- Under the substantially equal periodic payment exception, the account owner must withdraw a substantially equal amount from an IRA annually for five years or until the taxpayer reaches age 59½. The amount that must be withdrawn is based on the taxpayer’s life expectancy.
- Three safe-harbor methods are available for calculating the annual withdrawal amount: (1) the required minimum distribution method, (2) the fixed amortization method, and (3) the fixed annuitization method. Each method produces a different annual withdrawal amount.
- The IRS has been willing to approve taxpayers’ requests to use reasonable variations on the safe-harbor methods to calculate their annual withdrawal amounts.
- Modification or termination of the SEPP before the end of the required time period results in a retroactive application of the 10% penalty to all previous withdrawals and interest charges from the date each withdrawal was received until the modification or termination occurred.
In the current economic climate, unexpected circumstances may cause many individuals to consider early withdrawal of IRA funds. Minimizing the tax consequences of these withdrawals means carefully considering opportunities to avoid the 10% penalty applicable to premature distributions.
While all distributions from a traditional IRA will be subject to income tax under the annuity rules in Sec. 72, an additional penalty is imposed on the taxable portion of distributions occurring before the owner reaches age 59½.1 Exceptions apply that allow the IRA owner to avoid this penalty if specific conditions are met. For example, withdrawals due to the owner’s death or disability are not subject to the penalty.2 Distributions for qualified medical expenses,3 qualified education expenses,4 first-time homebuyers,5 and health insurance premiums paid by the unemployed,6 as well as qualified hurricane distributions7 and qualified reservist distributions for the military,8 are all exempt from the penalty (subject to certain requirements). Distributions that are part of a qualified series of substantially equal periodic payments made annually to the IRA owner are also exempt from the penalty.9
Of all the techniques available to avoid the early distribution penalty, the substantially equal periodic payments (SEPP) alternative is the most universally available. It does not rely on preexisting conditions such as medical or educational expenses but is available to any IRA owner who is willing to calculate and sustain withdrawals according to the rules specified by the IRS. Thus, it provides an opportunity to fund early retirement or meet payments on long-term debt obligations without the imposition of the penalty. This can mean substantial savings on large IRA withdrawals.
This article examines the formulas provided by the IRS as safe-harbor methods for calculating the SEPP amount. Factors in the formula that are subject to discretion are highlighted, and the discussion focuses on how they may be used to adjust the payment to a preferred amount and still avoid the premature distribution penalty. Using these factors to establish the SEPP amount in response to the IRA owner’s objectives requires making choices at the beginning of the payment series.
For example, a client wishing to meet short-term financial needs with IRA distributions may prefer to maximize the SEPP withdrawal over the shortest possible time period. This objective would be met by choosing among several approved formulas, interest rates, and life expectancies to calculate the maximum payment. However, if the objective is to fund a long-term commitment such as early retirement, the client may prefer to limit the payment amount in order to extend the series over a lengthy period. In this case the same factors could be used to maximize the distribution period. Strategic choices made in establishing the SEPP will permit the IRA owner to exercise some control over the amount to be received consistently over the payout period.
SEPP CalculationUnder the SEPP exception, the account owner must withdraw substantially equal amounts from the IRA annually. The annual payment is calculated based on a period equal to the owner’s life or life expectancy, or the joint lives or joint life expectancy of the owner and his or her designated beneficiary.10 The payments must continue for a minimum period that extends to the later of five years or the owner’s reaching age 59½.11 For example, payments beginning at age 58 must continue to age 63, while payments beginning at age 50 must continue to age 59½. Modifications to the payment amount before this time period expires will trigger a retroactive application of the 10% penalty.12 The payment can be altered without penalty due to death or disability.
The amount to be withdrawn annually is calculated according to one of three methods provided by the IRS in Rev. Rul. 2002-62.13 The method chosen to calculate the first year’s distribution must be used in subsequent years. The three methods approved by the IRS include (1) the required minimum distribution method, (2) the fixed amortization method, and (3) the fixed annuitization method. There are no restrictions on the IRA owner’s choice of methods. The optimal choice will depend on the owner’s liquidity needs, life expectancy, applicable interest rates, and expected tax rates during the payout period. Once selected, the method must be applied consistently. Modification or termination of the payment amount will generally have negative consequences.
The factors that give the IRA owner discretion over the amount of annual distribution in the SEPP include choice of method, life expectancy table, interest rate, and account balance. Within allowable limits, these factors will collectively determine the payment’s amount.
The MethodsThe three methods provided by the IRS are generally interpreted as safe harbors for the payment calculation. Numerous letter rulings suggest that variations on these methods will be tolerated as long as they are consistent with the requirements of Sec. 72(t)(4). As the following discussion indicates, the three methods can yield different payment amounts, so the choice should be made carefully. An example of the payment amounts calculated under each of the three methods and the factors influencing the computation are included following this discussion.
Required Minimum Distribution MethodUnder the required minimum distribution method, the annual payment may vary from year to year. The payment amount for any tax year is determined by dividing the account balance’s value for that year by the owner’s current life expectancy obtained from one of three IRS tables.14 A separate calculation is required for each year. As the value of the account balance changes and the owner’s life expectancy declines, there will be a potential variation in the payout from year to year. This change in the amount is not considered a modification as long as the calculation method is consistently applied.
Unlike the other two methods, payment under this method is not affected by the interest rate selected for the calculation but is significantly affected by the choice of life expectancy table and by the investment return on the account. The rate of return on investment will affect the ac-count balance used in the yearly calculation, which may cause this payment to differ significantly from the payment under either of the fixed formulas.
Fixed Amortization MethodThe fixed amortization method results in a consistent annual payment throughout the series. In the first year the payment is determined by amortizing the account balance over a specific number of years (based on the life expectancy from one of the IRS tables) with the chosen interest rate. The payment amount will remain level throughout all subsequent years and thus will not be affected by the investment return in the account.15
Fixed Annuitization MethodThe fixed annuitization method also results in a level annual payment throughout the series. The annual payment is derived by dividing the account balance in the first year by an annuity factor that is the present value of an annuity of $1 per year beginning at the IRA owner’s age and extending through his or her life expectancy. The annuity factor is derived from the mortality table in Appendix B of Rev. Rul. 2002-62, with the interest rate selected by the owner.16 Unlike the other two methods, under fixed annuitization the IRA owner is not free to choose between life expectancy tables.
These methods can yield considerable variations in the payment amount. They are affected differently by the life expectancy measure and interest rate. The required minimum distribution method is influenced only by life expectancy. The annuitization method is affected only by the interest rate because it is restricted to a specific life expectancy table. The amortization method includes both the life expectancy and interest rate factors. The opportunity to choose among these unique methods provides the IRA owner with an element of control over the SEPP.
Other Relevant FactorsIn addition to selecting a suitable calculation method, there are several other variables through which the payment amount can be adjusted. There is some discretion as to the life expectancy table that will be used, the interest rate that will be applied, and the date on which the account balance to be used is established. Some restrictions and definitions apply to these factors, but the IRA owner does have some latitude in using them to establish a SEPP that falls within a specified range of payment levels.
Life Expectancy TablesThree life expectancy tables are available to determine the distribution period. They are (1) the uniform lifetime table,17 (2) the single life expectancy table,18 and (3) the joint and last survivor table.19 The first two tables will base the distribution period on a single life expectancy and the third will base it on the joint life expectancy of the owner and his or her beneficiary. The taxpayer must use the table used in the initial year consistently in subsequent annual calculations for the required minimum distribution method.
The taxpayer’s age on his or her birthday in the year of the distribution is used to obtain the correct number from the table. If the joint and last survivor table is used, the IRA owner’s age is cross-referenced with the beneficiary’s. Specific rules apply in determining who the appropriate beneficiary is for each year a calculation is made.20 The beneficiary is determined as of January 1 of the distribution year without regard to changes that may occur during the year or may have occurred in prior years. The oldest beneficiary’s age is used to derive the value from the table if more than one is named at the beginning of the year. Thus, if a 50-year-old account owner has two beneficiaries aged 25 and 55, the 55-year-old would be the designated beneficiary. The value from the joint and last survivor table in this case would be 38.3, obtained by cross-referencing the age of the 50-year-old and the 55-year-old. If the taxpayer selects the joint and last survivor table for the required minimum distribution method, then for any year in which there is no designated beneficiary, the taxpayer must use the single life expectancy table.
The discretion to choose among life expectancy values is available for the required minimum distribution method and the amortization method. The fixed annuitization method does not provide a choice among tables. Life expectancy is derived from the mortality table in Appendix B of Rev. Rul. 2002-62. The calculation of the annuity factor used in this method, which is based on the probability of being alive during any year in the remaining life span and on the present value of $1 to be received in that year, is complex.
The life expectancy table selected by the taxpayer can have a dramatic effect on the payment calculation. The shortest life expectancy is obtained from the single life expectancy table, while the longest will be obtained from the uniform lifetime table. For example, an IRA owner who is age 55 would have a value of 41.6 from the uniform lifetime table and 29.6 from the single life expectancy table.
The value obtained for life expectancy is inversely correlated to the amount of payment derived in the calculation. Thus, the choice among tables should be made in deference to the IRA owner’s goals. If the objective is to extend payments over the longest possible period, the higher value obtained from the uniform lifetime table will be preferred. It is important to note that this higher life expectancy not only results in a longer distribution period but also minimizes the annual payment relative to other choices. The single life expectancy table will generally always be appropriate if the IRA owner wishes to maximize the annual payout.
Interest RateIf the fixed amortization method or the fixed annuitization method is used, the IRA owner must select an interest rate. The rate must be no more than 120% of the federal midterm rate for either of the two months immediately preceding the initial month of the first distribution.21 This rule places a ceiling on the interest rate but does not impose a minimum rate, so a zero interest rate is also acceptable.22
These guidelines result in a range of interest rates from which the IRA owner may choose. As the federal midterm rate becomes higher, the amount of discretion available in manipulating the payment calculation increases. Selecting the highest possible interest rate within the acceptable range will maximize the payment amount. For example, if 120% of the midterm rate is 6%, the maximum annual payout on a $100,000 balance with a single life expectancy for a 55-year-old owner would be $7,301 under the amortization method. In contrast, the payment could be minimized by selecting a zero interest rate, resulting in a payment amount of $3,378.23
Although the selection of an interest rate can have a significant effect on the payout under the amortization or annuitization methods, it has no effect on the required minimum distribution method, which relies only on life expectancy and the annual account balance. The allowable interest rate is an important factor because it introduces a significant amount of variation among the three methods and thus expands the choices available to the IRA owner.
Account BalanceThe IRS provides some latitude in determining the account balance that will be used in application of the payout formulas. The IRA owner may elect the date on which the value of the account balance is determined. The balance used must be “reasonable . . . [and] based on the facts and circumstances.”24 For example, if distributions begin on July 15 for an account with daily valuations, it would be reasonable to determine the balance based on the value from December 31 of the previous year through the July 15 distribution when applying the required minimum distribution method. For subsequent years, the value on December 31 of the previous year or a date within a reasonable period before the distribution would be appropriate in applying that method.25 An IRA with a balance increasing from $100,000 on December 31 of the previous year up to $125,000 by the starting date of the distribution should provide a substantial range of values from which the IRA owner could select a starting balance.
It is important to note that under all three methods, SEPPs are calculated based on the account value as of the elected date. Any subsequent addition to the account balance other than investment gains or losses, including a nontaxable transfer of a portion of the account balance to another retirement plan or a rollover by the owner of the amount received resulting in that amount not being taxable, will be considered a modification of the series of payments and perhaps trigger the 10% penalty. 26
In the payment calculation it is not necessary to aggregate the account balances for all IRAs owned; only the balance of the IRA(s) from which the SEPP is drawn is considered. This gives some flexibility in adjusting the account balance by splitting or combining IRAs in advance to obtain the desired balance. The entire balance of the account used for the SEPP must be included in the calculation. The IRS has not permitted the use of a portion of the account balance to establish the payment amount.27 It is permissible, however, to include more than one IRA balance in the initial payment calculation. Distributions must then come from one or both of these accounts. If a taxpayer needs additional distributions, establishing a SEPP from another IRA account while simultaneously continuing an existing SEPP is not prohibited. The second series of payments must begin in a calendar year subsequent to the establishment of the existing SEPP.28
Example: P, a 55-year-old IRA owner with a $100,000 account balance, wishes to withdraw a SEPP. Assume that 120% of the federal midterm rate immediately before the initial distribution is 4%. P may choose an acceptable interest rate ranging from 0% to 4%.Exhibits 1 and 2 illustrate a range of payment calculations for P. Exhibit 1 shows the payment amount across several interest rates that would be in the acceptable range.
Exhibit 1: Payment levels across SEPP interest rates and methods
Panel A: Single life expectancy table (initial value 29.6 at age 55)
|SEPP formula interest rate||Required minimum distribution||Fixed amortization method||Fixed annuitization method|
Panel B: Uniform lifetime table (initial value 41.6 at age 55)
|SEPP formula interest rate||Required minimum distribution||Fixed amortization method||Fixed annuitization method|
Results in Panel A are calculated based on the single life expectancy table with a value of 29.6; those in Panel B are based on the uniform lifetime table with a value of 41.6. The calculations show that the required minimum distribution method is sensitive to the selection of life expectancy, but it is not affected by the interest rate selected. Both the amortization and annuitization methods are affected by the interest rate applied in the formula. Because the annuitization method requires the use of the mortality table in Rev. Rul. 2002-62, the payment amount is not sensitive to the choice of tables.
The payment amounts shown in Exhibit 1 would remain fixed throughout the series except for the required minimum distribution payment, which is recalculated for each subsequent year. Because of the fixed nature of the amortization and annuitization methods, payments are unaffected by investment performance. Under the required minimum distribution method, the value of the account balance is updated for each annual payment calculation. Thus, investment performance will affect the stream of payments generated under this method.
Exhibit 2: Five years of required minimum distribution method payments across actual rates of return on investment
Panel A: Single life expectancy table (initial value 29.6 at age 55)
|Year 1||Year 2||Year 3||Year 4||Year 5||Total|
|Value from table||29.6||28.7||27.9||27.0||26.1|
|6% investment return||3,378||3,569||3,756||3,966||4,188||18,857|
|4% investment return||3,378||3,501||3,615||3,746||3,881||18,121|
|2% investment return||3,378||3,434||3,477||3,534||3,591||17,414|
Panel B: Uniform lifetime table (initial value 41.6 at age 55)
|Year 1||Year 2||Year 3||Year 4||Year 5||Total|
|Value from table||41.6||40.7||39.7||38.7||37.8|
|6% investment return||2,404||2,542||2,694||2,856||3,019||13,515|
|4% investment return||2,404||2,494||2,594||2,697||2,798||12,987|
|2% investment return||2,404||2,446||2,495||2,546||2,589||12,480|
Exhibit 2 contrasts a five-year stream of payments under the required minimum distribution formula across several possible investment returns compounded annually. Even with a return as low as 2%, the required distribution will increase throughout this period. If the objective of the SEPP is to withdraw a series of payments over the shortest possible period to meet an immediate obligation rather than to create a supplemental income for life, the strategy should consider the expected return. The last column of Exhibit 2 shows the total payout the IRA owner could expect with the different rates of return. Although the amount is not highly variable across rates of return, it is significantly affected by the choice of life expectancy table. The average total withdrawal for the five years declines by 28% when the single life expectancy table is replaced by the uniform lifetime table.
While the payment calculation under the two fixed methods is not affected by investment return, it is noteworthy to consider the impact of these methods on the account balance. In the event of a market downturn, a large fixed payment under the amortization or annuitization method could quickly deplete the IRA account balance. The desirability of this impact would depend on the IRA owner’s objectives in establishing the SEPP. If the goal of maintaining an IRA balance supersedes the desirability of a SEPP withdrawal, it may be beneficial to take advantage of the one-time opportunity to permanently switch to the required minimum distribution method.
Both exhibits show the effect of choosing the single life expectancy table versus the uniform lifetime table. Although not included in this analysis, the joint and last survivor table would yield the same value as the uniform lifetime table when the beneficiary is 10 years younger. A lower value will be obtained if the beneficiary is less than 10 years younger (or is older) than the owner, and a greater value will be obtained if the beneficiary is more than 10 years younger than the owner. Thus, there is even more discretion in the payout calculation if there is a designated beneficiary.
The payments generated from this analysis ranged from a high of $5,824, derived from the amortization method with a 4% interest rate used in the SEPP formula, to a low of $2,404, derived from the minimum distribution formula used in conjunction with the uniform lifetime table. The IRA owner in this example would have the discretion to select payments within the ranges shown here.
A comparison of the formula results indicates that a taxpayer can use the choice of interest rate to control the payment amount under the fixed amortization or the fixed annuitization method. This same element of control is not available in the required minimum distribution formula. It is significant, however, that if the IRA owner wishes to alter the amount of the payment during the series, he or she may take advantage of the one-time switch from either of the fixed methods to the required minimum distribution method.
The discretion that may be exercised over the factors noted above provides opportunities to derive a payment schedule that will meet liquidity needs and minimize taxes to the maximum extent possible. IRA owners who want to take out distributions for the short term could begin the substantially equal series in their early to mid-50s and terminate the payments after reaching age 59½. As long as the payment series had continued for at least five years, no penalty would be applicable. Thus, it would be possible to withdraw a fixed amount within a short period of time.
Modifications and TerminationsOnce an IRA owner begins to receive a series of substantially equal periodic payments, there must be no modifications to the payment except for death, disability, or the expiration of the required minimum payment period, which is the later of five years or the owner’s attainment of age 59½. Modification or termination before the end of the requisite time period results in a retroactive application of the 10% penalty to all previous withdrawals and includes appropriate interest charges from the date each withdrawal was received until the modification or termination occurred.29
The calculation method adopted for the SEPP must be applied consistently throughout the payment period. Any alteration would be considered a modification. However, a one-time change from the fixed amortization or fixed annuitization methods to the required minimum distribution method is permitted. This change is permanent and applies to all subsequent years. It is not considered a modification.30
In general, if an IRA owner were to cease receiving distributions during the requisite time period for a SEPP, this would be considered a modification that would trigger the retroactive penalties. The IRS has noted, however, that if the assets in the account are completely exhausted as a result of following an acceptable method of determining substantially equal periodic payments, the individual will not be subject to the additional penalty and interest. The cessation of payments will not be treated as a modification of the series of payments.31
While the cautions regarding modifications and terminations must be taken seriously, the IRS has shown considerable tolerance for variations in the safe-harbor methodology when the life expectancy and interest rates used by taxpayers did not circumvent the intentions of Sec. 72(t). For example, in letter rulings the Service has allowed taxpayers to employ a method that adjusts payments from both the amortization and annuitization methods on an annual basis using an updated account balance, interest rate, and life expectancy value.32 In these cases the taxpayer was using an acceptable life expectancy table consistently and an interest rate that was updated annually and did not exceed the ceiling of 120% of the federal midterm rates. In one case, the Service allowed a taxpayer to make a 3% cost of living adjustment to the payment each year.33 The IRS appears to be willing to consider alternative approaches that are based on reasonable variations of the safe-harbor methods. However, this must be established in advance of the initial distribution.
In general, the IRS has been willing to tolerate modifications that are due to external factors beyond the IRA owner’s control. For example, letter rulings have indicated that transfers out of a SEPP IRA and subsequent reduction of payments required by a divorce decree will not be considered a modification.34 Inadvertent rollovers or makeup distributions due to erroneous calculations have been overlooked when they resulted from trustee error.35
Exhibit 3: Planning opportunities and pitfalls in establishing SEPPs
|Planning opportunities||Pitfalls to avoid|
This flexibility does not seem to apply when the IRA owner is responsible for erroneous actions. For example, when an IRA owner transferred funds out of the SEPP IRA into another IRA during the payment series and then attempted to reverse the transfer and restore the funds, this was considered a modification of the series and the 10% penalty was imposed.36 In one recent case, the 10% penalty was imposed when a rollover into a SEPP IRA occurred three months before the end of the required five-year distribution period. This ruling was made despite the fact that the IRA owner had already received the final payment in the series at the time the rollover was deposited.37 These results suggest that unless a ruling is requested in advance of initiating the SEPP or is requested because divorce or trustee error has caused a violation of the SEPP rules, the error is likely to be met with imposition of the 10% penalty.
A review of the SEPP rules and the letter rulings obtained from the IRS suggests a series of planning opportunities that can be exploited as well as pitfalls that should be avoided when establishing a SEPP. Exhibit 3 identifies some strategies and pitfalls that investors should be aware of.
ConclusionTax-deferred accounts such as IRAs provide an excellent opportunity to accumulate funds for retirement. When it becomes necessary to liquidate these accounts before age 59½, it is important to be aware of the consequences and opportunities for making early withdrawals. The tools are available to take distributions and avoid unnecessary penalties. In many circumstances, strategic planning will allow the temporary withdrawal of funds with the original balance of the account still intact at the end of the period. Such a strategy would require careful consideration of each of the factors examined here.
Linda Burilovich is a professor of accounting at Eastern Michigan University, Ypsilanti, MI. Andrew Burilovich is with the law office of Andrew Burilovich, Ypsilanti, MI. For more information about this column, contact Prof. Burilovich at email@example.com or Mr. Burilovich at firstname.lastname@example.org .
1 Sec. 72(t).
2 Secs. 72(t)(2)(A)(ii) and (iii).
3 Sec. 72(t)(2)(B).
4 Sec. 72(t)(2)(E).
5 Sec. 72(t)(2)(F).
6 Sec. 72(t)(2)(D).
7 Sec. 1400Q.
8 Sec. 72(t)(2)(G).
9 Sec. 72(t)(2)(A)(iv).
11 Sec. 72(t)(4)(A)(ii)(1).
13 Rev. Rul. 2002-62, 2002-2 C.B. 710, was issued as a modification of IRS Notice 89-25, 1989-1 C.B. 662.
14 Rev. Rul. 2002-62, §2.01(a).
15 Rev. Rul. 2002-62, §2.01(b).
16 Rev. Rul. 2002-62, §2.01(c).
17 Rev. Rul. 2002-62, Appendix A.
18 Regs. Sec. 1.401(a)(9)-9, Q&A-1.
19 Regs. Sec. 1.401(a)(9)-9, Q&A-3.
20 Rev. Rul. 2002-62, §2.02(b).
21 This must be determined in accordance with Sec. 1274(d). See Rev. Rul. 2002-62, §2.02(c).
22 Midterm interest rates are available at www.irs.gov/taxpros/lists/0,,id=98042,00.html.
23 The payment could be further minimized by combining the zero interest rate with the uniform lifetime table, yielding a payment of $2,404.
24 Rev. Rul. 2002-62, §2.02(d).
26 Rev. Rul. 2002-62, §2.02(e).
27 IRS Letter Ruling 9705033 (11/8/96).
28 IRS Letter Ruling 200033048 (5/23/00).
29 Sec. 72(t)(4)(A)(ii).
30 Rev. Rul. 2002-62, §2.03(b).
31 Rev. Rul. 2002-62, §2.03(a).
32 IRS Letter Rulings 200432021 (5/11/04) and 200532062 (5/20/05).
33 IRS Letter Ruling 9816028 (1/21/98).
34 IRS Letter Rulings 200202075 (1/11/02) and 200202076 (1/11/02).
35 IRS Letter Rulings 200503036 (1/21/05), 200628029 (7/14/06), 200616046 (4/21/06), and 200631025 (8/4/06).
36 IRS Letter Ruling 200720023 (5/15/07).
37 IRS Letter Ruling 200634033 (8/25/06).