For richer or poorer; for better or worse; until death do us part…
How to choose? Traditional retirement account contribution Roth conversion ladder decisions have many similarities to the marriage decision. Marriages and Roth conversions are very long-term commitments, and they may or may not work out.
This article focuses on Roth conversion ladder strategies that are more likely to pay off in the long run. Much of the conventional wisdom from the financial media and the self-interested financial services industry is simplistic or wrong. Nevertheless, there are specific situations when Roth conversions are more likely to be profitable to you, your spouse and your heirs in the fullness of time.
Note that this article is paired with another article on the beliefs of some investors that Roth retirement assets are inherently more valuable to own than traditional retirement assets.
https://www.theskilledinvestor.com/VeriPlan/5657/roth-conversion-ladder-strategies/
This other case study article demonstrates, in detail, how to take into account the lifetime tax effects associated with traditional retirement assets versus Roth retirement assets. Such an analysis requires the construction of “side accounts” that capture the effects of lifetime taxes related to either traditional and Roth retirement assets.
This article is based on two primary sources of research and experience. The first is a very detailed analysis by Edward F. McQuarrie, Professor Emeritus in the Leavey School of Business at Santa Clara University. Professor McQuarrie’s analysis was published in 2021 and titled; “When and for Whom Are Roth Conversions Most Beneficial? A New Set of Guidelines, Cautions and Caveats.” You can download this research paper from the Social Science Research Network with this link:
https://ssrn.com/abstract=3860359
Professor McQuarrie carefully and quantitatively evaluated and ranked numerous Roth conversion strategies. An emeritus professor often has significant latitude on how they choose to spend their time. Given the June 2021 posting date of this research paper, Professor McQuarrie may also have made good use of unexpected pandemic downtime.
While Professor McQuarrie’s analysis is the foundational research source of this article, the second primary source is this article’s author, Larry Russell. For 20+ years, I have been the architect and chief developer of VeriPlan, a Microsoft Excel-based, do-it-yourself lifetime personal financial planning software application. Since 2006, VeriPlan has supported Roth retirement contributions analysis. Subsequently, the DIY ability to plan for a tax-optimal, year-by-year Roth conversion ladder to convert traditional retirement assets into Roth retirement assets was added to VeriPlan. VeriPlan’s Roth conversion feature projects specific, user-changeable yearly dollar amounts that could be converted within each federal income tax rate bracket. When this Roth conversion feature is combined with the VeriPlan Comparison Tool, a user can easily develop an optimal, income tax minimized, lifetime Roth conversion plan.
In addition to VeriPlan’s use by Internet DIY licensees, I have personally developed and refined hundreds of comprehensive lifetime financial planning projections models for individual and family advisory clients. Once I have fully fleshed out a client’s comprehensive projection model, the last analytic step is to use VeriPlan to perform a detailed Roth conversion ladder and lifetime tax optimization analysis. Given each client’s unique financial circumstances, I determine whether Roth conversions would every make sense, given their life expectancies, legacy objectives for heirs, and charitable intentions in retirement. If the projected tax recovery breakeven age is acceptable, I use VeriPlan to define a multi-year Roth conversion plan that is sensitive to graduated federal, state, and income taxes, to Social Security retirement benefit taxation, and to IRMAA Medicare subsidy reductions for higher income retirees.
Table 14 of Professor McQuarrie’s research paper is titled “Rank Order of Selected Conversion Scenarios from Most to Least Beneficial.” This article will focus on the most beneficial scenarios or strategies and help the reader understand why certain strategies are likely to be more beneficial. Professor McQuarrie’s research objectives were broader than this. In addition to identifying more beneficial strategies, Professor McQuarrie intended to debunk much of the flaws in the conventional “wisdom” about Roth conversions.
For example, in his paper’s Executive Summary, Professor McQuarrie stated that his research “effort exposed multiple flaws in conventional wisdom:
“* Future tax rates need not be higher for a conversion to pay off;
“* Nor is it all that helpful to pay the tax on conversions from outside funds;
“* Nor are Roth conversions especially beneficial for top bracket taxpayers as compared to middle class taxpayers;”
He followed these flaws in convention wisdom by stating that “Rather, the greatest benefit accrues to taxpayers who can make the conversion partly in the zero percent tax bracket, i.e., during a year with no other taxable income.”
Economic and investment tradeoffs with Roth contributions and conversions during working years
How should one evaluate whether a Roth conversion is better or worse? First, one needs to understand the proper and rational economic and investment tradeoff model. Because the evaluation is future-oriented, this analysis needs to be done in the face of substantial uncertainties about future outcomes. Given that a myriad of uniquely personal financial factors affect this investment future oriented analysis, a lifetime projection modeling tool is required to do the analysis properly.
This entire article’s analysis is focused on situations wherein the retirement saver could either:
A) Make a traditional retirement contribution to a traditional IRA or an employer-sponsored retirement plan such as a 401(k), 403(b), 457(b), etc. which would reduce taxable income and would thus result in lower federal, state, and local income taxes being paid for that year. Subsequent asset appreciation within the traditional retirement account would not be taxed, but age-related, income taxable RMDs would be required in retirement.
OR
B) Make a Roth retirement contribution that would not reduce taxable income for that year and thus would not reduce federal, state, and local income taxes at the outset. Subsequent asset appreciation within the Roth account would not be taxed, and there would be no additional taxation in retirement.Note that there are various circumstances wherein a retirement saver would not have the right to make additional tax-deductible traditional contributions, but could instead make either Roth contributions or traditional, non-deductible contributions. In such circumstances, since it makes no difference regarding taxation at the outset, Roth contributions obviously would be preferred.
Since the Taxpayer Relief Act of 1997, individuals have been able to make non-tax-deductible Roth contributions to IRAs instead of tax-deductible traditional IRA contributions. From 2006, an employee could also make non-tax-deductible contributions to the “designated Roth contribution” portion of 401(k) and 403(b) plans, if the employer added that feature to their contributory retirement plan. In 2011, the designated Roth contribution feature was also allowed for 457(b) plans. In addition, various rules were phased in that allowed within-plan conversions of traditional retirement assets into Roth assets with the payment of taxes at whatever total marginal federal, state, and local income tax rate the employee would then be subject to.
Roth contributions during working years versus a Roth conversion ladder in low tax years
The source of this graphic is Section 5 of VeriPlan’s Roth Analysis worksheet, which is titled: “Economic and investment tradeoffs between traditional and Roth retirement account assets.” Since this article contains only a brief summary, you could read the full text by downloading the most recent “VeriPlan User Guide” in PDF format and refer to Chapter 14.5. Use this automatic download link:
The conceptual graphic above summarizes the essential elements of Roth contributions versus a Roth conversion ladder investment analysis. Generally, are Roth contributions and conversions while working likely be a beneficial financial strategy? The answer is no, while you are working, because Roth contributions and conversions are likely to be suboptimal.
While working, employees tend to be in higher tax brackets than when they are retired. While working, Roth conversions or contributions would be taken at their then current marginal total overall federal, state, and local income tax rate, which is driven by total family taxable income.
The top half of this graphic illustrates the implications of Roth contributions and/or conversions while working. The longer downward arrows represent the implicit cash outlay invested in those Roth assets from one’s overall financial asset portfolio. That cash outlay equals the forgone tax savings that could have been achieved by making tax-deductible traditional contributions instead. Next, is the horizontal dashed line with arrowheads, which illustrates the lost potential asset appreciation on the foregone potential tax savings related to the those additional Roth contribution and/or conversion taxes outlays. Professor McQuarrie refers to this lost appreciation as the compounded “time value of money.”
Every investment model requires a payback that justifies the investment outlay. In this model, reduced future annual income taxes due to lower annual RMDs in retirement provide the potential stream of cash paybacks on the initial investment. However, the investment payback can only begin at age 73 or 75, when that worker’s RMDs on their accumulated traditional retirement assets would begin. (RMDs start at 75 for those who reach age 75 on or after January 1, 2033.)
In the investment model cumulative lost tax savings plus lost appreciation on those tax savings related to Roth contributions and conversions while working would begin to be recovered at 73 or 75. Each year thereafter traditional RMDs and associated federal, state, and local income taxes on those RMDs would be somewhat lower, because of any Roth contributions and conversions done while working. Professor McQuarrie terms this investment model as “intertemporal tax arbitrage.”
As the years pass in retirement, the tax savings on multiple years of somewhat lower RMDs would accumulate on a present value basis, which would take into account inflation over time. Given the age-adjusted equation used to calculate RMDs, these annual tax savings tend to rise over time and potentially would quicken the pace of the tax recovery. However, with each passing year, appreciation on the original foregone tax saving would increase and push the breakpoint forward.
At some point, there would be a breakeven age where the net then-present value of these retirement RMD tax savings will exceed the net then-present value of the original foregone tax savings and subsequent appreciation.
Slowing the approach of that Roth breakeven age is the fact that in most people’s circumstances their marginal tax rates will decline in retirement, when compared to their working years. In addition, in many US states retirement Social Security benefits are not taxed, which reduces state level retirement taxable income and tax rates — thus slowing the race to breakeven. Moreover, in some states RMDs and other traditional retirement account withdrawals are also not taxed or state income taxes are limited, which also reduces state taxable income and tax rates and slows the pace to break even. (Note that VeriPlan automatically takes into account all of these factors in its projections and Roth analysis features.)
Judging whether any investment is worthwhile requires an estimate of when the investment breakeven would occur. For Roth conversions and contributions, this breakeven can only occur later in retirement. The good news is that once the breakeven point occurs, the return on investment increases as the years proceed. The bad news is that many Roth conversion and contribution breakeven points may occur at a point well beyond a normal life expectancy, and thus be unacceptable to many investors.
Lowering the Roth investment model breakeven age with a Roth conversion ladder early in retirement
Given the likely very advanced breakeven age for Roth contributions and conversions done while working, what is that alternative? How can the projected breakeven age be lowered? The key is to wait until later in life after retirement to implement a Roth conversion ladder at lower tax rates in years with lower taxable income.
The added benefit of waiting until retirement is that you will then know how much you have accumulated in traditional retirement accounts that would be subject to RMDs. If your financial life progressed in a way wherein your total accumulated traditional assets are modest, then taxes on RMDs would tend to be low in retirement and a Roth conversion ladder might not be necessary or at least you might do fewer conversions.
On the contrary, if you have accumulated substantially higher amounts of traditional retirement assets at retirement, you could still take advantage of lower taxable income in earlier retirement years and do a series of annual Roth conversions. Furthermore, with ample assets you might even decide to retire early and draw down some of your traditional assets to live on. Assets enable choices. Earlier retirement spending can reduce future RMDs and RMD taxes, as well.
The lower half of the graphic above illustrates the alternative of not making Roth contributions or doing Roth conversions while working. Instead, one accumulates more traditional retirement assets and decides after retirement whether to convert some of the traditional assets to Roth assets when taxable income is likely to be lower and thus tax rates are likely to be lower. The graphic shows shorter vertical lines indicating fewer taxes paid, and it indicates a shorter dashed-line period of lost appreciation. These factors are likely to result in a payback breakeven age that can be significantly lower that the breakeven age that would be associated with making Roth contributions or doing Roth conversions while working.
As previously mentioned, Professor McQuarrie had broader objectives including testing convention wisdom about higher, lower, or similar tax rates between working years and retirement years. For example, on page 32, he states: “Over a lifetime of retirement savings by an affluent professional, most contributions, if made to a traditional 401(k), will have been subsidized at [federal marginal tax] rates of 32% (Trump), 33% (Bush and perhaps Biden), or 36% (Clinton). As documented in Table 3, a lifetime of saving the maximum statutory amount will barely suffice to nudge such a taxpayer into the 24% [federal marginal tax] bracket during retirement. Any switch to Roth contributions, at any point, will entail giving up a [federal tax] subsidy of 30-some percent to save on RMD taxes at 20-some percent. And a particularly aggressive combination of Roth contributions and Roth conversions might drive the projected TDA balance below $2 million, per Table 2 — i.e. begin to reduce future taxes assessed at the 12% marginal rate. That is just plain dumb.”
To simplify, the years when a Roth conversion of accumulated traditional retirement assets are more optimal and have a lower breakeven age are those years when taxable income is lower. Those years are NOT when one is working because the total federal, state, and local percentage income tax shield provided by tax-deductible traditional contributions is much more valuable. The only plausible working years exception would be related to a planned sabbatical or unplanned unemployment event, wherein taxable income and taxes are quite low in that particular year, including taxes on any assets necessary for living expenses to tide you over. Otherwise, the most opportune years for a Roth conversion ladder occur early in retirement when taxable income is lower — particularly if Social Security benefits can be delayed and RMDs have not yet begun.
Marginal versus weighted average total federal, state, and local income tax rates
In an particular year, your marginal federal, state, and local total income tax rate on a Roth conversion ladder will depend upon any other taxable income that must be on your tax returns in that year. Wage, salary, and bonus income in working years tends to be comparatively high, so total marginal tax rates must be high. Without higher income from work, there cannot be excess income to redirect as contributions into tradition retirement accounts. If you do not build up substantial traditional retirement assets, then RMD taxes would not be a problem in retirement, and the Roth conversion question would be moot.
If planned for, taxable income in the beginning years of retirement can be quite low and sometimes zero. A current 55 year old couple might plan to retire at 65. Their RMDs would not start until 75. At least their largest potential Social Security could be delayed until age 70, while it increases in value. Should they be so luck as to have a pension, they might be able to defer receipt while that check increases, as well.
For some retirement years, they might have minimal or zero other taxable income. They would be entitled to a $29,200 federal standard deduction (in 2024), which in effect has a marginal tax rate of 0%. With no other income, they could convert traditional assets into Roth assets up to that standard deduction, and pay no taxes. Their investment model breakeven would be instantaneous, and the resulting, but small annual RMD tax savings would occur and compound year upon year.
Using 2024 federal income tax rates for a married couple filing jointly, this is an example of the weighted average tax rate calculation for a $100,000 Roth conversion, when there is no other taxable income:
A) $0 income tax on the first $29,200, which equals the standard deduction. (Remaining to be taxed: $100,000 minus $29,200 equals $70,800)
plus
B) $2,320 income tax on the next $23,200, which equals the width of the 10% income tax bracket. (Remaining to be taxed: $70,800 minus $23,200 equals $47,600)
plus
C) $5,712 income tax on the next $47,600, which is all taxed at 12%, because the width of the 12% income tax bracket exceeds $47,600)
totaling
D) $8,032 total federal income tax on the $100,000 Roth conversion, which equals a weighted average federal income tax rate of 8%
This “no other taxable income” in retirement conversion example above results in an 8% average conversion tax rate. Thus, it is inevitable that the investment breakeven age will be lower than the same $100,000 Roth conversion done earlier while working, when a marginal federal income tax rate of 22%, 24%, 32% or higher might have applied.
Obviously, any state and local income taxes would increase either total tax rate above, but when a couple is subject to state and local income taxes those rates would be proportionally smaller additions. Furthermore, just like the federal income tax system, states with graduated income tax rates assess lower rates on lesser retirement incomes.
Furthermore most states exclude Social Security benefits from retirement taxation, and some reduce or eliminate taxation on traditional withdrawals and RMDs. For information on how each state does or does not tax Social Security retirement income or traditional account withdrawals and RMDs, see Section 4 of VeriPlan’s Retirement Planning worksheet or Chapter 3.4 of the VeriPlan User Guide.)
Note: Arranging for the lowest marginal tax rates in the early years of retirement requires some asset planning. If one does not have other assets to pay living expenses already saved in taxable accounts, then it would be necessary to withdraw additional amounts from traditional retirement accounts and pay associated taxes. Whether taxes would be due on any assets in taxable account would depend on their tax basis. Spending cash assets would not involve taxation, while appreciated stock assets would incur some level of long-term capital gains (LTCGs) at the federal level. Federal income tax rates are higher that LTCG tax rates, but these two tax systems are linked and interact.
Professor McQuarrie’s Appendix Beta is titled: “Understanding the Difference Between Marginal and Average Tax Rates and the Implications for Roth Accounts.” This appendix emphasizes the difference between marginal and weighted average tax rates. His Appendix Table 1 used 2020 federal income tax brackets to calculate the difference between the marginal tax rate at the top of each tax bracket with the weighted average tax rate that would actually be paid in cash up to that point. The weighted average tax rate is always noticeably lower than the marginal tax rate.
Below, I have updated the numbers to the 2024 federal tax rate schedule including the 0% standard deduction bracket. All of the tax rate calculations are very close to what Professor McQuarrie published for 2020, because federal tax law requires that the standard deduction and the tax brackets be adjusted annually for inflation. (All 2024 average tax rates and rate differences are with two tenths of one percent compared to Professor McQuarrie’s 2020 calculations.)
US federal marginal income tax rates versus weighted average income tax rates in 2024
Columns:
A = 2024 US federal marginal income tax rate at the bracket ceiling
B = Taxable income including the standard deduction at the federal marginal income tax rate bracket ceiling
C = Cumulative 2024 US federal cash income taxes at the bracket ceiling
D = 2024 federal weighted average income tax rate at the bracket ceiling
E = Gap between marginal and weighted average tax rates at the bracket ceiling
These are selected comments by on marginal versus average tax rates by Professor McQuarrie (page 31):
“Federal income taxes have a progressive rate structure: and extra dollar of income at higher income levels gets taxed at higher rates. That extra dollar is said to pay the marginal rate that applies at that income level. …
“At middling and higher levels in the bracket structure, it becomes important to distinguish the average tax on income booked at that point, versus the marginal tax that would be paid on the next dollar earned. …
“In any progressive rate structure, the average tax rate in all but the top bracket must be lower than the marginal rate applied in that bracket. Because the U.S. bracket structure is now such a mess, the exact relationship between average and marginal rates as one moves through the brackets is not readily accessible to intuition.”
When this information is applied to the Roth conversion ladder strategy problem, it becomes clear that earlier Roth contributions and conversions while working and earning at higher marginal tax rates are substantially more costly to execute that those done in lower tax years.
Read my other article on the beliefs of some investors that Roth retirement assets are inherently more valuable to own than traditional retirement assets. This other Roth retirement asset case study article demonstrates, in detail, how to take into account the lifetime tax effects associated with traditional retirement assets versus Roth retirement assets. Such an analysis requires the construction of “side accounts” that capture the effects of lifetime taxes related to either traditional and Roth retirement assets.
