Would calculation of Net Legacy assuming "stretch" distributions change the ranking of Tax Lab scenarios?

In order to compare the effects of a distribution plan (e.g., “Taxable, Tax-Deferred, Tax-Free”, or “Roth conversions to the 24% bracket”) on beneficiaries, Tax Lab gives you a “Net Legacy” value and allows you to compare projected net legacies between distribution approaches in the same plan, or even between different plans (with the same distribution strategy or different strategies). 

In order to calculate this Net Legacy value, we assume the accounts are inherited by a couple (married filing jointly) and all funds are immediately withdrawn, subjecting any tax-deferred accounts to ordinary income and providing a step-up in basis for taxable accounts. (Roth accounts are likewise received free of tax.) No other income is assumed to be present that year. This is clearly a simplifying assumption: the beneficiaries would likely have other income and some beneficiaries may distribute their inherited accounts (especially Roth and IRA accounts) over several years. 

In fact, there are an infinite number of possible beneficiary tax and distribution situations. The lump-sum assumption is meant to provide a simple, understandable benchmark for comparing strategies. It is not meant to show exactly what the net legacy would truly be in all situations.

It's reasonable, however, to ask whether the lump-sum assumption favors or disfavors Roth conversions. A stretch approach, after all, would allow IRA funds to be withdrawn in smaller amounts and therefore to be subject to lower tax rates. So, we might worry that the lump-sum assumption disadvantages tax-deferred accounts like IRAs, making Roth conversion plans appear more advantageous than they actually are. It's important to remember, however, that inherited Roth accounts can also be stretched. So it's best to look at actual examples and compare stretch to stretch to keep the comparison reasonable.

The example below shows a simple case where the stretch scenario actually increases the advantage of doing Roth conversions. That doesn't mean we'd see this in every single scenario, but it's interesting that in a fairly straightforward example (not cherry-picked), the stretch helps the Roth conversions scenario, rather than improving the relative position of inheriting a traditional IRA.

Of course, no one can know ahead of time when an inheritance will be received or the tax situation of the beneficiaries, or how those beneficiaries will ultimately distribute the funds. So all examples like this are purely hypothetical.

A More Realistic Scenario

For this example, we assume a 62-year-old couple inherits all accounts in the plan and is subject to married filing jointly tax rates. The couple also has $100,000 in other ordinary income. We compare a lump-sum distribution scenario to one where the couple stretches the IRA and Roth via the “10-year with stretch” rule, where in years 1-9 they take minimum distributions required under IRS Table I and the remaining amount is taken in year 10. We apply actual 2026 tax brackets and thresholds as well as standard deductions. We also assume a 5% real rate of return and use a 3% discount rate for net present value calculations. (Different assumptions would, of course, yield different results, but the advantage to Roth conversions discussed below remains for a range of reasonable inputs.)

For our test case, we use an example plan that begins with $2 million in an IRA and $1 million in a taxable account. When it is run through Tax Lab, Roth conversions to the 12% federal bracket results in a final Roth balance of about $7.1 million. Running a Taxable, Tax-Deferred, Tax-Free approach results in a final taxable balance of $4.6 million and a final IRA balance of $2.6 million (about $7.26 million total gross of tax).

As you can see in the table below, a standard 10-year stretch increases the net-of-tax amount received by the beneficiaries in either case: for the Roth conversion case it increases by about $840,000 in net present value (with the discount rate applied to bring all amounts back to the point of inheritance); for the no-Roth-conversion case, by about $190,000. That means, at least for this example, stretching expands the difference between the scenarios. For a lump-sum distribution, Roth conversions add about $736,000 to net legacy. In other words, for this example, stretching makes the difference $1.38 million in favor of Roth conversions.

Let's take a look at the details.

No Roth Conversions

The scenario with no Roth conversions leaves an IRA and a Taxable account balance. Here we stretch the IRA balance through 10 years, using IRS Table I divisors for years 1-9 and a full remaining distribution in year 10. We use current tax rates for all years and real (today's dollars/adjusted for inflation) balances. The “Total” column shows net-of-tax distributions in each year. A total of $7.27 million is received net of tax over these 10 years. The net present value of these distributions is $7.256 million.

Note that since we are calculating net present value at the point of inheritance, we distribute the fully stepped-up taxable account right away.

Roth Conversions

The Roth conversion strategy for this simple case results in 100% of the final balance being found in a Roth account. Doing the same 10-year stretch results in $10.2 million in net-of-tax withdrawals, the net present value of which is $7.94 million.

Conclusion

This is of course just one example. But it is not cherry-picked. In fact, in a range of fairly standard plans we have not found examples where stretching tightens the gap between Roth conversions and non-Roth-conversion scenarios, or cases where the ranking is flipped. That doesn't mean such examples don't exist. But it does mean we should not conclude that the lump-sum assumption used in Tax Lab overstates the value of Roth conversions. In fact, it likely understates them. That is the desired effect - since Roth conversions are often large and disruptive, we don't want to overstate their impact. Furthermore, since Roth conversion modeling often extends for years and decades into the future - a period in which tax rates and tax situations may differ from what is modeled - it is a good idea to remain fairly conservative in modeling these scenarios..