Hey Nicolas - interesting piece. You could use your CMA as inputs to your Monte Carlo simulation. Any thoughts? Errors in your MC results then stem from errors in the methodology used to create the CMAs.