RE: new firecalc
Most curious though. I'm doing pretty normal stuff. Lump sum at the start, an annual withdraw with the default inflation increase per year and the result is consistently 0.5% worse than the old one.
I reported the same thing as I exercised the new version. I can understand saomewhat different results between the two when you target SWR's of less than 100% (because the new version drops the incomplete periods), but it seems to me like the results from the two versions ought to be a bit closer when modeling 100% safe scenarios.
Well, I wasn't specifying 100%, but I'm relieved to hear someone else is seeing the same effect.
I think the issue may be that it's not so much there are incomplete periods that are ignored . . . it's that the periods ignored were the 80's and 90's that were so strong. In terms of a historical profile, this would seem to mildly data mine negative years years/decades from history. But I'm only speculating as to the cause and the guy doing the work has thought about this stuff carefully. It would just be nice to have a mode available to precisely reproduce the old result, so that when results differ one knows exactly why.