Here's a solution that fitted me! It gives slightly different results, but almost the same!
Before a perform a simulation, I go to the "degradation losses" section and take note of all degradation and mismatch values for a given time horizon (in my case, 20 years), in a Excel spreadsheet. Then I untick the option to use degradation in the simulation.
Then I go to the batch simulation and change the variables "module quality" and "mismatch", and paste the values given in the spreadsheet.
But one should never forget the initial values (eg. if I use module quality -0.4% and mismatch 1%, these values have to be added every year)!
I don't know if I was clear enough, but if anyone have a question, feel free to ask!