You can extend how long your retirement savings last by bucking the conventional wisdom on how to take withdrawals. Even if you are not in retirement yet, you will want to pay attention because the strategy may impact your decisions about what retirement savings accounts to maintain and contribute to.
Conventional wisdom says retirees should withdraw savings from their taxable accounts first, tax-deferred accounts (e.g., traditional IRAs, 401(k)s, etc.) second and tax-exempt accounts (Roth IRAs) last. The general idea is to let the money in the most tax-favored accounts grow as much as possible. In the current issue of the Financial Analysts Journal, Baylor professor (and long-time AAII Journal contributor) William Reichenstein along with Kirsten Cook of Texas Tech University and William Meyer of Retiree Inc. challenge this notion. They say by altering how savings are withdrawn, and even by using Roth IRA conversions and recharacterizations, a retiree can extend the lifespan of his or her portfolio.
At the basis of their rationale is the fact that the aftertax value of funds in a tax-deferred account (TDA) and a tax-exempt account (TEA) are the same if the tax rate is flat. As such, the retiree’s strategy should be to avoid being bumped into a higher tax bracket and take advantage of years when he or she has a large amount of deductible expenses (e.g., big medical bills). This can be accomplished by taking funds from the taxable account first (where taxes are paid on capital gains and dividends, but not on distributions) and then capping withdrawals from the TDAs to the point at which the retiree would be bumped into a higher marginal tax bracket if a larger withdrawal were taken. Any additional dollars needed to fund living expenses should be taken from the TEA, which has tax-free distributions (as long as certain basic conditions are met.)
Before I address the strategy, I want to explain how funds in a TDA and a TEA could have the same aftertax value because I realize this may seem like a strange concept. Let’s assume that an investor in the 28% marginal tax bracket makes a one-time out-of-pocket contribution to a retirement account in the pretax amount of $10,000. He realizes a return of 7.5% per year for 25 years and then withdraws the entire amount. The initial $10,000 contribution to the traditional IRA grows to $60,983. (All $10,000 is used because the contribution occurs on a pretax basis). The ending balance is taxed at 28% when he makes his only withdrawal, lowering the aftertax balance to $43,908. The initial contribution to the Roth IRA would be $7,200. (He cannot afford to contribute the full $10,000 on an aftertax basis.) This balance grows also to $43,908 and is withdrawn in full with no taxes paid on the withdrawal.
(A higher marginal tax rate in retirement would shift the outcome in favor of the TEA, while a lower tax rate in retirement would favor the TDA. Even if one expects the marginal tax rates to increase in the future, an investor could pay a lower marginal tax rate if his or her annual income is lower in retirement than it was during the working years. One argument for incorporating Roth IRAs is to diversify against the risk of not falling into a lower tax bracket.)
Going back to withdrawal strategies, Reichenstein, Cook and Meyer propose withdrawing from the taxable account first, while simultaneously converting part of the tax-deferred account (e.g., a traditional IRA) to a tax-exempt account (e.g., a Roth IRA). The maximum amount of the conversion should be just below the limit at which the retiree would be pushed into a higher tax bracket. Money needed to pay for taxes on the conversion would be paid out of the taxable account. Once the taxable account is drained, the retiree would withdraw from the TDA up to the point that he or she is not pushed into a higher tax bracket, with any additional money needed to cover living expenses withdrawn from the TEA.
This is where a retiree can take advantage of the tax law. Withdrawals from the TDA can be increased during years with big deductions (e.g., high medical bills) without pushing the retiree into a higher tax bracket.
A more advanced strategy—and one that would extend the amount of time retirement savings last—would be to split a Roth IRA conversion into two equal amounts at the start of each year with one half placed into an account holding only stocks and the other placed into an account holding only bonds, with the taxes on the conversion being paid from the taxable account first and the tax-exempt account second (but only after the taxable account has been fully drained). At the end of the year, the account with the worst performance would be recharacterized back to a TDA. In other words, if stocks fared worse than bonds, the year’s conversion of dollars from the traditional IRA into a Roth IRA holding stocks would be undone. The conversion into the bond account would remain unchanged, allowing the retiree to maintain the Roth IRA conversion that is most favorable from a tax standpoint. The conversion amount into each account would be the maximum amount that does not push the retiree into a higher tax bracket. (This advanced strategy can be done with more than one asset class, with the obvious trade-off of more complexity.)
The Week Ahead
Earnings season will hit full stride with nearly 150 members of the S&P 500 reporting. Included in this group are Dow Jones industrial average components International Business Machines (IBM) on Monday; Du Pont (DD), Travelers (TRV), United Technology (UTX) and Verizon (VZ) on Tuesday; Boeing (BA), Coca-Cola (KO) and McDonald’s (MCD) on Wednesday; and 3M (MMM), Caterpillar (CAT), Microsoft (MSFT) and Procter & Gamble (PG) on Thursday.
The economic calendar is light. March existing home sales will be released on Wednesday. Thursday will feature March new home sales and the April PMI manufacturing flash index. March durable goods orders will be released on Friday.
The Treasury Department will auction $18 billion of five-year TIPS on Wednesday.
The views and opinions expressed herein are the author's own, and do not necessarily reflect those of EconMatters. © EconMatters All Rights Reserved | Facebook | Twitter | Free Email | Kindle