro winsorize(dset, vars, oset); data temp; set &dset; run; proc sort data=temp; by fyear; run; proc means data=temp p1 p99 noprint; var &vars.; by fyear; output out=winvalue p1=p1value p99=p99value; run; data temp1; merge temp (in=x) winvalue; by fyear; if x; run; data &oset.; set temp1; if &vars. < p1value then &vars. = p1value; if &vars. > p99value then &vars. = p99value; run; %mend winsorize; I am thinking to use a specific method for panel data analysis. I am still searching and I will share the code if I found any useful code. Thank you. mspak

mspak · ‎04-15-2012

Thanks Shiva, I downloaded this article too. I read some academic articles but it is unknown on how these methods to be carried out by using SAS. For example: 1. An iterative method (refer: http://www.ecob-consulting.com/assets/pdf/outliers.pdf) 2. High-Breakdown Estimators (refer to attached: American Statistical Association Portal :: Multivariate Outlier Detection With High-Breakdown Estimators - Journal of the American Statistical Association - Volume 105, Issue 489 ) Do anyone have the idea to work out these methods of outlier detection by using SAS? Thank you for any suggestion, if any. Regards, mspak

mspak · ‎04-15-2012

Good days to all, I am doing corporate finance research, in which panel data (ie. longitudinal data) is collected. All the observations are firm-years. I wish to detect the outliers before running any regressions. I read some of the articles from SAS, but there is no single method used to address the needs of panel data. I am writing to ask if it is possible to get some useful references (ie. books or articles or macro functions) on the outlier detection for panel data (ie. longitudinal data) using SAS. Thank you in advance for any helps and suggestion. Regards, MSPAK

mspak · ‎04-12-2012

Thanks Ksharp and other experts' help, I have learned a lot from the discussion in SAS forum. Regards, mspak

mspak · ‎04-12-2012

Thanks Ksharp again, As I used sum() function, so I include comma between them. If I remove the comma, the following error message appeared: ERROR: Summary functions nested in this way are not supported. I used SUM instead of operator of MINUS (ie. -) as SUM will provide a value even there is missing value either on a.ROA1 or b.ROA1. Regards, Mspak

mspak · ‎04-12-2012

Hi Ksharp, I ran the following program, it seems leading to my desired output: proc sql; create table Jones_matchedROA1 as select unique a.tic,a.fyear,a.sic2,a.aem, a.roa1, b.roa1, abs(sum(a.ROA1, -b.ROA1))as DIFF_ROA1, min(calculated DIFF_ROA1)as CLOSEST_ROA1, sum(a.aem,-b.aem)as Jones_PDA from geog.mydata a, geog.mydata b where a.tic ne b.tic and a.fyear=b.fyear and a.sic2=b.sic2 group by a.tic,a.fyear,a.sic2 having DIFF_ROA1=CLOSEST_ROA1; quit; Only ONE observation per firm-year will be derived by providing "having DIFF_ROA1=CLOSEST_ROA1" clause, that is the one with lowest difference. Can anyone confirm this is correct? Thank you very much. Regards, MEI SEN

mspak · ‎04-11-2012

Hi all, I tried to open this file. This data can be opened. Thank you. Regards, MEI SEN

mspak · ‎04-11-2012

Hi all, This is the file. I tried to upload the file from my notebook's desktop instead of my hard disk. I am not sure whether it works. mspak

mspak · ‎04-11-2012

Hi again, One correction: The lowest difference between ROA1 of the firm itself and the matched firm would be defined as closest ROA1. As such, I calculate the difference as = min(abs(sum(a.ROA1, -b.ROA1))). 1. the difference between ROA1 2. as the difference can be positive and negative, so I used ABS (absolute value). 3. In order to further restrict to only firm, so I should choose the lowest value. Thanks again. mspak

mspak · ‎04-11-2012

Thanks Ksharp, I changed the file name as "mydata" for this new file. I have the following variables: 1. tic - company identification code 2. fyear - financial year 3. ROA1 - performance measure, returns on assets 4. SIC2 - 2-digit SIC industry classification code 5. AEM - it is a residual as an output from proc reg I wish to have an output file with the following columns: 1. tic 2. fyear 3. Jones_PDA Jones_PDA is the difference between the firm's AEM with the another matched firm in the following criteria: a) same 2-digit SIC code b) same fyear c) with the closest ROA1. The lowest difference between ROA1 of the firm itself and the matched firm would be defined as closest ROA1. As such, I calculate the difference as = min(abs(a.ROA1, -b.ROA1)) There are many firms can be matched with the firm based on SIC code and fyear. However, I would wish to restrict the ONLY ONE matched firm, that is, the one with the lowest difference in terms of their ROA1. In order to avoid self-matched, so I provide a restriction a.tic ne b.tic. Finally, I would wish to calculate the differences in AEM between the firm and the matched firm (ie. Jones_PDA). I hope this make my points clearer. Thank you in advance. mspak

mspak · ‎04-11-2012

Dear all, I wish match each firm-year (tic-fyear) observation with another from the same two-digit SIC code (sic2) and year (fyear) with the closest return on assets in the current year, ROA1. Then, I wish to obtain a performance-matched AEM (ie Jones_PDA) for all firm-years, by subtracting the AEM of the firm with the closest ROA that is in the same industry as the firm. The closest ROA = min(abs(a.ROA1, -b.ROA1)). The lowest difference between ROA1 would be defined as closest ROA1. I ran a sql using the following program, however, it provides Jones_PDA for all the matched firms. I only wish to have the Jones_PDA for the matched firm with closest ROA instead. It does mean that I wish to have Jones_PDA for only ONE matched firm. proc sql; create table Jones_matchedROA1 as select a.tic,a.fyear,a.sic2, min(abs(a.ROA1, -b.ROA1)) as closest_ROA1, sum(a.aem,-b.aem) as Jones_PDA from geog.data a, geog.data b where a.tic ne b.tic and a.fyear=b.fyear and a.sic2=b.sic2 group by a.tic,a.fyear,a.sic2 order by a.tic,a.fyear; quit; Could anyone guide me on how to modify the program/code to achieve my desired output? Thank you in advance for advices. Regards, mspak

mspak · ‎03-20-2012

Thanks Ksharp, It took a longer time to process the the program than the previous one Regards, mspak

mspak · ‎03-20-2012

Thanks for suggestions. mspak

mspak · ‎03-17-2012

Good days to all; I wish to generate the predicted values and residual values by running an Ordinary Least Square (OLS). I used the following program: Proc reg data=GEOG.USA; Model TA=_ASSET CHANGE_SALES PPE_AT/p r; Id FNA AEM; output out=PRED_USA p = predicted_y r=residual_x; Run; Quit; However, I would like to provide an additonal rules on the above regressions: I have the other variables such as SIC (Industry Classification Code) and fyear (financial year). I wish to estimate the above predicted values and residual values based on the above cross-sectional model for each SIC code for each year (ie. SIC-fyear). How should I modify the above program to incorporate the additonal criteria? Thanks for any helps. Regards, mspak

Online Status	Offline
Date Last Visited	‎03-26-2017 10:35 AM

Re: Select most recent row with value

Re: Select most recent row with value

Re: Select most recent row with value

Select most recent row with value

Re: Optimal Lag length

Optimal Lag length

Creating zipcodes with FIPS codes

Re: Interpolation

Interpolation

Re: Proc Panel Warning message

Re: Select most recent row with value

Re: Select most recent row with value

Re: GMM using Proc Panel

Re: ZIPCODE FOR EACH COUNTY and INTERPOLATION

Re: Linear interpolation

Outlier detection

Re: AR test

Re: Combine Datasets using Inexact Character Variables in SAS

Combine Datasets using Inexact Character Variables in SAS

Re: Comparison between largest and second largest value

Find an outstanding amount after an event

Re: Outlier detection

Re: Outlier detection

Outlier detection

Re: Matched observations

Re: Matched observations

Re: Matched observations

Re: Matched observations

Re: Matched observations

Re: Matched observations

Re: Matched observations

Matched observations

Predicted value and residual value for each SIC-Year

Invalid argument to function ZIPCITYDISTANCE Invalid argument to funct...

Predicted value and residual value for each SIC-Year