For years financial and tax advisors have counseled their clients to make Roth conversions when deemed expedient, but typically not to the extent the same pushes the client into a significantly higher federal income tax bracket. After the SECURE Act, does this strategy always still make tax sense?

Take, for example, this scenario: A couple, both age 65 and recently retired, have accumulated a combined taxable IRA of $2 million. They are expecting no other significant sources of retirement income, other than Social Security of $60,000 per year. The couple estimates their current combined life expectancy at 20 years.

Especially given the likelihood of higher individual income tax rates beginning in the year 2026, if not earlier, common planning advice for this couple may be to withdraw taxable IRA funds earlier and to a greater extent than is required by the tax law, and then roll this amount (likely after tax, in this fact situation) either into a nontaxable Roth IRA, to the extent the amount withdrawn exceeds the required minimum distribution amount for the year, or into some other form of no-tax (e.g., life insurance or municipal bonds) or low-tax investments. This can no doubt be a sound income tax planning strategy.

The question remains, however, what amount is the optimum annual amount to withdraw from the taxable IRA? There are two basic alternative approaches—the so-called “tax table approach,” where the focus is on not causing the couple to be pushed into a significantly higher current income tax bracket, and the so-called “amortization table approach,” which ignores current income tax brackets and instead focuses of lowering the total income tax liability of the couple and their children, after the couple’s death.

Under the “tax table approach,” the couple may choose to withdraw $80,250 per year (or about 4% of the initial IRA value) for the first seven years, because this will keep them in the 12% federal income tax bracket, and out of the 22% bracket (applying 2020 tax brackets). After that (i.e., age 72) the couple will be forced to take the larger required minimum distribution (“RMDs”). The federal income tax on the withdrawal during the first seven years would be $9,235 per year, or $64,645 over the seven-year period, assuming tax rates do not change and the couple is able to file jointly the entire period. Assuming a 5% growth rate, the couple’s taxable IRAs would be worth approximately $2.2 million after year seven. The total tax on the RMDs from year 8 through year 20 would be $215,343, for a total tax on the RMDs of $279,987 during the couple’s estimated 20-year life expectancy.

Under the “amortization table approach,” the couple would instead add their 20-year estimated life expectancy to the 10-year maximum period over which the couple’s children must withdraw the balance of the taxable IRAs after the couple’s death, and "amortize" their taxable IRAs over 30 years. Assuming a 5% growth rate, equal annual withdrawals would be $128,837. The federal income tax on this larger amount would be $19,923 per year, or $398,478 over the couple’s estimated 20-year life expectancy, again assuming tax rates do not change and the couple is able to file jointly the entire period.  The couple thus pays $118,491 ($398,478 - $279,987) more in income taxes under the amortization table approach than under the tax table approach. [For simplicity, this comparison ignores the impact of Social Security on the IRA owner's income taxes, since the impact will be largely the same under each of the conversion options, and is therefore largely irrelevant for purposes of the comparison made in this article.]

Under the amortization table approach, the amount remaining in the taxable IRA at the couple’s projected death in 20 years will be approximately $1 million; under the tax table approach the amount remaining in the taxable IRA in 20 years will be approximately $2.2 million.

Now we need to compute the approximate annual withdrawal amount to the children after the couple’s death, under the two approaches, assuming no growth and equal annual withdrawals over 10 years and a 5% growth rate. Consistently under the 30-year amortization table approach, these annual withdrawals (on the approximately $1 million starting base) would be $127,279. Under the tax table approach, these annual withdrawals (on the $2.2 million starting base) would be $280,013. 

Now assume the couple has one child, and that this child’s annual taxable income, excluding the equal IRA payments, is $150,000. The child’s total annual taxable income during the 10-year payout period would be $277,279 under the amortization table approach and $430,013 under the tax table approach. 

Assuming 2020 tax tables and that the child tax status is married filing jointly (and ignoring for this purpose any potential tax on Social Security payments), the child’s annual tax liability would be $54,706 under the amortization table approach (or $547,060, over 10 years) and $100,094 under the tax table approach (or $1,000,940, over 10 years), a difference of $453,880 over 10 years.

This amount must then be compared to the $118,491 lower lifetime tax amount of the tax table approach versus the amortization table approach, for a net tax savings in favor of the amortization table approach over the tax table approach, over the entire 30 years, of $335,389. This tax savings could be even higher if the child was in a higher income tax bracket.

It can be argued that, while this tax savings in favor of the tax amortization table approach is substantial, it does not reflect the time value of the loss use of the $118,491additional tax payments during the lifetime of the couple. However, this potential loss in the time value of money must be balanced against the potential that one of the two spouses will likely die some years before the other, so by not withdrawing the additional amount earlier, when the couple’s tax bracket was essentially half the tax bracket of the widow or widower, these two competing factors can be viewed as essentially cancelling each other out. Also remember tax rates could rise in the future, so withdrawing a larger amount earlier may also be beneficial from this perspective.

The numbers can obviously be run a variety of ways, and of course there are countless different client fact patterns presented to us as advisors. The purpose of this article is merely to illustrate that traditional Roth conversions strategies need to be challenged in light of the SECURE Act, to ensure that our clients are not foregoing a significant potential family income tax savings by not exploring all of the Roth conversions approaches available to them. 

James G. Blase, CPA, JD, LLM, is founder of Blase & Associates LLC, a St. Louis-area law firm practicing primarily in estate planning, tax, elder care, asset protection, and probate and trust administration. Mr. Blase is also an adjunct professor at the St. Louis University School of Law and in the Villanova University Charles Widger School of Law Graduate Tax Program. Mr. Blase is also the author of the 2020 book:  Estate Planning for IRAs and 401Ks:  A Handbook for Individuals, Advisors and Attorneys.