Use of lags in supplementary cards

[randerson]

Is there a limitation/bug in the use of supplementary cards in the @cats dialog? If I specify "dum" as an option, lags are accepted in the # supplementary card. If I specify "exo" as an option and use the same supplementary card -- Cats crashes with a Rats error re seeking to use a series numbered out of bounds. The obvious workaround is to create the lagged series with a Set instruction, entering the lagged series on the supplementary card -- then all is well. I did not see in the manual any caution regarding the use of lags in supplementary cards. Thank you.

[TomDoan]

EXO variables are completely different from DUM variables. EXO variables are treated just like the endogenous variables (inclusion in lagged differences and the cointegrating relationship), except that there is no equation for them. DUM variables aren't included in the cointegrating space and don't have lags added automatically.

[randerson]

Thanks, Tom. Yes, all of that I understood. What surprised me is that the "standard" Rats lag notation -- { x } is accepted when "dum" is used but traps a fatal error when "exo" is used. I had anticipated that the standard notation would work in both cases.