i find it too nerve-wracking to do this... i wish i could model the behavior to have the confidence to do this. but haven't sat down and drawn up a spreadsheet.
i assume when the MNAV is jumping much higher than 2.1 is a good time to sell covered calls and MNAV lower than 2.1 is a good time to buy them back. but so many variables it seems impossible to make a model that is better than "buy and hold".
i'm sure somebody with a gigabrain is able to do it. the best person RN that i try to monitor for this is Punter Jeff and Ben Werkman on twitter.