I am attempting a Roth optimization. First I manually set the priority to "No Conversion." The output is shown on the upper part of the attached page. I then enabled the conversion and got the best result for 24% Bracket. This is the middle output on the attached page. Then I attempted to use the PRC Optimization. Input for optimization was Beginning Year = Left Blank; Duration = 20 years. The result is shown on the bottom of the attached page. The optimized result sets the maximum bracket to 10%, which essentially results in no conversion. This is a much worse result that I achieve through manually optimization.
Not sure if there is a bug or if I am doing something incorrectly. The odd thing is that I think I was getting a correct result a week or so ago, when I was looking at this. As far as I know I haven't changed any of the parameters.
Happy to send along and export file if needed.
Using Gold V3.3
It looks like tax-wise, the manual will do better. But there are plenty of other factors at play. I think PRC judges the "best result" by your ending net worth, not the taxes saved. Have you gone to tabular results for the different test cases?
There can be other factors at play here - such as PRC seeing that you really don't need the money due to other income sources being higher than your expenses. Plus, the impact of the conversions on things other than your marginal tax bracket - social security taxation, premium tax credit if you're using ACA, IRMAA if you're getting Medicare.
As well, there can be reasons other than math to do the conversions even if they're not optimal to your end net worth - i.e. you want the money you leave behind to heirs/charity to be tax-free.
I'd check the tabular results for the ending net worth in those 3 cases and report back. I'm interested.
Thanks for the reply. The graphs I included in each file show the net savings. But just for kicks I went to the tabular data and pulled the Net Worth. At the last year of life expectancy here are the Net Worth values.
Baseline No Conversion = $7,712,952
Manual Optimization = $9,469,952
PRC Optimization = $7,712,952
Since the PRC Optimization is setting a tax bracket level were no conversion can occur, the net worth is going to match the result were no conversion is forced. Clearly there is a much better result for the manual optimization. Of course it takes until the 28th year of retirement for the conversion to show a greater net worth than doing no conversion. I am assuming PRC would optimize for net worth at the end of retirement.
I really have no doubt that doing a conversion in some form will be optimal in the long run. I'm just trying to figure out why PRC Roth Optimizer seems to be settling on a result that is clearly contrary to what I can show just manually changing the conversions tax bracket. I'm either doing something wrong or there is some kind of bug here. I'm betting I must be doing something wrong.
Newbie here, and very interested in further answers to this. Could it be related to asset allocation? As I am setting up my portfolio, I have it set to Roth 100% stock as a priority. But I guess since all 3 trials involve whatever your asset allocation is, that wouldn't be the case. Hmmm.
@lmolino If this is still an issue, would you please send me (mail@pralanaconsulting.com) an export file so I can investigate?
Thanks,
Stuart
just an FYI - I am seeing the same results (Optimizer recommending no conversions even though I can see a significant benefit with manual entries).
I ran the Roth Optimizer with the original input file (that came with Pralana), and the optimizer works well. But for my portfolio, it recommends no conversions.
As a related aside, IMHO Excel PRC would be greatly improved with better documentation - When experts such as @hines202 post "I think" about a feature it seems clear we all could use better education on, and understanding of, the tool. Also previously suggested Case Studies which could prove useful.
Yes, this Forum is an outstanding source of knowledge but a thorough, in-depth Official PRC User Manual would be invaluable. Not an easy task but nonetheless an important component of the best software.
Hoping webPRC recognizes and meets the need...
@bo3b For sure, I've been a fan of RTFM since the early days of computing. Sadly, few do these days, which is why we see user manuals becoming a thing of the past. People want real-time chat support and don't seem to want to read any more. Our just-in-time society. Kudos to Stuart for providing one of the best I've seen, especially in these dire times of user manuals!
That said, when I say "I think" it's usually because I'm away from my computer and not able to verify (as I like to do) via the manual, forum, etc before posting something as fact. Or, I'm being sloppy in my language 🙁
I'm planning to do a book on financial planning with PRC (the new web version) that will hopefully be a good companion to the user manual, with lots of corner cases and real usage examples driven from my many, many clients. I just had one that came to me and said "I want you to show me how I can pay zero taxes in retirement, I've been reading books on the topic..." Uh oh. Most of that stuff is click-bait or based on tax law, etc that has changed. And the client has a high SS payment and lots of pretax money, and will be in RMD-land in a few years! I was able to accomplish it for them, though. And, I showed how they will be knee-capping their net worth in the process. But, they don't want to pay taxes in retirement. Lots of weird and unusual stories like that, which I hope to put in the book. Kiss Your Money Hello (link in sig) has done pretty well.
Also, I wrote in to the folks on The Retirement and IRA Show podcast when they asked about any good consumer-facing tools that help plan Roth conversions. They talked about it on the show, which has a pretty huge following! That was in Oct I think, I may have posted here. It was also discussed just recently, in the last few episodes. I don't agree with them on everything but the show is awesome.
Get ready for the revolution in self-financial planning and booting out the scammy advisors/planners!
@pizzaman I actually had an editorial on the subject in the old PC Week rag back in the mid 90s 🙂 Title was RTFM. Still have it around here somewhere. It was the old-school version of "Let me Google that for you."
@taxhater The Roth optimization algorithm in PRC2023 finds the optimum set of marginal tax brackets with the limitation that it uses the SAME bracket in every year of up to two time periods. By using that as a starting point and then manually tweaking the tax bracket on a year-by-year basis, you might be able to gain an improvement. The algorithm we're including in PRC2024 (Excel) and the web version of the tool (coming soon) will be better; it'll find the optimum tax bracket for each and every year.
Stuart
@bo3b OK, Bob, I'll bite. I'm fairly sure you realize there's already an Official PRC User Manual available via the Pralana website. I realize there's room for improvement in anything, but the current manual is 188 pages in length, so it's not a superficial document. You mentioned Case Studies. What else would you like to see improved?
Stuart
@smatthews51 Hi Stuart, I can no longer recreate this issue. So the problem seems to be resolve. Thanks.
