Hi all,

I would like to know how I can create auxiliary variables (if at all possible) without them representing expectations. The example below should clarify.

I would like for example to write down the following model

```
x = rho * x(-1) + shock;
aux_var = x(+1);
var = aux_var*x(+2);
```

If I write the model this way, I understand that

aux\_var = E_t[x_{t+1}]

and

var = E_t[x_{t+1}] E_t[x_{t+2}]

Now I would actually like to solve for

var = E_t[x_{t+1} x_{t+2}]

Would it be possible to supply some auxiliary variable to write the model in such a way? One solution would naturally be to write down explicitly

```
var = x(+1) * x(+2);
```

However, I would like to avoid such a solution since the expressions for x(+1) and x(+2) could potentially be big and replicating them would be cumbersome.

I hope this was clear enough and happy to clarify. Any suggestions are very much appreciated.

Thanks!

From what I understand, declaring variables in the model blo