Largrangain Multiplier as State Variables?

Hi all,

Usually Largrangain multipliers appears as static or forward-looking variable. But in my model, Mu(-1) appears…

Could a Largrangain multiplier be a state variable ? How to understant it ? Is Mu(-1) determined in the end of last period ? Standing at period 0, is Mu(-1) determined in period -1 ?

But Mu(-1) is determined before I make any decision. If no decision was made, there is no optimization, and hence no Largrangain multiplier existing…

So…Mu(-1) makes no sense, isn’t it ?


Are you talking about your own modling or one of Dynare’s routines?
Generally speaking, depending on the setup it can happen that Lagrange multipliers appear with a lag. It is rare, but not impossible. It simply means that you need to take your previous decision into account today when Mu(0) is decided upon.