@hecht790 If you use consumption spending, or actuarial, it's kind of a die with zero strategy. Of course, if you're 40 it's way different than if you're 80, in terms of the path and accuracy. The tool becomes more precise over the years as you do your updates each January. Many things will change - tax rates, RMD ages, Roth rules, your plans, etc. All we (and the tool) can do is use math and lots of assumptions/guesses.
As well, since it was mentioned, be very careful folks, about doing that 100% equities in your Roth account, as many of the social posts/blogs recommend because of tax-free growth. At some point, if that money is indicated as needed for spending, you are then subject to sequence risk if it's still 100% in equities. That's why it's good to use the capabilities to modify your asset allocation plans over your life phases. Sure, if you're 30 years old and no need/plans to use the Roth for decades, go for it.
When Pralana automatically rebalances assets (mode 1 or mode 2), does it it add the realized capital gains from rebalancing activities in the taxable account, to taxable income?
@hecht790 If you use consumption spending, or actuarial, it's kind of a die with zero strategy. Of course, if you're 40 it's way different than if you're 80, in terms of the path and accuracy. The tool becomes more precise over the years as you do your updates each January. Many things will change - tax rates, RMD ages, Roth rules, your plans, etc. All we (and the tool) can do is use math and lots of assumptions/guesses.
Piggy-backing off your comment: Is the current Roth optimizer valid for the actuarial, consumption smoothing, or any other "spending down to zero" mode?
I'm thinking not, because you are wanting to maximize your total spending over your lifetime, not your EOLV for your heirs. Thinking this through, we actually might want to maximize the discounted variable/discretionary spending over your lifetime (i.e. npv).
Notice the tvm issue doesn't arise in the case of EOLV roth conversions since comparisons are "apples to apples" using one data point: namely, EOL. (Also, any dollar saved without a Roth conversion gets automatically shoved back into investing, so tvm is informally "built-in".) But with a "spending down to zero" strategy, I am frankly not sure if you'd need to incorporate the time value of money or not. Any dollar not taken out of the system (because you don't convert), stays in and gets reinvested, which leads to more money taken out later. So I don't think you do. On the other hand, many people might have a preference for money spent sooner rather than the same dollar later. And if so, you would want the option to discount the value by a user set amount to compare.
So, having thought through, maybe the true test to analyze roth conversions for a "spend to zero" strategy is to compare a simple arithmetic total of all the money withdrawn discretionary/variable over the lifetime, with an option to discount the totals over time (NPV) if the user had a decreasing utility over time. For instance, if 50k spend this year was more valuable than the same 50k spent in fifty years, maybe because you're young and can utilize the money to the max; but this would be personal to every user and many would just choose 0 discount.
Thoughts? Is Pralana Roth optimizer not valid for the actuarial and consumption smoothed withdrawal methods? And is my suggestion a way Pralana might be built to analyze Roth conversions in those cases?
You do have to iterate between the Roth optimizer and the consumption smoother as neither knows about the other. So if you run Roth Optimizer first, then run the Consumption Smoothing, then re-run the Roth Optimizer and re-run the Consumption Smoothing. Since you automatically see whether a change produces more or less spending from the smoother, you know the value of the change, expressed as a steady, inflation adjusted, spending stream. If you want to know the present value, it's just the spending stream times the numbers years in the plan, since your discount rate = inflation and the program already accounted for inflation in the smoothed result.
@ricke 1) But will iterating between consumption smoothing and the roth optimizer really give you the correct answer, given the optimizer only cares about EOLV? For the consumption smooth spending option wouldn't we have to have the optimizer maximize "total discretionary/variable spending" for its aim rather than highest EOLV? After all, by definition the EOLV is 0 whether you do roth conversions or not. And the actuarial spending doesn't use smoothing in any case, so is the Roth conversion valid at all with that spending option?
Come to think of it, is the Optimize Withdrawal Priorities invalid for the same reason? Is EOLV really what matters when deciding between taxable, IRA or Roth spending order? (That might be worth its own topic post.)
2) Regarding the discount, what you say makes sense: with today's $s there is no need to do anything else other than sum up the discretionary (aka variable) spending over the plan to compare Roth vs Non-roth or any other two options in terms of total spending. Since there is almost always a small legacy, due to rounding errors etc, you could always get fancy and do Sum (spending) + EOLV to get apples to apples. Being Devil's advocate, maybe one reason one might add an additional discount amount above CPI would be if you valued future spending less than current spending. (Say you would value 50k discretionary now more than the same 50k when you're 86, because you are healthy enough to use it to the max.) In that case an additional discount might be appropriate above CPI to evaluate two spending options over time. I personally wouldn't go that route, since I have a feeling when I'm 86 I'll value my 50k just as much as I do now, even if I use it for completely different things than I would now.
The goal of the Consumption Smoothing is to find the constant level of spending that will fully deplete your accounts at EOL. The back and forth iteration with the Roth Conversion Optimizer still works because the Roth Optimizer will base its answer on the smoothed spending, so if a Roth Conversion helps, it leaves a residual EOL balance, and increasing that balance is the goal of the Roth Optimizer. Then you re-run the smoother, it ups the spending to bring the End of Life value back to zero, then you repeat the Roth Conversion. Since Roth Conversions are only going to make a few percent difference in most cases, it only takes a couple of back and forth iterations before this converges.
If you want to know the present value, it's just the spending stream times the numbers years in the plan, since your discount rate = inflation and the program already accounted for inflation in the smoothed result.
I think I'm going to turn in a feature request to add total spending table to the monte carlo results page. In addition to the graph, there is currently a table of EOLV values (deterministic/monte carlo/total savings/effective savings). It would nice to have an additional column or second table of total spending as well right next to it.
Values of interest for the suggested table: Lifetime spending total (deterministic and median monte carlo); and total and median discretionary (non-essential/variable spending) spending.
Some valuing legacy would be interested only in the EOLV, as is currently tabulated. Others (like me) only in the total consumption. Many others would be interested in both legacy and consumption, perhaps the sum of the two.
