Roth conversion, maybe a simple solution?
I, too, have a desire to model a Traditional IRA to Roth conversion. I have read all the posts I could find on the topic and came away with you can kinda do it if you bend the situation to fit ESPlanner. I wanted to make what seems to me on the surface to be a simple change to ESPlanner to better handle this and maybe even other situations. Given the lifting of the conversion AGI limit next year, I think having a way to model the conversion more precisely would be a big plus.
The most discussed solution in the forums - leaving conversion amount out of Traditional IRA balance, using taxable receipts for the conversion amount to generate the tax bill, and putting corresponding contributions in the Roth to build up the Roth balance - fails to take into account the income the Traditional IRA balance is earning before conversion. This might be small of a one year conversion in the first year of a profile but what if you want to make small conversions in many multiple years to stay in lower tax brackets.
The feature that ESPlanner is missing is a way to remove money from the IRA balance in specific years. A simple way to implement that feature may be to just allow negative contributions on the Retirement Accounts->Contributions screen. That screen already allows you to set specific amounts for specific years If you implement it so that a negative amount has no tax effect then it does nothing but reduce the Individual, Employer or Roth balance available for earning income and making distributions in future years. I think this feature might also let you model a charitable contribution directly from an IRA and give you a start at modeling taking a loan from your 401k.
I suggest implementing the feature as a negative contribution because the UI on that screen has everything needed, except for allowing negative amounts. And being a programmer I can at least envision the possibility that non-taxable negative contributions might be able to make their way through the Compute Engine without any/many changes.
I realize negative contributions might not be as intuitive a concept as Withdrawals but I think it would take more UI work on the Withdrawals screen to implement. But if you wanted to take on more work for the sake of intuitive use then copying the UI from Contributions to Withdrawals and adding an option to allow you to specify whether the amounts to are be taxed or not. You then would have effectively created a way to model any distribution and/or conversion pattern someone wants to enter. Don’t remove the current method on the Withdrawal screen, instead make the user choose which method to use – based on age start and finish or based on specifying specific amounts by year.
RSS
You can get ESplanner to calculate the exact amount of withdrawals from your TIRA without any programming changes. It is a two step process. For example, suppose you plan to convert 100% of your TIRA in eight equal amounts beginning at age 62 and completing at age 69. Let's also assume that the rate of return you put on your retirement accounts is correct. You do the first run to get ESP to report how much the withdrawals will be for each of the eight years by using the "Smooth Withdrawals" tab on the Retirement Accounts page. Set the percentage of retirement account assets to be spent to 100% and set the age range of the withdrawals to begin at age 62 and end at age 69. In this run the TIRA shows the actual initial amount. Then run Create Reports. On the Retirement Accounts page of the Detail section ESP will show what the withdrawal will be for each year.
You then setup a second ESP run. On this run you zero out the TIRA account intially and show the withdrawals from the previous run as special income that is contributed to the Roth IRA each year.
Since this method uses ESP to calculate the withdrawal amounts, they will be correct within the error bars that apply to ESP.
I didn notice another discrepancy that creeps into the final calculation however. When I zero out my TIRA for the second run, that approximately halves my net worth. The second run shows the correct effect of the Roth conversion, but since my net worth is so much lower it calculates a large purchase of life insurance to protect the living standard of my wife in case I die early in retirement. It's not too much trouble to back out the cost of the insurance using the spreadsheet output to get a more correct annual consumption level.
Hope this helps.
You actually help make my point. You can do the conversion scenario if you bend ESP. But it gets complicated fast if you try to take into account all the details. Your method makes the Net Worth fall short. The method suggested elsewhere in the forum misses income of the TIRA assets before conversion and underreports Net Worth in the early years. I think I’ve come up with a method that takes into account everything but it is a complex use of special receipt and expenditures and manipulation of a reserve fund supported by a spread sheet to calculate the input values.
It would be much simpler if ESP could handle it with just a few inputs. My suggestion of negative contributions was an attempt at creating a method with few inputs which might not be difficult to implement in the UI and Computation Engine.
Given that a lot of people will be asking this question now because of the lifting of the AGI limit on conversion, I think it would be a good marketing tool to advertise ESP has a “built-in” way to do it.
Hi Greg, I like your suggestion. Let me go over this with Dick. We have our hands full working on some other cool improvements, but email me in 10 days on this at kotlikoff@gmail.com and I'll let you know if we think this easy fix you suggest is kosher insofar as it won't mess something else up. Clearly, it's trivial to implement in the UI. best, Larry