Hybrid Withdrawal Strategy Works for George and Martha
by
Jordan H. Leibman J.D.
Neal J. Rothman Ph.D.
Richard B. Dull Ph.D.
© Copyright 1998, Jordan H. Leibman, Neal J. Rothman, Richard B. Dull
Like many college professors about to retire this year, George White was amazed and gratified to discover that his "defined contribution" pension account had grown into a whopping seven figure fund. Shortly after his retirement, Professor White converted his entire retirement accumulation, including this pension account, to a Roth IRA. Although this decision will likely prove more tax efficient than several other withdrawal options that were available to him, the yield from his strategy over the length of his retirement will probably fall short by about a quarter million dollars of that which he might have earned had he opted for what we call a hybrid withdrawal strategy. Here’s why:
A "defined contribution" pension plan is one in which the employer contributes to the employee’s retirement account an amount equal to a specific percentage of his salary, producing over time an accumulation that can rise or fall depending on how successfully the contributions are invested. During the accumulation phase the employee is often given some discretion to allocate the contributions made on his behalf to several funds invested in diverse asset classes such as equities, fixed income securities, real estate, and short term money market funds. In addition to his employer’s contributions, George was permitted to make limited personal contributions to a supplementary "salary reduction account" using before-tax dollars, a program similar to that in which private sector employees make personal contributions to their 401(k) accounts.
Although Professor White could boast a long and successful academic career, it was one marked by a modest annual salary. However, White had proven to be an aggressive and successful fund picker, so that on January 1, 1998, the year he would turn 70½ , George and his wife Martha, age 67, were faced with the task of devising a withdrawal strategy for George’s two retirement accounts, now valued at $1.2 million. The growth of the couple’s nest egg had been greatly augmented by its being sheltered from income taxes during the accumulation phase, but now George was required to begin making withdrawals from his accounts according to IRS rules that were designed to exhaust the funds over the couple’s joint life expectancy -- a period of 22 years. Moreover, these withdrawals would be subject to federal and state income taxes at ordinary income rates.
George’s university retirement plan permitted him to accomplish the withdrawal in two principal ways: he could annuitize, that is, he could receive periodic payments over his and Martha’s life in retirement, or he could convert his funds to a traditional IRA. The basic annuity arrangement provides for equal payments based upon a contractually agreed upon interest rate, but also available were actuarially equivalent annuity variations having guaranteed payment periods irrespective of the dates of death. The plan provided that any equities fund accumulation would be withdrawn in the form of a variable annuity, a further variation which shifts the risks and benefits of fluctuating asset values and interest rates from the issuer to the annuitant. Despite the existence of these variants designed to accommodate individual preferences, annuity contracts in general tend to be indifferent to the dynamics of progressive tax rates.
The distribution rules for traditional IRAs require the owners to make minimum withdrawals beginning at age 70½ with an amount equal to the IRA value multiplied by the fraction of one divided by the owner’s and his "calculation beneficiary’s" -- in this case, Martha’s -- joint life expectancy. The owner can recalculate this life expectancy every year or can lock in the life expectancy determined in the first year. Either way, the percentage of the retirement fund assets withdrawn starts small and increases each year as the joint life expectancy gets shorter.
For retirees with large retirement accumulations, annuity contracts and traditional IRAs have the potential for removing more funds from tax-sheltered status early in the withdrawal phase than the retirees need to meet their consumption and income tax requirements. These retirees are, therefore, forced to invest the annual surpluses, either in instruments that are subject to income tax, or in tax exempt instruments that carry reduced interest rates. In short, these surplus funds are likely to earn less over time than investments of equal risk located in sheltered retirement accounts. Although the traditional IRA minimum distribution rules can produce less of these early surpluses than the equal-payment annuity strategy, both of these withdrawal strategies suffer from the infirmity of "de-sheltered" investment.
George and Martha determined that they needed no more than $75,000 on which to live comfortably in 1998, and that amount would be sufficient thereafter if it would increase with the rate of inflation, which they predicted would average a steady 3% over their life expectancy. They also predicted they would earn 7½% interest on funds in their sheltered accounts, but on surpluses they had to invest outside of sheltered accounts they expected they would earn only 5.1%. They also calculated they would have $10,500 in itemized deductions in 1998 exclusive of State income taxes. They determined this amount would rise thereafter with inflation to which the State taxes would have to be added back in to obtain the total amount of their itemized deductions.
In late 1997, George read that Congress’s Roth IRA legislation would permit him to convert his traditional IRA to Roth IRA form. In the spring of 1998, the leading administrator of higher education retirement funds in the U.S., TIAA/CREF, approved Roth conversion for its traditional IRA account holders. For professors like George, this meant that, if his and Martha’s adjusted gross income for 1998 were no more than $100,000, he could roll over his entire accumulation of $1.2 million, first into a traditional IRA and from there into a Roth IRA. The amount rolled over would be taxable, but it would not count in calculating the $100K income limitation on conversions. Moreover, if he completed the rollover to Roth in 1998, the income he would have to recognize from the conversion – the full $1.2 million – could be recognized over a four year period. In subsequent years, Roth rollover income will have to be entirely recognized in the year of conversion.
The principal advantage of Roth conversion is that there will never be further income taxes on the accumulation. In addition, there is no mandatory withdrawal requirement while the Roth IRA owner and his spousal beneficiary are alive. George and Martha will, therefore, be able to maintain during their joint lifetime the investment of their annual surpluses at sheltered, undiscounted, before-tax rates.
Because George converted all his retirement assets into a Roth IRA in 1998, he has four years to recognize the income. But even so, the tax on an additional $300,000 of ordinary income for four straight years poses a significant financing problem. Because George is over 59½ he can dip into his $1.2 "contribution" without triggering an early withdrawal penalty, but he will lose some of the advantage from the four-year-income-averaging, because, under the 1998 amendments, an accelerated withdrawal of contributions will become immediately taxable.
Because George and Martha are hypothetical, we can assume they have a line of credit at 8½% that can provide them with cash to live on and funding for income taxes for the first five years of George’s Roth IRA. Withdrawal of "earnings" (as opposed to "contributions") prior to the expiration of this five year holding period will subject those earnings to taxation (but not to a 10% early withdrawal penalty). We will also assume the couple has additional investments which produce a small amount of taxable income ($3000) in 1998), and still other investments, part of which will be taxable upon their sale, that they will cash in to help finance their conversion strategy. They plan to pay the 8½% loan back in the sixth year from funds accumulating in George’s Roth IRA. Following the holding period both the contributions to, and the earnings of, the Roth IRA can be withdrawn tax-free entirely at the discretion of the over 59½ year old owner.
So, have George and Martha found a winning strategy? Without knowing any more than the above about their finances, we can state with some confidence that this couple has made a big mistake. Although they will have the psychic income of being entirely free of income tax concerns for the rest of their lives, they will have only a trickle of taxable income after the first four years to which their itemized deductions and personal exemptions would otherwise apply. They also have lost most of the benefit of the low 15% tax bracket. These tax benefits foregone over their remaining 18 years of tax-free life expectancy is a huge loss that they paid for with unnecessary taxes on an unnecessarily large initial rollover. The Roth legislation permitting conversions does not require full conversions. All of George and Martha’s retirement assets were in convertible form. To preserve their deduction benefits, they should have retained some of those assets in traditional IRA form and converted less to Roth. Which raises this question: If retirement assets are best split into Roth and traditional IRA components, how does one determine the optimum split? Recall that this couple was content with the annual buying power of $75,000 in 1998 dollars. Recall also, that, at this spending level, George and Martha will generate annual surpluses. Therefore, we can expect them to amass a residual fund (or estate) upon the completion of their joint life expectancy.
We argue here that the withdrawal strategy that produces the maximum residual fund is the optimum strategy for George and Martha. For them, we have determined that optimum strategy to be a hybrid composed of a Roth IRA component for maximum tax exempt saving power and a traditional IRA component to maintain a flow of taxable income so as to preserve the value of deductions and also to lower the marginal and average tax on the initial Roth conversion. Remember that George and Martha start with a finite fund of $1.2 million in 1998. The amount siphoned off in the first year will affect financial results for each of the remaining 21 years of their life expectancy.
To determine the optimal split, the authors created a worksheet showing the financial flows for George and Martha for the entire estimated period of retirement, 1998-2020. As Bar 5 on the following graph shows, an initial conversion by George of $442,000 of his $1.2 million initial fund to Roth -- leaving $758,000 in traditional IRA form -- will produce a residual fund in 2020 of $1,767,552. Compare this with basic annuitization, which produces a residual fund of $1,202,428 (Bar 1); a full traditional IRA strategy that yields $1,417,789 (Bar 2); and a full Roth conversion -- Professor White’s option -- that produces $1,516,951 (Bar 3). In each of these four scenarios George and Martha are assumed to have taken advantage of the 1998 four-year income averaging rule.
If the income averaging rule were not in effect -- as will be the case in 1999 and beyond -- a full conversion to Roth will prove much more costly. Bar 4 illustrates this result: a greatly reduced residual fund yield of $1,020,641. However, adopting a hybrid strategy under the condition of no income-averaging can improve things considerably for George and Martha. Bar 6 illustrates an initial Roth conversion of only $100,000 – to keep initial taxes low – followed by a series of five additional annual rollovers to Roth, ranging in size from $60,000 to $80,000. The timing and size of these withdrawals were determined by applying the optimization program of one of the major commercial spreadsheet packages to the data. The residual fund under this scenario is $1,623,261. Notice that this strategy is internally financed. George can finance annual shortfalls by drawing down the contributions he has made to the Roth IRA through his conversions. So long as George and Martha’s adjusted gross income, exclusive of their rollover amount is $100,000 or less in each Roth rollover year, there is no tax or penalty on early withdrawal of contributions for those over 59½, nor is there accelerated taxation in the absence of income averaging.
Interestingly, this yield can be further enhanced by following the rollovers with a series of ten additional withdrawals from the traditional IRA component in excess of the minimum requirement. Under this plan, the residual fund grows to $1,639,293 (Bar 7). The reason why the early removal of traditional IRA money from its sheltered status increases the residuum is because the traditional IRA is a more tax-efficient source of consumption funds than is the draw-down of the retiree’s Roth IRA balance.
Finally, Bar 8 illustrates another internal financing method available under a hybrid strategy. George converts $442,000 to Roth in 1998 leaving $758,000 in traditional IRA form. In this scenario, four-year-averaging is assumed. Shortfalls in the first five years are met by "additional withdrawals" from the traditional IRA component, ranging in size from $5,000 to $61,000. These withdrawals represent taxable income, but unlike Roth rollover withdrawals, these additional withdrawals are included in determining whether the $100K limitation on Roth conversions has been exceeded. To avoid excess adjusted gross income in 1998 -- the year of Roth rollover -- the taxpayer may seek to delay his first additional withdrawal until 1999. This internal financing strategy yields a residual fund of $1,692,709.
The hybrid internal-financing strategies (Bars 6,7, and 8) are less tax-efficient than the hybrid loan+savings-financing strategy (Bar 5), but they are far more efficient than 100% annuitization (Bar 1) or 100% IRA conversion, whether traditional (Bar 2) or Roth (Bar 3). A well designed spreadsheet program can analyze and optimize the full range of withdrawal and financing strategies for large and small accumulations and for the many cases where employees are subject to some involuntary annuitization through "defined benefit" pension plans. Typically, under these, the retiree gets periodic payments of fixed size until he, or he and his spouse, die.
Personal financial analysts generally agree that the 1997-98 Roth legislation requires an extraordinary degree of individualized retirement tax analysis and planning. While this task is easier said than done, it is doable, and is clearly worth doing.
__________________________
Jordan H. Leibman, jleibman@iupui.edu, is a Professor of Business Law at the Indiana University Kelley School of Business-Indianapolis; Neal J. Rothman, nrothman@math.iupui.edu, is Emeritus Professor in the Mathematics Department at Indiana University Purdue University in Indianapolis, and Richard B. Dull, rdull@iupui.edu, is Assistant Professor of Accounting in the Kelley School-Indianapolis.
