About kessler

kessler · ‎09-06-2022

Alexey, If you want to programmatically generate your list of restrictions then you can use the SAS macro language. Here's an example that shows how a RESTRICT statement could be generated from values stored in a data set. Marc data d; do t = 2 to 12 by 2; output; end; run; data _null_; length eqs $ 1024; set d; eqname = 'eq'||left(put(_n_,6.)); call symput(eqname, "lhs = " || t); run; %macro restricts; restrict %do i = 1 %to 6; %unquote(&eq&i) %if &i ne 6 %then, %end; ; %mend restricts; proc tmodel; ... your model code ... %restricts; quit;

kessler · ‎08-29-2022

Alexey, If you're new to PROC MODEL it can be difficult to know the best syntax to express your model. I'd be happy to help if you could clarify what data you have and what you are trying to estimate. If your data consists of many contract prices (price) then your FIT statement (ignoring extraneous options) would just be fit price; How you write out the right hand side of your price model depends on how time, discount factors, amounts (remaining and real) are represented in your INUPUT data set (e.g. what are the rows and columns in your INPUT data set). If you can specify these details it should be possible estimate your model. Marc

kessler · ‎08-29-2022

Yes. You can specify multiple restrictions by separating them with commas on the RESTRICT statement: restrict x1*x2*a+x3*b*0.25 = 60, x1*x2*a+x3*b*0.5 = 75, x1*x2*a+x3*b*0.75 = 79; Also, you may be interested in using PROC TMODEL to estimate your model. PROC TMODEL uses the same syntax as PROC MODEL, but it has an improved non-linear optimization engine that is well suited to problems with nonlinear restrictions. Marc

kessler · ‎05-04-2020

PROC MODEL requires that your model is specified using a mathematical expression including explicit parameters rather than just using a list of variables. To do a two stage least squares estimation for your model you could use something like the following: proc model data=new; am = p_batie*batie + p_ln_ta*ln_ta + p_ROA*ROA + p_leverage*leverage + p_growth*growth + p_age*age + p_CPA*CPA + p_big4*big4 + p_ibsize*ibsize + p_loss*loss + p_loss_lag*loss_lag + p_liquidity*liquidity + p_INVplusAR_TA*INVplusAR_TA + p_btm*btm; fit am / 2sls outv=vdata vardef=n kernel=(bart,0,); instruments _exog_ employ LD; title '2sls results'; quit;

kessler · ‎01-14-2020

If you want all your results in a single data set you'll have to concatenate them yourself. You could do something like this: proc model data=have; y1=a+b*x; y2=c+d*x; y3=e+f*x; fit y1 / gmm kernel=(bart,2,0) vardef=n; ods output parameterestimates=est1; fit y2 / gmm kernel=(bart,2,0) vardef=n; ods output parameterestimates=est2; fit y3 / gmm kernel=(bart,2,0) vardef=n; ods output parameterestimates=est3; quit; data ests; set est1 est2 est3; run;

kessler · ‎01-14-2020

I understand those are the equations to be estimated; however, those equations can only be estimated simultaneously if the data used to estimate each equation are linearly independent of the other two equations. If you would like to estimate each equation independently (or any two out of the three independently of the third) using the HAVE data set that you provided you would need to specify two (or three) FIT statements in PROC MODEL: proc model; y1=a+b*x; y2=c+d*x; y3=e+f*x; fit y1 / gmm kernel=(bart,2,0) vardef=n;; fit y2 / gmm kernel=(bart,2,0) vardef=n; fit y3 / gmm kernel=(bart,2,0) vardef=n; run;

kessler · ‎01-14-2020

I don't think the issue here is how you are expressing the model in PROC MODEL. It is the data you are using to estimate the model. The three equations you want to estimate are linearly independent because they each contain a different error term, 'u,' 'v' and 'w;' however, the HAVE data set computes the y3 error term as linear combination of the y1 and y2 error terms. One way you can avoid that PROC MODEL warning is by adding an independent error term to y3 in the DATA step so that it better represents the model for y1, y2, and (y2-y1): y3=y2-y1+rannor(1) Marc

kessler · ‎08-05-2019

You can restrict your solution conditionally by using the OPTIMIZE option on the SOLVE statement. Below is an example of a simple case where only the solution for the 9th observation is restricted to be less than or equal to 8. Please let me know if you had something else in mind. data d; do t=1 to 10; output; end; run; proc model data=d; y = t**2/10; solve y / optimize out=o; restrict ifn(t=9,y,0) <= 8; quit; proc print data=o; run;

kessler · ‎05-31-2019

The moment conditions depend on the formulation of your model. A simple model with a single moment condition which includes your 6 independent variables, 1 response variable, and an autocorrelated error term could be specified as: proc model data=work.gmmtest; parms a1 a2 a3 a4 a5 p r; endogenous resv; instruments x1 x2 x3 x4 x5 c1; resvstruct = resv - (a1*x1 + a2*x2 + a3*x3 + a4*x4 + a5*x5 + r*c1); eq.m1= resvstruct -p*lag(resvstruct); fit m1 / gmm; quit; -Marc

kessler · ‎02-12-2019

Typically, when there is only one dependent variable in your model you use all the instrumental variables to instrument both endogenous variables; however, if you want to limit which instrumental variables are used to correct for the endogeneity of each endogenous explanatory variable you could use the following: proc tmodel data=yourdata; eq.a = y - ka*a; eq.p = y - kp*p; fit a p / ols 2sls; instruments (a, b c d) (p, q r s); quit;

kessler · ‎02-11-2019

In both PROC MODEL and its multithreaded replacement, PROC TMODEL, you can estimate your model using two stage least squares with the following statements: proc tmodel data=yourdata; y = ka*a + kp*p; fit y / 2sls; instruments b c d q r s; quit; Let me know if you need any further clarifications on how to use PROC (T)MODEL for your problem.

kessler · ‎07-17-2018

For a time dependent model with panel data PROC TMODEL should be able to accommodate your needs. The TMODEL procedure was introduced in SAS/ETS version 14.3. It is a multithreaded version of the MODEL procedure, and it supports the CROSSSECTION statement for specifying panel data.

kessler · ‎09-20-2017

Although the subject of your posting is "Panel data" I don't see how the panel structure of your data is spcified in your PROC MODEL estimation. PROC MODEL does not have built in facilities to handle panel data for this model. The large standard errors may be a symptom of collinearity in your data for this model. The message about convergence indicates that the data may underspecify the parameters in your model as well. To know for sure what's causing these results it would be necessary to examine the REGION data set you are using.

kessler · ‎08-31-2017

PROC MODEL provides many facilities for implementing models like this one. Besides using macros you can use arrays, or the SUM function: proc model ...; endo y; exog x1-x10 ind1-ind10; parms a l t1-t10; array t[10] t1-t10; array x[10] x1-x10; array ind[10] ind1-ind10; array tmp[10]; do i = 1 to 10; tmp[i] = exp(-l*(i-1))*x[i] + t[i]*ind[i]; end; y = a + sum (of tmp1-tmp10); fit y; quit;

kessler · ‎04-26-2017

If I understand your question correctly you would just need to model 'a' using one or more parameters that are also estimated. In the original PROC MODEL example 'p' plays the role of this parameter.

Online Status	Offline
Date Last Visited	‎09-06-2022 11:51 PM

Re: proc model : vector of restrictions

Re: proc model optimization

Re: proc model : vector of restrictions

Re: two stage least squares coding problem

Re: Multiple GMM Regressions by PROC MODEL

Re: Multiple GMM Regressions by PROC MODEL

Re: Multiple GMM Regressions by PROC MODEL

Re: proc model - conditional resrict

Re: GMM moment conditions for time series?

Re: 2SLS regression - multiple endogenous variable

Re: proc model : vector of restrictions

Re: Abbreviation of long sums in proc model

Re: Proc Model log likelihood

Re: Parameter estimation with proc model and ODEs

Re: How to do feasible generalized least square (FGLS) in SAS?

Re: proc model : vector of restrictions

Re: proc model optimization

Re: proc model : vector of restrictions

Re: two stage least squares coding problem

Re: Multiple GMM Regressions by PROC MODEL

Re: Multiple GMM Regressions by PROC MODEL

Re: Multiple GMM Regressions by PROC MODEL

Re: proc model - conditional resrict

Re: GMM moment conditions for time series?

Re: 2SLS regression - multiple endogenous variable

Re: 2SLS regression - multiple endogenous variable

Re: SUR (seemingly unrelated regression) for panel data in SAS

Re: Panel data

Re: Abbreviation of long sums in proc model

Re: Outputting 'state prediction' time series with Proc Model Switchin...