Optimize Roth conversions & drawdown between account types


I recently bought ESPlanner & it seems like a very powerful tool. There are a few areas I'd like to optimize, but I'm not sure if ESPlanner provides that capability.

1. Recommend/optimize a Roth conversion plan between when I retire and start Social Security
2. Optimize spending across accounts (taxable, tax-deferred, and Roth IRA), for both pre & post Social Security.

I'll have about 10 years between retiring & starting Social Security, with financial assets about evenly split between taxable and tax-deferred accounts, and I'd like to optimize my drawdown plan between account types and making Roth conversions (before starting SS).

Any thoughts if ESPlanner could help with these optimizations. If not, any other recommended tools?



dan royer's picture

In the retirement screen input area, you will see a place where you can specify (change the order) of the ROTH account (before or after the other accounts). Whether to convert or not is a more difficult thing to model. You have to compare the existing plan with a model that converts say a trad IRA to a ROTH using special expenditures to account for any expenses that might be related. Search the forum for specifics. I've found that it typically makes very little difference, but each case is different.

ESPlanner can definitely help with this. I'd start with sensitivity testing. This can be extremely detailed, but if you know which specific variables individually are positive to your results, these can be combined to see if you can increase their benefits.

Because you plan to start SS several years after retiring, you should consider how to "bridge the gap" in income for those years. Spousal SS, part-time work, IRA withdrawals, Roth withdrawals, special withdrawals, withdrawal order, regular assets vs. tax-sheltered assets, percent of non-annuitized assets to be spent, annuities, pensions, paying off mortgage before retirement, downsizing or eliminating other debts, etc. can all make a difference. In many cases, when to take SS will have a significant impact and sensitivity testing can help you optimize your overall results when combined with the other financial levers.

Once you have a solid "map" created to fill your income "gap" years, then you can make tweaks to optimize the results. Of course, you don't have to do all of this, but I've found it worth the effort.


Brian, can you elaborate on what you mean by "sensitivity testing"? What is the purpose and method?

Edit: Never mind. You answered my question more fully below.

Thanks for the quick responses!

Dan - thanks, I did see where you can select the retirement withdrawal order in the retirement screen input area. The PDF reports lump the retirement accounts together (e.g. "Retirement Withdrawals" column in the Total Income report, and "Retirement Assets" column in the Details Retirement Accounts report), so it's not easy to track the Roth IRA vs. 401k/tIRA asset balance through the years.

I thought an optimum withdrawal strategy (& tax management) during the years after retirement but before SS, would be to spend taxable assets to "bridge the gap" and pay Roth conversion taxes as I move funds from 401k/tIRA to Roth IRA, so my RMD is smaller when I hit 70 & SS would be taxed less. But I'm not sure how to model and test that scenario.

So, essentially adding the taxable account into the mix with retirement accounts, and selecting the withdrawal order between all.

Brian - thanks for your inputs. I need to become more familiar with the program and sensitivity testing.



Yours is a scenario I've looked at as well. Here's how I recall doing it (don't have ESPlanner in front of me at the moment).

1) Set your smooth withdrawal age on all retirement accounts to 70.
2) Specify the amount you want to convert for each year between 60 and 70 as a special withdrawal from your 401k.
3) Specify the amount you want converted to Roth (less than or equal to the special withdrawal) as a Roth contribution in those years.

I think those are the steps. In my case, this was the only scenario in which Roth conversion made sense. Will check my ESPlanner scenario later and update if necessary.

jpridder - thanks for your help.

I read another forum thread that discussed a different method of inputting data for Roth conversions. This method was to reduce the Retirement assets value by the amount you plan to convert, take special receipts (taxable) for that amount, and then add Roth contributions for that amount.
The thread is at: http://www.esplanner.com/forum/roth-conversion-1.
Actually, the last post on that thread if from you, explaining it with an example.

Maybe both methods are valid, but the program exhibited a strange behavior if I use the method you mention in this thread (enter Retirement special withdrawals from my 401k for the amount I plan to convert). I only have 401k & Roth IRA retirement assets (no traditional IRA), but I to enter "Individual" rather than "Employer" special withdrawals, for the withdrawals to show in the results (appeared that "Individual" & "Employer" were switched).

Hi retireGuy,

Your original heading is "Optimize Roth conversions & drawdown between account types" which is where my previous comment came into play - especially "optimize".

I tested a massive number of inputs (basically every ESPlanner variable that was remotely relevant), then combinations to "optimize". This was a huge effort. However, it can be approximated with much less effort. I found a big difference in results (consumption) that can be realized depending on your choices although this may be different for each person.

Here's what I'd do to simplify from what I've read in your post:

1. Set your baseline ESPlanner profile and record key variables (e.g. SS choices, Roth, taxable, sheltered, key ages, etc.) in Excel, then record your annual consumption. I'm assuming you have flat consumption to make it simpler.
2. Come up with a short list of variables you are interested in testing against. Probably the same key variables above. Record high/mid/low options for each. Keep this simple and do not do more complex steps such as Roth conversions (save those for later).
3. Run maybe 10-30 sensitivity tests using various combinations from #2 and record annual consumption to compare against baseline in #1. Hopefully, this will give you a feel for what options are positive/negative and by how much (I used $$ and % to compare them).
4. With more effort, refine/repeat #2 and #3 to narrow into better choices or (combinations of choices) until you are comfortable with your profile and are not finding obvious ways to increase consumption.

With some preparation/planning, you could probably do this in about an hour or two.

5. If you want to put in more effort, other variables such as those in my previous comment along with Roth conversions could improve on the results.

If you have some time, perhaps over the summer, to research and play around with other options, my guess is you'll find a nice boost in your results. The reason I left Roth for #5, is that I found bigger improvements in other areas (e.g. gain in consumption vs. my baseline profile). Once I went through the "low hanging fruit", I went back and tried Roth conversions and other variables to see if they would boost my results.

Hope this is helpful.


Thanks Brian, I appreciate your recommendations and taking the time to write it up.
When I return from travel & get some time, I'll do just that.

dan royer's picture

Just want to make sure Retireguy sees this thread: https://www.esplanner.com/case-studies/convert-your-ira-roth

Thanks Dan, I did read the Roth conversion case study thread.

As far as optimizing the amount of Roth conversions before RMDs and Soc Sec start (to provide the highest smooth discretionary spending), is the best strategy to estimate your tax bracket when RMD & SS start, and enter Roth conversions in prior years up to the max of the next lower tax bracket. Then iterate a few times to see if it's beneficial to also do some conversions at the next higher tax bracket (same bracket as when RMD & SS start). It would be great if there was an ESPlanner option to iterate and provide the optimum Roth conversion amount. Of course all this is dependent on knowing future tax rates, which we don't.

Hi retireGuy,

There's no single strategy that works best for all people here. If you can spread out the conversions so you don't breach the next tax bracket, you're probably in good shape regarding optimizing. Of course, there are restrictions such as delayed withdrawals of converted funds for five years, penalty for early withdrawals, etc. to consider.

There can be situations where you have large amounts in traditional IRA/401k and assume future tax rates are unfavorable so want to convert large amounts to Roth over a number of years although with the tax rate uncertainty that you mention. Sensitivity testing can definitely help here.

One suggestion that I've used to cope with the uncertainty is to develop a "range of reasonable estimates" using some "what if" assumptions. This way you can model conversions with/without various tax increases or other scenarios that seem somewhat reasonable.


We use cookies to deliver the best user experience and improve our site.