Negative loss (Stochastic Volatility model)

[urikokp]

I have checked the post, but it doesn’t seem to be the same case.
That model diverges to -inf.