Negative variable in equilibrium

I have a very basc question.

I have a model with two agents. One of them is a lender and has “negative debt” (let us call it d). The other one is a borrower and has positive debt (let us call it b). When Dynare simulate the IRF, I do not know how to interpret a positive deviation of “d”. Does it mean that the lender is lending more or less?

Thank you again.

That depends on your sign convention in the model. If d=-0.8 in steady state means a person is indebted, an IRF where d goes up by 0.1 means that d goes from -0.8 to -0.7.

In contrast, if d=0.8 (i.e. positive) in steady state and the IRF is -0.1, it goes from 0.8 to 0.7.

Linearization or log-linearization do only affect the scaling but not the sign and thus do not alter anything.