Fmincon significance; which statistics to use?

I've used the fmincon function to optimize a utility function in a portfolio selection problem. Now that I have the parameters and the standard errors, I would like to draw conclusions about the significance of the model as a whole and its parameters individually. To make a comparison with a OLS regression: like an F-test to see if the regression coefficients are jointly significant and a T-test to see if the coefficients are individually significant.

Can I just divide the optimized coefficients by the calculated standard errors (like a T-test) or do I need to do something else, and how do check for joint significance?

Kind regards,

Martin