Precisely. Now, the question becomes "How does one test for equal intercept and coefficient?". In order to perform this test, it would be most convenient to use a regression/ANOVA procedure which allows >class variable specifications. Either of the procedures GLM or MIXED >will do. Then, tests for equal intercept and equal slope coefficient >are made by including firm and the interaction of firm with lagged >profit. Code for this test would be as follows (using PROC GLM): > >proc glm data=mydata; > class firm; > profit = firm|lagprofit; >run; > >The test of firm will test equality of intercepts, while the test of >the interaction firm*lagprofit will test equality of slopes. > >Dale > > >>Date: Mon, 9 Jul 2001 13:29:11 -0400 >>Reply-To: John Jones <jonesj@PHARMARESEARCH.COM> >> John Jones <jonesj@PHARMARESEARCH.COM> Re: This is a difficult one! Please Help!!! SAS-L@LISTSERV.UGA.EDU >>Adelina, >> >>Why do you not simply test for equal intercept and coefficient? If they >>are equal then so too is LRPP. >> >>John >> >> >>At 06:35 PM 7/9/01 +0200, Adelina Gschwandtner wrote: >>>Hello Everybody, >>> >>>I have a dataset with 187 firms and two variables :Profits (Prof) and >>>Lagged Profits (LP). Times series for 50 years (t). >>> >>>First I run the first order autoregressive process (AR1) for each >>>company (i): >>> >>>profit_it=intercept_i + coefficient_i*lagged profit_i + error term >>> >>>Then I calculate the long run projected profit rate (LRPP) for each >>>company: >>> >>>LRPP_i=intercept_i/(1-coefficient_i). >>> >>>Then I want to test the hypothesis that all long run projected profit >>>rates converge to a common competitive level by restricting all firms to >>>have the same long run projected profit rate. >>> >>>But how do I do this???? >>> >>>A colleague of mine suggested to test the AR1 for the whole dataset >>>under the following restrictions: >>> >>> profit_it=intercept_i + coefficient_i*lagged profit_i + error term >>> >>> Restrictions: LRPP_1=LRPP_2 >>> LRPP_1=LRPP_3 >>> LRPP_1=LRPP_4 >>> LRPP_1=LRPP_5 >>> ............. >>> LRPP_1=LRPP_187 >>> >>>Then I should get an F-Value and if this value is greater than the >>>critical value (1.30) I can reject the null hypothesis that all long run >>>projected profit rates converge to a common level. >>> >>>But how do I write this in SAS??? >>> >>>I know how I do the AR1 by company: >>> >>>proc reg data=new1 outest=est tableout ; >>>model prof=LP; >>>by coname; >>>proc print data=est; >>>run; >>> >>>I know then how by using the estimated output I can calculate the LRPP: >>> >>>data new2(keep=CONAME INTERCEP LP); >>>set est; >>>run; >>> >>>data new3; >>>set new2; >>>Lambda=LP; >>>run; >>> >>>data new4; >>>set new3; >>>beta=1-lambda; >>>run; >>> >>>data new5; >>>set new4; >>>LRPP=INTERCEP/beta; >>>run; >>> >>>But how should I write the 187 restrictions???? >>>Can someone please help?!?Please! >>> >>>Thanks a lot for any suggestions, >>>Adelina. >>> >>> >>> >>> >>>¤º°`°º¤ø,¸¸,ø¤º°`°º¤ø,¸¸,ø¤º°`°º¤ø,¸¸,ø¤º°`°º¤ø¤º°`°º¤ø,¸¸,ø >>> >>>ADELINA GSCHWANDTNER MAG. >>> >>>University of Vienna >>>Department of Economics >>>BWZ, Bruennerstr.72 Tel:(00431) 4277 37480 >>>A-1210 Vienna Fax:(00431) 4277 37498 >>> >>>Mail: gschwand@econ.bwl.univie.ac.at >>>Adelina.Gschwandtner@univie.ac.at >>> >>>http://www.univie.ac.at/Wirtschaftswissenschaften/gschwand/ >>> >>>¤º°`°º¤ø,¸¸,ø¤º°`°º¤ø,¸¸,ø¤º°`°º¤ø,¸¸,ø¤º°`°º¤ø¤º°`°º¤ø,¸¸,ø >> >> >> >> >> >>_________________________________________________________ >>John T. Jones -- Phone: (910) 509-4814 -- JonesJ@PharmaResearch.com >> >>Rise above the obstacles in life and focus on the positive. °`°º¤ø¤º°`°º¤ø,¸¸,ø >>... for the measure given is the measure you will receive....

--------------------------------------- Dale McLerran Fred Hutchinson Cancer Research Center mailto: dmclerra@fhcrc.org Ph: (206) 667-2926 Fax: (206) 667-5977 ---------------------------------------

------------------------------------------------------------ --== Sent via Deja.com ==-- http://www.deja.com/