I have a question about Jensen’s inequality and Dynare. I am working with an Asset Pricing Model that involves computing the risk free rate which is the inverse of an expectation. The time t+1 interest rate is

R^{f}_{t+1}=1/[E_t (M_{t+1})]

The way I include this in the model section is:

```
m = DELTA^THETA * exp((-THETA/PSI)*dc(+1) + (THETA-1)*log(rc))
rf = 1/m;
```

My understanding is that Dynare will interpret this as Rf=E_t (1/M_t+1), which is not the same as the inverse of expected M due to Jensen’s inequality. Is my understanding correct? If so, how does one get around this problem.

Many thanks in advance.