When to Claim Social Security Retirement Benefits

Transcription

1 When to Claim Social Security Retirement Benefits David Blanchett, CFA, CFP Head of Retirement Research Morningstar Investment Management January 10, 2013 Morningstar Investment Management

2 Abstract Social Security (SS) is the largest source of retirement income for most Americans. This paper provides the reader with an overview of the SS retirement system and offers insight into key factors that should be considered when determining when to begin receiving SS retirement benefits. Five separate tests are performed, each of which considers a component that is important to the optimal claiming decision, such as life expectancy, taxes, the cost of purchasing equivalent insurance, and the benefits of the surviving spouse. Three claiming scenarios are considered: receiving benefits early (e.g., at age 62 versus 66); delaying benefits past full retirement age (e.g., age 66 versus 70); and the maximum realistic delay period (e.g., at age 62 versus 70). The results of this analysis suggest most retirees would be best served delaying SS benefits until at least full retirement age (FRA) or later, and that delayed SS benefits are especially valuable for females, married couples, retirees who expect to invest in relatively conservative portfolios during retirement, and retirees who have longer life expectancies. The effective return achieved by a retiree from making the optimal SS decision can significantly exceed the return he or she could potentially earn by investing the money received from starting benefits earlier and investing the difference, especially in today s low interest rate environment. We find the optimal Social Security claiming decision can generate 9.15% more income for a hypothetical retired married couple, which creates an annual equivalent financial planning alpha (or gamma) of +0.74% per year. The author thanks Alexa Auerbach, Jim Daley, and Francisco Torralba for helpful comments. Published in the Journal of Personal Finance in 2012, volume 11, issue 2 Page 2 of 30

3 When to Claim Social Security Retirement Benefits Social Security (SS) retirement benefits are a big deal. Social Security is the largest income source for most retirees and determining when to claim benefits is an incredibly complex decision. This paper provides an overview of Social Security retirement benefits and offers insight into some of the key factors that should be considered when determining when to claim those benefits. Five separate tests are performed, each of which considers a component that is important to the SS decision process, such as life expectancy, taxes, the cost of purchasing equivalent insurance, and the benefits of the surviving spouse. Three scenarios are considered: receiving benefits early (e.g., at age 62 versus 66); delaying benefits past FRA (e.g., at age 66 versus 70); and the maximum realistic benefit delay period (e.g., at age 62 versus 70). This analysis suggests that benefits from delaying the receipt of SS benefits can be significant. We find that females, married couples, retirees who expect to invest in relatively conservative portfolios during retirement, and retirees who have longer life expectancies are likely to benefit most from delaying SS benefits. On the other hand, retirees who have shorter life expectancies or invest more aggressively and believe they can achieve a relatively high return on their retirement portfolios would likely be better off taking SS earlier. We also find that optimal SS claiming decisions can add 9.15% more lifetime income for a married couple, based on a hypothetical married couple. This 9.15% increase in retirement income is equivalent to financial planning alpha (or gamma) of +.74% per year during the entire retirement period. The Importance of Social Security Benefits Social Security retirement benefits are the largest single income source for retirees, representing between 40% and 51% of a retiree s aggregate income (Butrica et al. (2004) and Rhee (2011), respectively). According to Rhee (2011), the bottom 25% of households (by income) will rely on SS for 87% of income, versus 74% for the middle 50%, and only 30% for the top 25%. Entmacher (2009) notes that without SS, nearly one in two seniors would be poor, whereas with SS fewer than one senior in ten is poor. Butrica et al. (2004) estimate that 91% of current retirees, 92% of near-term retirees, and 94% of baby-boomer retirees will receive SS benefits 1. The Social Security Administration (SSA) notes that 96% of workers are currently covered under SS. With respect to SS benefits, as of November 2007, the last month the SSA published its detailed monthly OASDI information 2, 73.2% of eligible workers opted to take early SS benefits versus waiting until FRA. According to Reno and Lavery (2010), of those retirees turning 62 in 2006, 43% of men and 48% of women took benefits at the earliest opportunity. This marks a decline from 1985, when 51% of men and 62% of women turning 62 that year claimed benefits at the earliest opportunity. Determining Social Security Benefits Social Security retirement benefits are based on lifetime earnings. A worker s actual earnings are increased (i.e., indexed) to account for changes in average wages since the year the earnings were received, based on 1 Either their own or their spouse s benefit 2 Page 3 of 30

4 the National Average Wage Index 3 (NAWI). The NAWI is calculated annually by the SSA based on wages subject to federal income taxes. The highest 35 years of indexed earnings for the worker are used to compute the average historical earnings. The monthly average becomes the Average Indexed Monthly Earnings (AIME). The actual monthly retirement benefit, called the primary insurance amount (PIA) is based on AIME. While the percentages of the PIA formula are set by law, the dollar amounts, or bend points, change annually based on changes in the NAWI. The bend points for the year 2012 are $767 and $4,624. The benefit received is 90% of the AIME amount under the first bend point, 32% of the AIME amount between the first and second bend point, and 15% of the AIME amount over the third bend point. The maximum possible AIME value for 2012 is $9,175, which means the maximum SS benefit at FRA is $32,851 per year. Full retirement age ( FRA ) is the age when someone is eligible to receive full SS benefits. FRA for SS retirement benefits is age 66 for individuals born between 1943 and This group consists of individuals approximately between the ages of 58 and 69 as of This is the cohort of retirees currently facing the decision of when to take SS benefits 4 and therefore is the focus group for this research. Someone can elect to receive SS retirement benefits as early as age 62. The impact on the benefit from claiming benefits at various ages is shown in Table 1, where the FRA is assumed to be age 66. Table 1: Social Security Benefits Assuming a Full Retirement Age of 66 % of FRA % of Earliest Example Retirement Age Benefit Benefit Differences ($) * , , , , ,320 * Full Retirement Age Source: Social Security Adminstration Someone who elects to receive SS at age 62 will receive a benefit that is 75% of the value he or she would have received if he or she waited to FRA (age 66). Postponing retirement past age 70 is never optimal from an income perspective, since SS benefits will not increase past that age. For an individual who is younger than FRA and continues to work after claiming SS benefits, the benefits may be reduced in proportion to the individual s labor earnings. In particular, if the worker is less than FRA, the SSA will deduct $1 from the benefit payments for every $2 earned above the annual limit. For 2012, the annual limit is $14,640. In the year the individual reaches FRA, the SSA will deduct $1 in benefits for every $3 earned in excess of a higher limit ($38,880 in 2012). Starting in the month the individual reaches FRA, SS benefits are no longer capped Which is why age 66 is the assumed FRA for this analysis. Page 4 of 30

5 Social Security Benefit Taxation SS retirement benefits may be subject to taxation based on the total income of the retiree. Currently, no more than 85% of someone s SS retirement benefits can be subject to federal taxation. The rules regarding the taxation of SS benefits are outlined in IRS Publication 915. The amount of SS benefits subject to taxation is based on the recipient s combined income. Combined income is calculated by adding adjusted gross income to nontaxable interest plus half of the SS benefit. The tax thresholds to determine the amount subject to taxation were determined in 1983 with the idea that only the wealthy would pay taxes on their SS retirement benefits. SS benefits are not taxed up to the first threshold, but after the first threshold, up to 50% of total SS income is subject to federal taxation. This tax is applied to the entire amount of SS income, not just the incremental income above the threshold. After the second threshold, up to 85% of SS income is subject to taxation. The thresholds are not indexed for inflation, and are $32,000 and $44,000 for married couples, and $25,000 and $34,000 for single individuals (as of 2012). Slight changes in income can subject a SS benefit receipt to what is known as a tax torpedo, where the marginal tax rate on SS increases dramatically at the income thresholds. For example, a single worker with a combined income of $24,950 would be subject to no tax on his or her SS benefits. If the worker received an additional $50 in income, then 50% of the SS benefits would be subject to federal taxation. Therefore, it is very important to be aware of the total income of an individual and the respective tax thresholds. Taken to the extreme, if all of someone s income (e.g., $30,000) were SS income, he or she would be subject to no tax. In contrast, if this worker received the same $30,000 as IRA income, some tax would likely be due. Spousal Benefits Even if the SS recipient s spouse has never worked, his or her spouse is eligible for SS retirement benefits, equal to one-half of the primary worker s retirement amount at the spouse s FRA. Whereas the spouse s benefit is based on the primary worker s benefit, the spousal benefit amount is not determined by when the primary worker claims SS, but rather by when the spouse begins claiming benefits. If the primary worker elects to receive benefits early, the spouse can still receive his or her full spousal benefits by waiting until full retirement age. If the spouse is eligible for retirement benefits based on his or her own work history, he or she will receive the greater of his or her own earned amount or the amount based on the spouse s benefit. If the spouse is eligible to receive a pension for work not covered by SS, such as government or foreign employment, the amount of her SS benefits may be reduced. If a spouse has never worked under SS, she can begin collecting benefits as early as age 62. However, if the benefit begins early, the amount will be permanently reduced. A spousal benefit is reduced 25/36 of one percent for each month before normal retirement age, up to 36 months. If the number of months exceeds 36, then the benefit is further reduced 5/12 of one percent per month. This reduction factor is applied to the base spousal benefit, which is 50% of the worker s primary insurance amount (PIA). The impact on the PIA amount varies with the primary worker s PIA itself, whereby the reduction could potentially be greater than 25% 5. If the primary worker elects to receive benefits early while the spouse waits until FRA to begin receiving benefits, it is possible for the spouse to receive $0.67 for every $1.00 received by the primary worker, which is greater than the base 50% benefit. 5 If the worker s primary insurance amount is $1,600 and the worker s spouse chooses to begin receiving benefits 36 months before his or her normal retirement age, we first take 50% of $1,600 to get an $800 base spousal benefit. Then we compute the reduction factor, which is 36 times 25/36 of one percent, or 25%. Applying a 25% reduction to the $800 amount gives a spousal benefit of $600, which is 37.5% of the primary insurance amount. Page 5 of 30

6 Spousal benefits do not include any delayed retirement credits the primary worker may receive; therefore, there is no advantage to waiting to begin collecting spousal benefits after the spouse reaches FRA. If the primary worker is at FRA, he or she can apply for retirement benefits and then request to have payments suspended. That way, the worker s spouse can begin receiving a spouse s benefit and the primary worker can continue to earn delayed retirement credits until age 70. If the spouse has reached full retirement age and is eligible for a spouse s benefit and his or her own retirement benefit, he or she has a choice. The spouse can choose to receive only the spouse s benefit now and delay receiving his or her own retirement benefits until a later date. If retirement benefits are delayed, a higher benefit may be received at a later date based on the effect of delayed retirement credits. Survivor Spousal Benefits While the spousal benefit is not affected by when the primary worker claims SS, the spousal survivor benefit is affected, because the spousal survivor benefit is based on the respective benefit of the spouse. A widow or widower who is at his or her FRA or older will receive the maximum of his or her own benefit and 100% of the deceased worker s benefit amount, as long as the couple was married at least 10 years at the time of the primary worker s death. According to the SSA website 6, at present there are approximately five million widows and widowers receiving monthly SS benefits based on their deceased spouse s earnings record. SS provides 58% of the income for widows 65 and older, compared to 39% for all individuals and couples 65 and older (Entmacher 2009). Survivor spouses are eligible to receive benefits as early as age 60. For those spouses with an FRA of 66 (i.e., those born between 1943 and 1954), the benefit is reduced.396% for each month before FRA that the applicant takes the benefit. Therefore, if the spouse elects to take spousal benefits at the earliest age possible (age 60), the benefit, based on primary worker s PIA, will be reduced to 71.5% of the initial amount. Survivor spousal benefits are an important consideration for the primary worker when he or she claims for SS because they represent a residual benefit available should the spouse outlive the primary worker. Even if each member of a couple is receiving a benefit based on his or her own earnings, the survivor spousal benefit would still be important for whichever spouse has the smaller benefit, especially if that spouse has a longer life expectancy. Literature Review There is significant disagreement in past research on what the optimal claiming age is for SS benefits. Determining the optimal age for SS requires complex calculations given the number of variables that must be considered. Therefore, no single study can be expected to provide an answer for every case. A slight change in some of the assumptions (e.g., future tax rates, market returns, life expectancy, future benefit changes, etc.) can dramatically affect the result. This section includes an overview of past works to provide the reader with insights about what authors and experts have determined. Mahaney and Carlson (2007) examined a strategy in which retired individuals bought an annuity with personal retirement savings to bridge the period from age 62 to 70 and delayed claiming SS until age 70. This strategy out-performed a strategy of taking SS early while preserving funds in their other retirement savings. As reasons for this finding, they point to the robust increase in SS benefits due to delayed claiming; the expenses (such as commissions, management fees, and advisor charges) related to IRA investments; the tendency for retirees to invest more conservatively as they age (thus reducing investment returns); and the tax treatment of SS benefits versus other retirement income. 6 Page 6 of 30

7 Meyer and Reichenstein (2010) note that for single taxpayers with average life expectancies who will not be subject to an earnings test, the present value of SS benefits is approximately the same no matter when benefits begin. This supports the premise that SS benefits have generally been noted to be actuarially fair. They suggest that singles with short life expectancies should begin benefits early and those with longer life expectancies should delay. For an average couple, on the other hand, the decision revolves around spousal and survivor s benefits. The present value is usually maximized when the lower-earning spouse begins benefits as soon as possible (as long as those benefits would not be lost due to the earnings test), while the higher-earning spouse delays benefits until age 70. Meyer and Reichenstein (2012) review the implications of delayed benefits on a portfolio and find that delaying benefits can add more than ten years of longevity to a portfolio. They note that the additional longevity from delaying SS decreases for higher levels of wealth. Therefore, less-wealthy clients concerned with longevity risk should be especially interested in delaying SS benefits. They note that if early retirement is desired, one should wait until age 64, but if an individual does not retire at age 64, then he or she should retire no later than age 67. Ryan (2010) notes that the candidates who have the potential to gain the most from delaying (or resetting, if benefits have already started) SS retirement benefits would likely have one or more of the following characteristics: first, sufficient confidence in the ability to survive the break-even period; second, an aversion to market risk and a desire to trade riskier investment assets for an increased SS benefit; third, a wish to minimize the risk of outliving their portfolio; and finally, adequate non-tax-deferred assets from which to repay past benefits. With respect to the exact age at which to claim SS benefits, Rose and Larimore (2001) find age 62 is the optimal age for men and women. Munnell and Soto (2007) find the optimal age to be 62 for men and 68 for women. Sun and Webb (2009) find the preferred retirement age to be 62 or 69 for men and 67 or 70 for women depending on their risk aversion. McCormack and Perdue (2006) find age 66 to be the optimal retirement age. Cunningham and Erickson (2009) note age 62 is the optimal age for males with income less than $30,000, or 66 if greater, and 66 for females (regardless of income). Docking, Fortin, and Michelson (2011) calculate that the optimal age is 64 for men and 67 for women. When to Start Taking Benefits For single workers, the decision of when to claim for SS benefits is less complex than for married couples, since spousal survivor benefits are not a concern. A common approach used to indicate whether it is better to take benefits early or to delay benefits is a break-even analysis. A break-even analysis produces a value that equalizes the benefits of retiring to those of delaying retirement. The break-even calculation can be solved in terms of age, return, or a combination of the two. If the break-even is expressed as a life expectancy value, the worker should delay claiming SS benefits if he believes he will outlive the break-even period. In contrast, if the break-even is expressed as a return, the worker should take benefits early if he believes he can achieve a higher return by taking the payments earlier and investing that money. Here is a list of important considerations that will affect each individual differently; each of these considerations will be reviewed in the following sections: Page 7 of 30

8 1. Ability to delay benefits. In order to delay SS benefits, an individual or couple will need sufficient assets to provide income during the pre-ss years of retirement. 2. Life expectancy. This is important for both the primary worker and the spouse. Generally, the healthier someone is, the more beneficial it can be to delay receiving benefits, since the retiree is expected to live longer during retirement. Along these same lines, a younger, healthy spouse also increases the potential value of delayed SS benefits due to spousal survivor benefits. 3. Tax considerations. As previously reviewed, SS benefits have a unique tax structure that make them superior (if only slightly at higher income levels) to other retirement income, such as a traditional IRA. Slight changes in income can subject a retiree to a torpedo tax, dramatically increasing the tax bill. 4. Benefit reduction considerations. The Senior Citizens Freedom to Work Act of 2000 allowed seniors to file and suspend their benefits upon reaching FRA, which enables the benefits of a worker to continue to accrue delayed retirement credits (DRCs). 5. Spousal considerations: the age of the spouse relative to the primary worker is important as noted in number two above. This is perhaps one of the most complex considerations. While the election to receive benefits is not irrevocable, reversing the decision is allowed only during this first year of receiving benefits. In those instances, the retiree must also pay back to the government what he or she has already received. A retiree could then start receiving benefits at some later date, thereby increasing his or her SS benefit. An important consideration for retirees who want to delay retirement (something that is highlighted numerous times on the SSA website), is Medicare benefits. If one plans to delay receiving benefits and continue working, it is in the worker s interest to sign up for Medicare three months before reaching age 65, regardless of when he or she reaches full retirement age. Otherwise, Medicare medical insurance, as well as prescription drug coverage, could be delayed, and one could be charged higher premiums. Medicare is also an important consideration when thinking about the actual benefit a retiree will receive since the Part B monthly premium is automatically deducted from the SS benefit. These premiums range from $99.90 (the standard premium) to $ in 2012, depending on the recipient s modified adjusted gross income (MAGI). Note that this does not include the prescription drug monthly amount, which can be as high as an additional $66.40 per month (in 2012). While SS benefits are received monthly, research on benefit maximization typically assumes annual benefit payments, which will be the primary approach used in this analysis. This is both for simplicity purposes and because mortality tables only include annual survivorship rates. While Docking, Fortin, and Michelson (2011) make note of the inaccuracy of a number of past research studies on optimal SS retirement claiming ages that assume annual versus monthly benefits, they neglect to mention the insignificance of the difference. In a 62 vs 66 analysis (discussed next) the internal rate of return (IRR) using annual SS benefits is 9.10% versus 9.14% using monthly values. We would contend a 4 bps difference is insignificant in light of the other material assumptions that need to be made (e.g., inflation, mortality rate, taxation, other forms of income, etc.) when determining the optimal age to claim SS. Page 8 of 30

9 Analysis There is a large number of potential scenarios that could be tested to determine the optimal age at which to claim SS benefits. For this analysis, though, there will be three primary test scenarios: retiring at age 62 versus age 66, 62 vs. 70, and 66 vs. 70. The first test (62 vs. 66) determines the potential benefit of taking benefits at the earliest possible age versus delaying to FRA. The second test (62 vs 70) determines the potential benefit of taking benefits at the earliest possible claiming age versus delaying as long as reasonably possible to create the largest possible difference (76% higher income than at age 62). The final test (66 vs 70) determines the potential benefit of taking benefits at FRA versus delaying as long as possible. These three scenarios cover the three most significant choices faced by a retiree. They assume that the retiree invests all benefits. Unlike a number of past studies, a standard (baseline) scenario is not created for the analysis for comparison purposes. Instead, five different tests are conducted, each of which seeks to provide insights about the variables that affect the optimal claiming age. The first test is a simple break-even analysis that determines how much an investor would need to earn on his or her invested benefits to be indifferent between claiming ages. The second test is a break-even analysis that incorporates taxes to find out how many years the retiree would need to survive in order to break even based on different after-tax income goals and levels of SS income. The third test incorporates mortality into the analysis and determines the break-even returns necessary for a single worker (i.e., does not consider potential spousal survivor benefits). The fourth test determines the return required in different scenarios in order to purchase an inflation-adjusted immediate annuity (i.e., to self-fund the incremental difference in lifetime income by investing the early SS benefits versus receiving the delayed benefit). Finally, the fifth test calculates the value of spousal survivorship benefits, which are an important consideration for married couples. It s important to note that in the second test that incorporates taxes, we need to make an assumption about the rate of return the SS recipient earns on invested benefits. This is significant because the higher the return the longer it takes to break even. One potential starting place to determine what is a reasonable return assumption is to look at how older Americans are actually allocating their financial assets (i.e., portfolios). Median equity allocations for US households from age 60 to age 90 (i.e., retirees), obtained from the 2010 Survey of Consumer Finances are included in Figure 1. Page 9 of 30

10 Figure 1: Median Equity Allocations for Retirees Median Equity Allocation (%) y = x R 2 = Source: 2010 Survey of Consumer Finances Age The average equity allocation for Americans from age 60 to 95 is approximately 24%, although the equity allocation tends to decrease with age. A portfolio with a 24% equity allocation (and 76% bond allocation) is expected to have a nominal (before inflation) return of approximately 5%, based on Morningstar s current 20-year market forecasts (as of June 2012), with long-term inflation estimated to be approximately 3%. For the second test, we will assume a nominal return of 5% as the base return assumption (which equates to a real or inflation-adjusted return of approximately 2%). Some readers may question a 5% base return as too aggressive given the current interest rate environment (with government bonds yielding less than 2%), while others may contend it is too conservative. In either case, it is important to note that this 5% would be the after-fee total return of the portfolio. An investor who expects to achieve an annual return of 6% but pays 1% total annual money management fees would achieve a net return of 5%, which is equivalent to the base assumption. Also, most of the tests include the break-even returns that would need to be achieved, and are therefore not based on this 5% nominal return value. Break-Even Internal Rate of Return The primary metric used to demonstrate the potential benefit of delaying benefits is the break-even return. The break-even return is the annual compounded nominal after-fee return the investor would need to achieve on his or her invested benefits over a given time to be indifferent between claiming benefits early or delaying. For example, Figure 2 shows that in the 62 vs 66 analysis, the internal rate of return is -10.1% at year 10. This means that the investor who took benefits at 62 would have to have an average annual loss of 10.1% to be even with someone who delayed benefits until age 66 if they both died at age 72. The longer an investor lives, however, the more attractive it is to delay benefits. Had these investors lived to 92, the one who took benefits at 62 would have had to earn almost 10% per year on the invested benefits to be even with the investor who delayed benefits until age 66. Page 10 of 30

11 Figure 2: Internal Rates of Return for Various Time Periods vs vs vs 70 Internal Rate of Return (%) Years Survive After Receiving Initial Benefit One of the most noticeable points in Figure 2 is how much longer it takes for the 62 vs 70 scenario to break even. The reason for the longer time period has to do with the length of the delay from when benefits can be received initially (age 62) versus eventually (age 70). This eight-year delay requires a significant amount of time to overcome. The required returns for the 62 vs 66 and the 66 vs 70 are virtually identical, because they are both four-year periods and require a much a shorter break-even point than the 62 vs 70 scenario. This can be attributed to the four-year gap from the early to delayed benefits, versus the eight-year period for the 62 vs 70 scenario. While determining the break-even period provides useful information, it is in many ways overly simplistic because it does not consider more complex items, such as taxes, life expectancy, or spousal survivor benefits. Life expectancy is especially important, though, because while the retiree may be a little worse off if he or she passes away early, the implications of a lower SS benefit are significant for those retirees who end up living in retirement for a long time. Break-Even Analysis With Taxes Taxes are important because not all types of income are treated equally (or really taxed equally) during retirement. Social Security income has a more favorable tax treatment than other forms of income (such as income from a traditional IRA), and the implications of this differential tax treatment also vary by retiree. For this section, we examine the impact of delaying benefits under the assumption that the retiree must pull money from an IRA, which has less favorable tax treatment than Social Security, to fund retirement during the delay years. All tax rates and tax assumptions (e.g., exemptions, standard deductions, etc.) are based on 2012 federal tax rates. State taxes are ignored. We assume that the individual or couple takes the standard deduction for the respective filing status (either single or married filing jointly) and has no additional income, dependents, etc., other than the base assumptions. Similar to the previous analysis, spousal survivor benefits are ignored. While in the previous break-even analysis we solved for the return an investor would need to achieve to be indifferent about when to take benefits, in this analysis we solve for the number of years the retiree would Page 11 of 30

12 need to survive to be indifferent. We assume a 5% nominal return on the IRA based on achieving a target level of inflation-adjusted after-tax income every year. The desired after-tax income amount is based on the initial income for the first year, which is then increased by inflation (3%) during retirement. The SS payment amounts are the other key base assumption, where different amounts of SS income are assumed for different after-tax income targets. The remaining variable, or plug, is the value that must be withdrawn from savings in order to fund the remaining target after-tax income need. For this analysis we assume all savings are in a Traditional (pre-tax) IRA. The amount required to be withdrawn from the IRA is the difference between the SS benefit and the target after-tax income goal. The three base scenarios are conducted for a single couple and married couple, since the tax implications differ between the two household types. The results from the analysis are included in Table 2. Table 2: Break-Even Social Security Years Incorporating Taxes Total Annual Social Security Benefit at Earliest Age (000s) $10 $20 $30 $40 $50 $60 $10 $20 $30 $40 $50 $60 Married Couple 62 vs 66 Single 62 vs 66 Target After-Tax Income (000s) $ n/a n/a n/a n/a n/a n/a n/a n/a $ n/a n/a n/a $ n/a $ Target After-Tax Income (000s) Married Couple 62 vs 70 Single 62 vs 70 $ n/a n/a n/a n/a n/a n/a n/a n/a $ n/a n/a n/a $ n/a $ Target After-Tax Income (000s) Married Couple 66 vs 70 Single 66 vs 70 $ n/a n/a n/a n/a n/a n/a n/a n/a $ n/a n/a n/a $ n/a $ As the reader can see from Table 2, the number of years required to break even can differ significantly within the same base scenario (e.g., 62 vs 66, 62 vs 70, and 66 vs 70) based on tax implications. For example, the minimum number of break-even years in the 62 vs 66 scenario was 13.6 years for a couple targeting an aftertax income of $60,000 with $40,000 in SS benefits. In contrast, in that same scenario, the maximum number of break-even years was This difference (7.3 years) represents an increase of 53.7% in years from the minimum (13.6 years) to the maximum scenario. This is due to the tax torpedo that causes benefits to be taxed at higher rates at certain income levels. The disparity of years to break even was similar among the single scenarios. Taxes had virtually no impact for the single scenarios with after-tax incomes above $60,000 or the married filing jointly with after-tax incomes above $80,000. However, taxes were material in some cases below those levels, because benefits would be subject to full taxation at this point. Page 12 of 30

13 Introducing Mortality The previous two sections explored the potential benefits of delayed SS based on an IRR calculation and then the impact of taxes was reviewed. In this section, the implications of life expectancy (i.e., mortality experienced) are reviewed given different rates of return on invested benefits. Life expectancy is an important consideration for anyone deciding whether to delay the receipt of SS benefits because he or she must live long enough after the potential initial beginning age (i.e., age 62) to make the delay worthwhile. For this analysis, life expectancies were based primarily on the 2007 SSA Periodic Life Table, 7 but we also consider the life expectancies from the Society of Actuaries Annuity 2000 Mortality Table. The SSA Periodic Life Table contains the life expectancies for the average American, while the Annuity 2000 Mortality Table includes life expectancies based on the longevity of healthier people who are more likely to purchase annuities. Kreuger (2011) has noted that the Annuity 2000 Mortality Table is often used as the basis for mortality projections for healthier, more affluent (middle to upper class) populations. The differences in the respective survival probabilities for the two mortality tables are noted in Appendix I. While SS benefits are the same regardless of gender, this analysis will focus on males and females separately. Females have longer life expectancies than males, and therefore the potential benefits from delaying retirement should be higher for females. This analysis does not consider potential survivor benefits, since these are reviewed separately. For the analysis, the same three base scenarios are tested (62 vs 66, 62 vs 70, and 66 vs 70). Since inflation is assumed to be a constant 3%, the only variable that changes within each simulation is how long the retiree is assumed to live. If we change the discount rate within a simulation and run a large number of simulations, we can determine, for each discount rate, the probability of an individual being better off delaying SS retirement benefits or taking them earlier. These results are included in Figure 3. Figure 3: Probability of Being Better off Delaying Social Security Benefits for Various Returns Probability of Being Better off Delaying Benefits (%) Male Female vs vs vs 70 Nominal Rate of Return (%) Nominal Rate of Return (%) The higher the return the retiree expects to achieve on his or her invested benefits, the better off he or she is likely going to be taking SS benefits earlier versus later. The results from Figure 3 suggest that an investor must believe he or she will achieve returns around 7.0%-8.3% or higher in order to be better off receiving SS benefits early (at age 62) versus delaying benefits to FRA (age 66). In contrast, in the 66 vs 70 scenario, the investor would only need returns in the 4.6%-6.6% range or higher to be better off delaying benefits. This is because despite the fact they are both four-year periods, the mortality experience changes for older individuals, affecting the required returns. 7 Page 13 of 30

14 It should not surprise the reader that the required returns are higher for females when compared to males, because females have longer life expectancies, on average, than males. We can use the values in Figure 3 to determine various break-even returns, which is the required compounded nominal return on invested benefits where the retiree would be better off delaying benefits 50% of the time. The break-even return for the 62 vs 66 scenario was 7.0% for males versus 8.3% for females. A more complete set of break-even returns are included in Table 3. As a reminder, the base break-even assumption is based on a 50% probability of being better off. Table 3: The Break-even Return to Be Better off Delaying Benefits for Various Life Expectancy Probabilities Probability of Being Better off Delaying Benefits 75% 60% 50% 40% 25% 75% 60% 50% 40% 25% Choice Male Female 62 vs vs vs Mortality based on the Social Security Adminstration 2007 Periodic Life Table Probability of Being Better off Delaying Benefits 75% 60% 50% 40% 25% 75% 60% 50% 40% 25% Choice Male Female 62 vs vs vs Mortality based on the Society of Actuaries 2000 Annuity Table Table 3 includes the required returns for various probabilities for both the SSA 2007 Periodic Life Table as well as the Society of Actuaries 2000 Annuity Table. As noted previously, the 2000 Annuity Table includes mortality rates for healthier Americans (i.e., those who would generally consider purchasing annuities). Since the life expectancies are longer, the required returns are higher (since the higher benefits from delaying retirement are going to be received for a longer time period, on average). Stated differently, the longer the retiree is expects to live, the higher the return he or she must be able to earn on the invested benefits in order to claim earlier. The break-even rate serves as a hurdle rate, and if the retiree does not believe he or she can meet or exceed the break-even rate, delaying benefits is going to be the best option. Focusing on the 50% probability, i.e., the return required for the retiree to be indifferent between delaying benefits, all but one of the three tests (66 vs 70) had a required return less than the base target (5%). Since the required return for the 62 vs 66 is the highest this is the scenario where the retiree will benefit most from delaying benefits. These results suggest a retiree captures the greatest benefit from delaying benefits from 62 to 66, followed by delaying benefits from 62 to 70, and then 66 to 70. Insurance Perspective Social Security can be thought of as a form of insurance, where inflation-adjusted income is guaranteed for life by the U.S. government. One potential option for a retiree who is interested in guaranteed lifetime income, but who would like to invest the money for the short-term, is to claim benefits early, invest them for a time, and then use them to purchase an immediate annuity that has annual benefits that are adjusted for Page 14 of 30

15 inflation. In this section, we examine the return an investor would need to earn on benefits taken early to buy a large enough annuity to break even with a person who just delayed benefits. This analysis, and the resulting required returns, will differ from the previous test because the returns for this test are the annual compounded returns (after fees) the investor would need to achieve on invested benefits over just the period where the benefits are delayed. For example, in the 62 vs 66 scenario this would be a four-year period. In the previous analysis, the required return is the return that needs to be achieved on invested benefits over that individual s entire retirement. Since SS benefits are increased annually for inflation, the immediate annuity benefits must also be increased for inflation. Automatic payment increases, or cost-of-living adjustment riders, are relatively rare in immediate annuities purchased in the United States. According to a report by LIMRA (2010), only 7% of immediate annuity contracts issued in 2008 and 2009 included any type of payment increase rider. One company that offers immediate annuities with inflation protection (tied to the Consumer Price Index, which is the inflation factor for SS) is Principal. The costs of these annuities are available online 8 and can be used to determine the potential cost (or benefit) from delaying the immediate annuity purchase decision. Figure 4: Income Rates for Inflation-Adjusted Immediate Annuities Initial Benefit as a Percentage of Cost (%) Male Female 100% Joint and Survivor (J&S) Source: Principal Age Figure 4 includes the cost of an immediate inflation-adjusted annuity for a male, female, and joint couple (male and female, the same age, with 100% survivor benefit) as of August 1, These rates reflect the income as a percentage of the cost of the annuity. Given the payout rates, it is possible to determine the annual compounded return (net of fees) an investor must achieve to purchase an annuity at the delayed date to provide the same amount of guaranteed income. Note, this analysis does not consider the tax implications of an annuity, which is going to be less advantageous than SS, but it does provide insight as to the return required to generate the same benefit time (i.e., guaranteed inflation-adjusted income for life). This is a concept similar to what has been explored by Sass (2012) in a piece titled, Should You Buy an Annuity from Social Security? 8 It is worth noting the Principal rates are very completive. The author received quotes from Vanguard for different insurance companies offering inflation adjusted annuities and Principal generally offered the highest payout rate. Page 15 of 30

16 For example, the inflation-adjusted annuity income rate for a 66-year-old male is 4.83%. Therefore, every $1 of inflation-adjusted lifetime income will cost times that amount at age 66 (1 / 4.83% = 20.69). If we assume a monthly benefit of $750 is available at age 62 versus $1,000 at age 66, we can determine how much an investor must earn on that $750 per month over the four years in order to purchase an annuity at age 66 to provide $250 ($1,000-$750) per month in inflation-adjusted income for life. Continuing with the previous example, if we assume an annual inflation of 3%, the $250 initial difference in monthly lifetime income at age 62 will increase to $281 in four years (at age 66), or to $3,368 on an annual basis ($281*12=$3,368). Given a purchase cost of times (calculated in the previous paragraph), the retiree would need $69,686 to purchase an annuity at age 66 to create the same income he or she would have achieved by simply delaying benefits to full retirement age. The compounded annual return, net of fees, required for $750 to grow to $69,686 over four years is 31.73%. In other words, in order for an individual to be better off taking benefits early and purchasing an inflation-adjusted annuity versus delaying benefits until full retirement age, he or she would need to earn 31.73% per year on the invested early benefits to eventually purchase a large enough annuity given current annuity rates. It is worth noting that while a 3% assumed inflation rate is embedded in the primary analysis, if we remove inflation (i.e., assume the benefit payments stay constant) the required return, which would be the real or inflation-adjusted return, decreases to This could be approximated linearly by just dividing the nominal return result by the inflation rate ((( %)/(1+3%))-1) = 27.89%, which is approximately the real return of 28.02%. Both nominal returns (assuming a 3% inflation rate) and real returns results are presented in Table 4. Nominal returns are the primary returns used for the analysis, though, because they are the way most people think about market returns and are less abstract than real returns. Unlike the previous analysis that assumes annual benefit payments, this analysis assumes monthly cash flows due to the limited time horizon to invest the money. The current annuity rates are obviously very important because they determine the future cost of not delaying SS retirement benefits. Given the fact rates are currently near historic lows for annuities, the cost of this insurance is relatively high from a historical perspective. Should interest rates increase, the return required to purchase an annuity would decrease, ceteris paribus. In order to determine how the required return changes based on varying inflation-adjusted annuity income rates, different rates are tested, whereby the annuity rate is assumed to change. These results are included in Table 4. Page 16 of 30

17 Table 4: Returns Required to Purchase and Inflation-Adjusted Annuity with Early Social Security Benefit Income Change From Current Rates Annual Compounded Nominal Returns Required to Purchase an Equivalent Inflation-Adjusted Immediate Annuity at the Delayed Benefit Age (Inflation = 3%) 62 vs vs vs vs vs vs 70 Male Female -0.5% % 31.7 * 15.7 * 20.3 * 36.6 ** 18.1 ** 25.7 ** 0.5% % % % Annual Compounded Reall Returns Required to Purchase an Equivalent Inflation-Adjusted Immediate Annuity at the Delayed Benefit Age 62 vs vs vs vs vs vs 70 Change From Current Rates Male Female -0.5% % 31.7 * 15.7 * 20.3 * 36.6 ** 18.1 ** 25.7 ** 0.5% % % % * 66 year old male rate = 4.83% ** 66 year old female rate = 4.44% 70 year old male rate = 5.72% 70 year old female rate = 5.18% Source: The results in Table 4 suggest that it is highly unlikely a retiree will be able to invest the early SS benefit money and purchase an inflation-adjusted immediate annuity at the delayed benefit age given current annuity rates. The required annual nominal compounded returns are 31.7%, 15.7%, and 20.3% given current rates for the 62 vs 66, 62 vs 70, and 66 vs 70 scenarios, respectively. Even if rates increase by 1.0%, the required nominal returns to purchase an inflation-adjusted immediate annuity (ceteris paribus) would still be 21.4%, 11.7%, and 11.8%, respectively. Similar to previous tests, the highest required return (i.e., where the insurance is most valuable) is for the 62 vs 66 scenario. This suggests an individual interested in guaranteed lifetime income is better served delaying retirement, at least to FRA, in order to receive a higher eventual SS benefit versus trying to purchase an equivalent annuity. Of note is the fact the scenario with the lowest required return was 62 vs 70, which can partially be attributed to the decreasing effect of interest rates for older individuals. Modeling Spousal Survivor Benefits While spousal benefits are not directly affected by the primary worker s claiming age, spousal survivor benefits are going to be affected because the surviving spouse is entitled to the greater of his or her benefit or his or her spouse s benefit, assuming they have been married 10 years. Therefore, spousal benefits are an important consideration for the primary worker when determining at what age to claim SS benefits, because the potential benefit of the payments plus the potential residual benefit must be considered for a married individual (when his or her benefit is higher than his or her spouse s) versus just the potential benefit of the payments for a single individual. Page 17 of 30

18 In order to provide some insight as to the value of spousal survivor benefits, a Monte Carlo simulation was performed. Similar to the previous single mortality analysis, the SSA 2007 Periodic Life Table is used as the primary table for mortality rates, although the Society of Actuaries Annuity 2000 Mortality Table is also considered, because it better approximates the life expectancies for healthier Americans. Life expectancies are an important consideration when modeling spousal survivor benefits because the spouse will only receive a spousal survivor benefit if he or she outlives his or her spouse. Therefore, the spousal survivor benefit is going to be more valuable for a primary worker with a spouse that is much younger (and female) versus a spouse that is older (and male). The first analysis to determine the potential value of spousal benefits is similar to the analysis conducted for Table 3, where the required nominal compounded return is determined based on various probabilities of being better off delaying benefits. However, unlike Table 3, this analysis incorporates a potential survivor spousal benefit when determining the break-even return. The survivor benefit only has value, though, if the spouse lives longer than the primary worker. The first test assumes the spouse receives 100% of the delayed benefit. In this case, for example, for the spouse to receive 100% of the delayed benefit he or she would need to be receiving a SS benefit that is equal to or less than the benefit the primary worker would receive at the earlier retirement age considered for the scenario. For example, if the primary worker is eligible to receive a benefit of $750 per month at age 62 or $1,000 at age 66, the spouse would need to be receiving a benefit that is equal to or less than $750 for this test. For this test, spousal ages ranging between the spouse being eight years younger and eight years older are tested. Within a given simulation, the spouse is assumed to begin receiving the spousal survivor benefit as early as possible following the death of the primary worker. This could be as early as age 60, which is the earliest a person can receive spousal survival benefits, given an assumed FRA of 66. The age the survivor spousal benefits begin (if at all) is going to vary by simulation. The simulations are based on the idea that either the primary worker is male or female. If the primary worker is assumed to be male, the spouse is assumed to be female, and vice versa. This approach allows us to model the different potential spousal combinations that exist. Table 5 contains the results of this analysis. Page 18 of 30

19 Table 5: Break-even Return Required to Be Better off Delaying Benefits for Various Simulation Probabilities Assuming Spouse Receives Full Incremental Increase of the Delayed Benefit Probability of Being Better off Delaying Benefits 75% 60% 50% 40% 25% 75% 60% 50% 40% 25% Spouse Age Difference Versus Early Claiming Age Male is Primary Worker: 62 vs 66 Female is Primary Worker: 62 vs * * Younger Spouse Older Spouse Spouse Age Difference Versus Early Claiming Age Male is Primary Worker: 62 vs 70 Female is Primary Worker: 62 vs * * Younger Spouse Older Spouse Spouse Age Difference Versus Early Claiming Age Male is Primary Worker: 66 vs 70 Female is Primary Worker: 66 vs * * * Same Age Break-Even Return Younger Spouse Older Spouse The required returns in Table 5 are higher than those in Table 3, to varying degrees. We should expect this, because these required returns (in Table 5) consider both the primary worker s benefit and the spousal benefit versus just the primary worker s benefit. For example, the break-even (50% probability) required compounded nominal return for the 62 vs 66 scenario not considering spousal benefits (Table 3) was 7.0%. If we assume both spouses are the same age and the primary worker is a male, the break-even return increases to 9.3% (Table 5). While an increase in the required return from 7.0% (single worker) to 9.3% (married worker) is notable, the increases for the other two scenarios (62 vs 70 and 66 vs 70) were more significant. For example, the breakeven (50% probability) nominal required compounded return for the 66 vs 70 scenario not considering spousal benefits (Table 3) was 4.6%, and was the lowest break-even required return across the three scenarios. Given that this return was below our target return for the analysis (5%), it would likely be worthwhile for a single investor to not delay benefits past age 66. However, when spousal benefits are considered and the primary worker is male, the required compounded nominal return increases from 4.6% to 8.4%, which is 83% higher Page 19 of 30

20 than the single break-even return (4.6%). In fact, when looking at the probabilities, there is a 75% chance the primary worker would be better off delaying benefits if he can earn a return of 4.7% or more (assuming the spouse is the same age). Unlike Table 3, none of the returns at the 75% likelihood level are negative in Table 5. This means the required return is much higher for even the worst one-in-four scenarios. The break-even returns increase even more when mortality is based on the Society of Actuaries Annuity 2000 Mortality Table. These results are included in Appendix 2, and are an abbreviated version of the results in Table 5, whereby only the break-even returns (50% probability) are included. The previous analysis contained information about the compounded annual nominal return required to break even when the surviving spouse receives the entire incremental benefit from the decision about whether the primary worker delays benefits. What if, though, the spouse expects to receive less than the full incremental increase because his or her own benefit is relatively high? In this case, there would still be some advantage to delaying benefits, but less depending on the spouse s own benefit. For example, assume a male worker can receive $750 per month at age 62 or $1,000 at age 66 while the spouse is eligible for $900 per month. If the primary worker takes benefits early, then the spouse will stick with his or her own $900 benefit. However, if the primary worker delays benefits, then the incremental increase to the spouse is only $100 more ($1,000-$900 =$100), not the full $250. So, the spouse captures 40% of the primary worker s incremental benefit increase as calculated using equation 1:,0 Where: I* = Incremental increase in spousal survivor benefit D B = Primary worker s delayed SS benefit S B = Spouse s SS benefit E B = Primary worker s early SS benefit Note the incremental increase in spousal survivor benefit (I*) cannot be negative. If the spousal SS benefit (S B ) is greater than the primary worker s delayed benefit (D B ), then the incremental increase is zero (since it is bounded by the maximum function). Also, if the spousal benefit (S B ) is less than the primary worker s early SS benefit (E B ), the primary worker s early SS benefit should be considered for the incremental increase value (I*) since the primary worker s early SS benefit (E B ) is what the spouse would receive upon the death of the primary worker. The objective of this analysis is to determine how the required return changes based on varying levels of I*. The break-even required compounded nominal returns for various I* values are included in Table 6. These are the values that correspond to a 50% probability of the primary worker being better off either delaying benefits or taking them early, i.e., the return that should make the primary worker indifferent between delaying benefits or taking them early. Page 20 of 30

21 Table 6: Break-even Return Required to Be Better off Delaying Benefits for Various Simulation Probabilities Assuming Spouse Receives Only Some Percentage of the Incremental Delayed Benefit Increase Incremental Increase in Spousal Survivor Benefit (I*) 0% 20% 40% 60% 80% 100% 0% 20% 40% 60% 80% 100% Spouse Age Difference Versus Early Claiming Age Male is Primary Worker: 62 vs 66 Female is Primary Worker: 62 vs Younger Spouse Older Spouse Spouse Age Difference Versus Early Claiming Age Male is Primary Worker: 62 vs 70 Female is Primary Worker: 62 vs Younger Spouse Older Spouse Spouse Age Difference Versus Early Claiming Age Male is Primary Worker: 66 vs 70 Female is Primary Worker: 66 vs Younger Spouse Older Spouse Mortality based on the Social Security Adminstration 2007 Periodic Life Table The reader should note that the 0% I* values in Table 6 are effectively the break-even values for single workers, since there is effectively no spousal survivor benefit. These are break-even returns noted in Table 3. The 100% I* values in Table 6 correspond to the break-even values where the spouse receives the entire incremental benefit from the primary worker delaying retirement, which is the same as the results in Table 5. The results for the Society of Actuaries Annuity 2000 Mortality Table are included in Appendix III. The simplest approach to determining the appropriate break-even return for various I* values would be to weight the values in Table 3 and Table 5 respectively. While this would provide some indication of the additional value in the relationship, the change is not completely linear. Half of the increase in the break-even return is achieved with I* values that are roughly 35%. In other words, even a small potential increase in survivor benefits has a meaningful impact on the required return. The Gamma Impact Blanchett and Kaplan (2012), BK herein, introduce a metric they call gamma to measure the additional expected retirement income achieved by an individual investor from making intelligent financial planning Page 21 of 30

22 decisions. In their initial research on the topic, BK focus on five fundamental financial planning techniques for retirees: a total wealth framework to determine the optimal asset allocation, a dynamic withdrawal strategy, incorporating guaranteed income products (i.e., annuities), tax-efficient decisions, and liability-relative asset allocation optimization. They find following a gamma-optimized approach can yield 29% more income for a retiree on a utility-adjusted basis, which is equivalent to achieving an annual return increase of 1.82%. This gamma-equivalent alpha of 1.82% represents the value a client could realize from working with a competent financial planner (or managed accounts provider). In order to determine the potential gamma available to retirees through optimal Social Security claiming an additional analysis is conducted using many of the base assumptions in BK, where an optimal approach (delaying SS benefits to FRA, or age 66) is compared to a naïve strategy, which is claiming benefits at age 62 (which is the most popular SS claiming age). For the gamma analysis, we assume the primary worker is a male and the spouse is a female, both of which are the same age. For the naïve (i.e., non-optimal ) strategy the primary worker will receive a benefit that is the 25% less than the FRA benefit from claiming at age 62 and the spousal benefit (50% of the primary worker s benefit) will be reduced by 30% from claiming at age The primary worker s SS benefit amount is assumed to be $20,000 at FRA, and we apply the reductions to the actual benefit versus the PIA for simplicity purposes. Similar to BK, we use the Annuity 2000 mortality table for the analysis. In order to determine the benefit of delayed SS retirement benefits, we calculate the median weighted net present value of expected benefits for the primary worker (including the spousal survivor benefit) and spouse for the two scenarios at age 62. Age 62 is used as a base age, because there is a cost associated with delaying SS benefits if the retiree dies during the period he or she could have already began commencing benefits. The analysis is conducted in real terms (i.e., today s dollars) where the nominal return is 5% and inflation is 3%, thereby the real discount rate is 2%. These are the same general return assumptions for the original analysis and the return for the base portfolio in the original BK study. Table 7 contains the differences in the benefit values. Table 7: Net Present Value Benefit Differences Primary ($) Spouse ($) Total ($) Claim Early 310, , ,649 Claim at FRA 337, , ,828 Difference 27,380 16,799 44,179 Delaying SS benefits to FRA would generate a median mortality weighted net present value in benefits that is $44,179 more than claiming early. In the original BK analysis the median mortality weighted net present value of SS benefits was equal to the value of the portfolio, therefore, following this line of reasoning the increase in the total SS benefits ($44,179), would represent an equivalent increase in assets to fund retirement (i.e., retirement income) of 9.15%. This 9.15% increase represents an example of the gamma achievable to the client making an optimal, or at least more optimal, decision regarding when to claim SS benefits. In order to determine how much additional annual return, or alpha, that would need to be generated to create 9.15% more income, a Monte Carlo simulation is conducted. For the simulation, we compare the amount of income generated from a portfolio invested in 20% equities with a 4% initial withdrawal rate, where subse- 9 Page 22 of 30

23 quent withdrawals are based on the original withdrawal amount but increased for inflation. We then change the average portfolio return by -2%, -1%, 0% (no change), +1%, and +2% to determine the impact the return differences have on the amount of income generated. These results are included in Figure 4. Figure 5: Relationship Between Change in Retirement Income and Changes in Returns for a 4% Initial Withdrawal Change in Return (%) Median Change in Retirement Income (%) By fitting a third-order polynomial to the curve in Figure 5, we estimate the equivalent annual return impact of a +9.15% increase in retirement income to be +.74%. This.74% is gamma-equivalent alpha which represents the equivalent effective alpha an advisor would have to generate to create 9.14% more income for a client during retirement. An advisor able to generate +.74% of alpha each year during retirement which could easily exceed 20 years would widely be viewed to have added significant value for his client. In this case, the.74% of alpha can be achieved by simply helping a client make a more optimal SS claiming decision. Unlike traditional alpha, this gamma-equivalent alpha is not a zero-sum game. Additional Considerations With the decline of defined benefit plans and pension plans, Americans are increasingly forced to manage their own retirement assets and create their own distribution strategies. While there are certainly benefits that can be achieved with this type of flexibility, research is starting to note significant potential costs with this self-funding approach. For example, Korniotis and Kumar (2011) note that while older investors tend to have more experience, the adverse effects of cognitive aging dominate the positive effects of experience. They conclude that older investors are likely to experience 3-5% lower annual returns on a risk-adjusted basis. This topic has also been explored by Finke, Howe, and Huston (2011), who note that financial decision-making is a form of crystallized intelligence that requires both memory and problem-solving skills. Unfortunately, they find financial literacy scores decline by about 2% each year after age 60. This research suggests older retirees are likely better served having a retirement income strategy that is professionally managed and does not require ongoing decision making. Delaying benefits and effectively purchasing a SS annuity is one way to remove the withdrawal/portfolio burden from a retiree since the income would be transferred from a portfolio (which would be used to fund the early years in retirement) to a form of guaranteed income (SS). Page 23 of 30

24 The analysis conducted for this paper was rather robust, yet it still is only able to provide general guidance for an individual seeking to determine when to claim SS benefits. Most people would likely be best served engaging an independent financial planner to assist with this decision. Additionally, there are online calculators that can help someone make the best decision possible. One example is com, which takes into account SS s earnings test, reductions for early retirement, re-computation of benefits, delayed retirement credit, family benefit maximum, windfall elimination provision, government pension offset provision, and option to file and suspend. Conclusions The gains that can be achieved from selecting the optimal age to commence Social Security benefits can be significant. This decision becomes even more important when considering the fact SS is the largest retirement income source for Americans. The factors affecting the SS claiming decision were explored in this paper, such as life expectancy, taxes, and the benefits available to a surviving spouse, primarily using a concept known as the break-even return. The break-even return is the annual nominal compounded rate of return a retiree would need to earn over his or her retirement, after fees, to be indifferent between claiming benefits early versus delaying benefits. We find the effective return achieved by a retiree from making the optimal SS claiming decision is likely to exceed the return the retire can make by taking benefits early and investing them, especially in today s low interest rate environment. We find an investor would likely need to earn an annual nominal compounded rate of return, net of fees, of more than 7% to be better off claiming benefits early. We also conduct an analysis to determine the financial planning alpha, or gamma, generated by optimal SS claiming. We find that delaying benefits can generate 9.15% more income for a hypothetical married couple, which is equivalent to increasing returns (i.e., alpha) by +.74% per year in retirement. In summary, we find that most retirees are going to be better served delaying SS benefits until full retirement age or later. Delayed SS benefits are especially valuable for females, married couples, and investors who expect to invest in relatively conservative portfolios during retirement. Delayed SS benefits can enable an investor to effectively achieve a rate of return that is much greater than he or she is likely to earn in the market. SS benefits can also provide a valuable hedge against longevity risk and offer a form of protection from the adverse effects of cognitive decline at older ages. Page 24 of 30

25 References Blanchett, David and Paul Kaplan Alpha, Beta, and Now Gamma. Working Paper. Butrica, Barbara A., Howard M. Iams, and Karen E. Smith The Changing Impact of Social Security on Retirement Income in the United States. Social Security Bulletin, vol. 65, no. 3. Cunningham, Donald F. and Paul R. Erickson The Combined Income Tax Effect on Early versus Normal Social Security Benefits for Single Individuals. Journal of Financial Service Professionals, March: Docking, Diane, Richard Fortin, and Stuart Michelson The Influence of Gender and Race on the Social Security Early Retirement Decision for Single Individuals. org/11conference/11proceedings/(b4)%20docking,%20fortin,%20michelson.pdf Entmacher, Joan Strengthening Social Security Benefits for Widow(er)s: The 75 Percent Combined Worker Benefit Alternative. January. Strengthening_Social_Security_for_Vulnerable_Groups.pdf Finke, Michael, John Howe, and Sandra Huston Old Age and the Decline in Financial Literacy. Available at SSRN: Jennings, William W., and William Reichenstein Estimating the Value of Social Security Retirement Benefits. Journal of Wealth Management, vol. 4 (Winter): Korniotis, G.M. and A. Kumar, Do Older Investors Make Better Investment Decisions? The Review of Economics and Statistics, vol. 93, no. 1: Kreuger, C Mortality Assumptions: Are Planners Getting it Right? Journal of Financial Planning, vol. 24, LIMRA Guaranteed Income Annuities Report. Mahaney, James and Peter Carlson Rethinking Social Security Claiming in a 401(k) World. Pension Research Council Working Paper. PRC WP McCormack, Joseph P. and Grady Perdue Optimizing the Initiation of Social Security Benefits. Financial Services Review, vol. 15: Page 25 of 30

26 Meyer, William, and William Reichenstein Social Security: When to Start Benefits and How to Minimize Longevity Risk. Journal of Financial Planning, vol. 23 (March): Meyer, William and William Reichenstein How the Social Security Claiming Decision Affects Portfolio Longevity. Journal of Financial Planning, vol. 25, no. 4 (April): Munnell, Alicia and Mauricio Soto When Should Women Claim Social Security Benefits? Journal of Financial Planning, vol. 20, no. 6: Munnell, Alicia, Alex Golub-Sass, and Nadia Karamcheva Strange but True: Claim Social Security Now, Claim More Later. April. Number 9-9. Rhee, Nari Meeting California s Retirement Security Challenge. research/caretirement_challenge_1011.pdf Reno, Virginia and Joni Lavery When to Take Social Security Benefits: Questions to Consider. Social Security Brief, January, No 31. Rose, Clarence C. and L. Keith Larimore Social Security Benefit Considerations in Early Retirement. Journal of Financial Planning, vol. 14: Ryan, Charles Social Security Reset: When Does It Make Sense? Journal of Financial Planning, (June): Sass, Steven A Should You Buy an Annuity from Social Security. Number 12-10, May. Sun, Wei and Anthony Webb How Much Do Households Really Lose by Claiming Social Security at Age 62? Center for Retirement Research at Boston College, CRR WP , March, Page 26 of 30

27 Important Disclosures 2013 Morningstar. All rights reserved. This document includes proprietary material of Morningstar. Reproduction, transcription or other use, by any means, in whole or in part, without the prior written consent of Morningstar is prohibited. The Morningstar Investment Management division is a division of Morningstar and includes Morningstar Associates, Ibbotson Associates, and Morningstar Investment Services, which are registered investment advisors and wholly owned subsidiaries of Morningstar, Inc. The Morningstar name and logo are registered marks of Morningstar, Inc. The above commentary is for informational purposes only and should not be viewed as an offer to buy or sell a particular security. The data and/or information noted are from what we believe to be reliable sources, however Morningstar has no control over the means or methods used to collect the data/information and therefore cannot guarantee their accuracy or completeness. The opinions and estimates noted herein are accurate as of a certain date and are subject to change. The indexes referenced are unmanaged and cannot be invested in directly. Past performance is no guarantee of future results. This commentary may contain forward-looking statements, which reflect our current expectations or forecasts of future events. Forward-looking statements are inherently subject to, among other things, risks, uncertainties and assumptions that could cause actual events, results, performance or prospects to differ materiality from those expressed in, or implied by, these forward-looking statements. The forward-looking information contained in this commentary is as of the date of this report and subject to change. There should not be an expectation that such information will in all circumstances be updated, supplemented or revised whether as a result of new information, changing circumstances, future events or otherwise. The results from the simulations described, while hypothetical in nature and not actual investment results or guarantees of future results, can provide an important look at strategies designed to help retirees reach their goals. Monte Carlo is an analytical method used to simulate random returns of uncertain variables to obtain a range of possible outcomes. Such probabilistic simulation does not analyze specific security holdings, but instead analyzes the identified asset classes. The simulation generated is not a guarantee or projection of future results, but rather, a tool to identify a range of potential outcomes that could potentially be realized. The Monte Carlo simulation is hypothetical in nature and for illustrative purposes only. Results noted may vary with each use and over time. This should not be considered tax or financial planning advice. Please consult a tax and/or financial professional for advice specific to your individual circumstances. Page 27 of 30

28 Appendix Appendix I: Survival Probabilities Current Age (%) Joint Age (%) Male: Annuity 2000 Table Female: Annuity 2000 Table 98 Joint: Annuity 2000 Table Death Age Death Age (Neither Alive) Male: SSA Periodic Life Table Female: SSA Periodic Life Table 96 Joint: SSA Periodic Life Table Death Age Death Age (Neither Alive) Page 28 of 30

29 Appendix II: Break-Even Return Required to Be Better off Delaying Benefits Assuming Spouse Receives Only Spousal Benefit Break-Even Return (%) Break-Even Return (%) 62 vs vs vs70 62 vs vs vs 70 Spouse Age Difference Versus Early Claiming Age Male is Primary Worker: 62 vs 66 Female is Primary Worker: 62 vs Younger Spouse Older Spouse Mortality based on the Social Security Adminstration 2007 Annuity Table Appendix III: Break-Even Return Required to Be Better off Delaying Benefits for Various Simulation Probabilities Assuming Spouse Receives Only Some Percentage of the Incremental Delayed Benefit Increase % of Delayed Incremental Benefit Increase % of Delayed Incremental Benefit Increase Recieved by the Spouse Recieved by the Spouse 0% 20% 40% 60% 80% 100% 0% 20% 40% 60% 80% 100% Spouse Age Difference Versus Early Claiming Age Male is Primary Worker: 62 vs 66 Female is Primary Worker: 62 vs Younger Spouse Older Spouse Spouse Age Difference Versus Early Claiming Age Male is Primary Worker: 62 vs 70 Female is Primary Worker: 62 vs Younger Spouse Older Spouse Spouse Age Difference Versus Early Claiming Age Male is Primary Worker: 66 vs 70 Female is Primary Worker: 66 vs Younger Spouse Older Spouse Mortality based on the Social Security Adminstration 2007 Periodic Life Table Page 29 of 30

30 About the Author David Blanchett David Blanchett, CFA is head of retirement research for the Morningstar Investment Management division, which provides investment consulting, retirement advice, and investment management operations around the world. In this role, he works closely with the division s business leaders to provide research support for the group s consulting activities and conducts client-specific research primarily in the areas of financial planning, tax planning, and annuities. He is responsible for developing new methodologies related to strategic and dynamic asset allocation, simulations based on wealth forecasting, and other investment and financial planning areas for the investment consulting group, and he also serves as chairman of the advice methodologies investment subcommittee. Prior to joining Morningstar in 2011, Blanchett was director of consulting and investment research for the retirement plan consulting group at Unified Trust Company in Lexington, Ky. Blanchett has authored more than 35 articles that have been published in InvestmentNews, Journal of Financial Planning, Journal of Index Investing, Journal of Indexes, Journal of Investing, Journal of Performance Measurement, Journal of Pension Benefits, and Retirement Management Journal. He won Journal of Financial Planning s 2007 Financial Frontiers Award for his research paper, Dynamic Allocation Strategies for Distribution Portfolios: Determining the Optimal Distribution Glide Path. Blanchett holds a bachelor s degree in finance and economics from the University of Kentucky, where he graduated magna cum laude, and a master s degree in financial services from American College. He also holds a master s degree in business administration, with a concentration in analytic finance, from the University of Chicago Booth School of Business, and he is currently pursuing a doctorate degree in personal financial planning from Texas Tech University. Blanchett holds the Chartered Financial Analyst (CFA), Certified Financial Planner (CFP), Chartered Life Underwriter (CLU), Chartered Financial Consultant (ChFC), Accredited Investment Fiduciary Analyst (AIFA), and Qualified Pension Administrator (QKA) designations. Page 30 of 30