Let me jump in here and try to provide some clarification. Again, there are two completely independent topics being discussed in this thread, or at least the title of this thread and the subsequent discussion are unrelated. Mid-year growth timing (as mentioned in the thread's title) refers to the point during the year when Pralana calculates the annual account growth for the various accounts, where the options are either 1) start of the year or 2) middle of the year. The other topic, and the one I think everyone here is trying to come to grips with, is how to account for ACTUAL account balances at some point during the first year rather than January 1 as Pralana is assuming. If you have experienced a large upturn or downturn in your accounts and want to account for it you can just adjust your starting balances accordingly and this will be effective for both deterministic AND historical analyses. Changing the returns for a portfolio time period associated only with the first year will be effective for deterministic projections but will have no effect on historical analyses because they use historical returns exclusively.
With all that said, what we really seem to need is the ability to accept initial account balances based on any start date, not just January 1. As I've mentioned before, there are quite a few complications in doing that but I've been giving the matter a lot of thought and have successfully (I think) prototyped a reasonable implementation in Pralana Gold (our Excel model) and will be working with Charlie to give serious consideration to putting that implementation into Pralana Online sometime soon. With that, you'd just specify your starting date and your account balances on that date and the model would make its projections from that point. This can only work, though, if we make certain assumptions, such as these: RMDs, unscheduled withdrawals and Roth conversions always occur on the last day of the year and are, therefore, still in the future when you specify initial account balances. There will be other assumptions, too, but we'll address that on another day.
Stuart
