Optimizing Your Retirement Income: What Works Best and Why

    by AAII Staff

    With the oldest baby boomers hitting 62 this year, and more than 70 million of them likely to enter retirement over the next 20 years, the hard truth is that only a small minority are accumulating enough savings to provide for their income needs during decades in retirement.

    This uncomfortable reality is particularly true given the overall rise in life expectancy, sharply rising medical costs, the trend toward more active and costly retirement lifestyles, and, not least, the relentless toll of inflation.

    For the financially fortunate with sufficient personal savings, Social Security benefits, and corporate pensions to meet all their retirement income needs, the main financial challenges of retirement are how to invest and spend wisely and perhaps provide for their heirs as well.

    However, more than 75% of all workers age 55 or older report having less than $250,000 in investments apart from their homes and pensions, according to a recent survey by the Employee Benefit Research Institute (EBRI). At a recommended initial withdrawal amount of 4%, that provides an income from their investments of just $10,000 in the first year of retirement. Nevertheless, those approaching retirement can improve their income and financial security in retirement depending on their flexibility and their approach to four big decisions that are usually under their control:


    • When they stop working.
    • When they start taking Social Security.
    • How they manage withdrawals from their savings.
    • How they allocate their assets.

    The first two can have a significant impact on the amount of income in retirement, while the second two affect the sustainability of that income over a 30-year retirement. Taken together, controlling these decisions will go a long way toward determining retirees’ overall security in retirement.

    Working Longer

    Generally, no single decision will improve pre-retirees’ potential retirement security as much as continuing to work even a few more years beyond the anticipated retirement date.

    Appealing or not, this is usually the best option for those who come up short on retirement savings.

    Unless pre-retirees enjoy a windfall or a sharp rise in their incomes late in their careers, those just a few years from retiring who have not saved enough will probably not be able to make up their shortfalls solely with increased savings levels or by investing more aggressively. They simply will not have enough time for their assets to compound.

    Studies we have conducted at T. Rowe Price show:


    • The long-term impact of a greater rate of savings at this late stage—even boosting saving from 15% of salary to 25%, for example—is certainly positive, but relatively marginal in terms of increasing annual retirement income from investments in just a few years.
    • Likewise, those who invest more aggressively as they approach retirement—moving, say, from 60% of their portfolio assets in equities to 70% or more—also are not likely to make up for lost time. And because of potentially greater investment volatility, this step could actually cause their portfolio balances to drop significantly just before or after their desired retirement date.
    • On the other hand, continuing to work full-time could increase pre-retirees’ expected annual retirement income from their investments, in today’s dollars, by about 7% for each additional year of work and contributions. Working an additional three years—say, from ages 62 to 65—and continuing to save 15% of salary could raise annual income from investments by 22%, or by as much as 39% by working an additional five years. (See Figure 1.)
    • Combining the two options—working an extra five years and boosting savings to 25% of annual earnings—would increase annual retirement income from savings by 50%.

    The logic behind this is simple: Those who continue working can contribute to their savings for a few more years, delay tapping into their nest eggs, and reduce the number of years that their assets will have to generate income in retirement—a powerful combination.

    Moreover (as we’ll discuss next), this strategy may enable them to delay when they start taking Social Security benefits, which can significantly increase those payments.

    Also, those who continue working may receive health and life insurance and prescription drug benefits from their employers—all expenses that more and more retirees have to cover themselves. (Retirees are not eligible for Medicare until age 65.)

    While many retirees may not want or be able to continue working in their same jobs full-time, they could still improve their potential income in retirement by working part-time in the same or another job.

    Though wages are likely to be reduced with part-time work, the same potential financial dynamics apply: Every dollar earned is one that doesn’t have to be withdrawn from retirement savings. Indeed, $20,000 in annual income from a part-time job is the equivalent of withdrawing 4% a year from an additional $500,000 in savings.

    Social Security

    For most people 65 and older, Social Security is the largest single source of their income, accounting for 40% on average, according to EBRI.

    Delaying taking Social Security benefits can significantly increase a retiree’s income. For example, those benefits (in today’s dollars) increase approximately 8% per year based on Social Security Administration formulas.

    Thus, delaying three years (from 62 to 65) results in a 27% increase in the purchasing power of a retiree’s Social Security benefits, and delaying until age 70 almost doubles the purchasing power of these benefits (about 88%). (See Table 1.) The potential gain in actual benefits could be even higher, since Social Security benefits are adjusted annually for inflation.

    Table 1. The Impact of Delaying Social Security Benefits
    Soc Sec
    % Increase
    Over Benefit
    at Age 62
    Calculations and Assumptions: Social Security payments calculated using the Quick Calculator on the ssa.gov Web site. This assumes an individual who is currently age 62 (with a full retirement age of 66), who is continuing to work and earning $100,000 each year until benefits begin. All figures reflect current dollars. Actual benefits would be higher to reflect future adjustments for inflation.

    Each year this individual continues working, his or her annual retirement income from Social Security would increase by about 8%, regardless of how much of his or her additional wages he or she saves annually.

    Sources: T. Rowe Price Associates and Social Security Administration.

    62 17,772
    63 19,044 7
    64 20,832 17
    65 22,644 27
    66 24,468 38
    67 26,664 50
    68 28,884 63
    69 31,128 75
    70 33,408 88


    Taking all three steps to increase potential retirement income—continuing to work, saving, and delaying Social Security—could increase the purchasing power of total retirement income from retirees’ combined investments and Social Security benefits by about 8% for each year after 62, or 28% in three years (as reflected in Figure 2, which also provides the underlying assumptions).


    And doing that from ages 62 to 70 would almost double total retirement income from investments and Social Security in today’s dollars.

    To boil this down, here is another way of looking at the overall benefit of working longer and delaying Social Security benefits. If a 62-year-old wants a 30% increase in the purchasing power of his total retirement income from investments and Social Security, then he could:


    • Retire in three years at 65 by saving 25% of his salary annually.
    • Retire in three-and-a-half years at 65.5 by saving 15% of his salary annually.
    • Retire in four years at 66 by spending rather than saving his additional earnings.

    (These illustrations assume that the retiree does not begin taking Social Security until he or she stops working.)

    Keep in mind that, for those who continue working and begin Social Security benefits prior to attaining full retirement age (66 for most boomers), some benefits could be temporarily withheld depending on the amount of wages earned.

    In general, analyzing whether pre-retirees should decide to take benefits early, at age 62, for example, or whether they would be better off in the long run by waiting for increased benefits until as late as age 70 really depends on whether they can afford to delay receiving benefits, whether they are married, and, to some extent, on how long they expect to live. It’s often easy to underestimate longevity, particularly because married couples may neglect to take into account their joint life expectancy when it comes to Social Security. T. Rowe Price financial planners, relying on various actuarial studies, urge married investors to plan for at least one spouse living until 95.

    Many financial planners used to recommend taking and spending your Social Security benefits as soon as you become eligible. But today, with greater longevity, delaying Social Society for as long as possible may be the best strategy if you can afford it.

    Whatever you do, take extreme care when making Social Security benefit decisions, not least because the annual amounts of payments (adjusted for inflation) for which you will be eligible generally will be locked in for the remainder of your life and for the remainder of your surviving spouse’s life, except in certain circumstances.

    Taking Withdrawals

    The third and fourth major decisions faced by pre-retirees—their withdrawal amounts and their asset allocation in retirement—boils down to figuring out how to maximize the amounts they can withdraw initially from their retirement savings without running out of money during their lifetimes.

    While working longer, saving more and delaying Social Security benefits can increase total retirement income, deciding on an appropriate initial withdrawal amount from portfolio assets and adjusting that amount as necessary can go a long way toward lowering the risk that retirees outlive their resources.

    In most cases, your ability to avoid running out of money is driven more by your initial and subsequent withdrawal amounts than by your asset-allocation strategy, which for many investors is counter-intuitive.

    T.Rowe Price simulation studies show that:

    • For a 30-year retirement, an initial withdrawal amount of 4% from a balanced portfolio of assets (with 3% annual increases in the withdrawal amount for inflation) would provide as high as an 89% chance of having assets remaining at the end of this period. A 5% initial withdrawal amount with inflation adjustments, on the other hand, reduces these odds to 40% to 65%, depending on the asset allocation strategy (Table 2).
    • If retirees suffer poor portfolio returns in the first few years of retirement, they should consider lowering their withdrawal amounts temporarily, or at least holding their annual withdrawals flat for a while instead of increasing them for inflation. Extensive analysis by T. Rowe Price has demonstrated that this approach is much more advantageous than, for example, attempting to counteract a market downturn by dramatically reducing the level of equities—and hence the long-term growth potential—in retirees’ portfolios.

    Table 2. How Much Can You Withdraw in Retirement?
    The estimated probability of maintaining several initial withdrawal amounts throughout a 30-year retirement without running out of money, depending on the investor’s asset allocation.
    30-Year Retirement Period
    Stock/Bond Mix
    80%/20% 60%/40% 40%/60% 20%/80%
    Simulation Success Rate (%)*
    7 28 19 7 1
    6 45 38 24 7
    5 65 63 57 40
    4 84 87 89 89
    Assumptions: Analysis assumes pretax withdrawals from tax-deferred assets and can be applied to any size retirement portfolio. In this study, the annual withdrawal amounts are increased by 3% for inflation.

    *The probability of having at least $1 in the portfolio at the end of 30 years. Portfolio performance is based on a Monte Carlo simulated probability analysis.

    Sources: T. Rowe Price Associates.


    Asset Allocation

    In general, making minor adjustments to a balanced portfolio in retirement has less impact on financial security than the other three decisions.

    However, pre-retirees often make the serious mistake of assuming that the safest path in retirement is minimizing equity exposure to lower their market risk. Instead, moderate exposure to equities is recommended for diversification, growth potential, sustaining real income, and providing a “cushion” to cover unexpected expenses during a 30-year retirement.

    Also, to increase the potential wealth that retirees could draw on in emergencies—or to possibly leave more money to heirs—retirees could opt for somewhat higher allocations to equities, though that does carry greater risk in market downturns.

    Retirees should maintain at least 40% in equities, even into their 80s, and keep no more than 30% of their assets in cash or short-term bonds.

    The basic dilemma is that if you have too much set aside for emergencies in cash, which usually has a very modest annual return, you run the risk of not keeping up with inflation and possibly running out of resources from which to take withdrawals. However, if you have too much invested in stocks, you lessen your ability to cope with market uncertainties and run the risk of having to sell equities during a market setback to provide for income or unexpected contingencies.

    The bottom line is to maintain a balanced, diversified portfolio—with moderate growth potential and a moderate risk profile.

    Flexible and Well Thought Out

    With all of these critical decisions—when to stop working, when to start taking Social Security, how much to withdraw from your portfolio in retirement, and determining the right asset allocation strategy—the overarching concepts are to maintain flexibility in your plans for retirement, and make thoughtful decisions regarding financial matters that are under your control.

    Such pre-retirement planning can help optimize your financial prospects for years to come.

    Explaining Monte Carlo Analysis Used in the T. Rowe Price Retirement Study

    The T. Rowe Price studies in this article were based on analyses using Monte Carlo simulations. Here is an explanation of the process, along with the assumptions that were used in the analyses.

    Monte Carlo simulations model future uncertainty. In contrast to tools generating average outcomes, Monte Carlo analyses produce outcome ranges based on probability-thus incorporating future uncertainty. In this example, savings data are based on average outcomes and retirement income data on Monte Carlo analysis.

    Material Assumptions Include:


    • Underlying long-term expected annual returns for the asset classes are not based on historical returns, but estimates, which include reinvested dividends and capital gains.
    • Expected returns-plus assumptions about asset class volatility and correlations with other classes-are used to generate random monthly returns for each class over specified time periods.
    • These monthly returns are then used to generate hundreds of scenarios, representing a spectrum of possible performance for the modeled asset classes. Success rates are based on these scenarios.
    • Taxes aren't taken into account, nor are early withdrawal penalties. But fees-average expense ratios for typical actively managed funds within each asset class-are subtracted from the expected annual returns.

    Material Limitations Include:

    • Extreme market movements may occur more often than in the model.
    • Some asset classes have relatively short histories. Expected results for all asset classes may differ from our assumptions-with those for classes with limited histories potentially diverging more.
    • Market crises can cause asset classes to perform similarly, lowering the accuracy of projected portfolio volatility and returns. Correlation assumptions are less reliable for short periods.
    • The model assumes no month-to month correlations among asset class returns. It does not reflect the average periods of "bull" and "bear" markets, which can be longer than those modeled.
    • Inflation is assumed constant, so variations are not reflected in our calculations.
    • The analysis does not use all asset classes. Other asset classes may be similar or superior to those used.

    Model Portfolio Construction

    Seven model portfolios were designed for effective diversification among asset classes. Diversification theoretically involves all asset classes: equities, bonds, real estate, foreign investments, commodities, precious metals, currencies, and others. Because investors are unlikely to own all these assets, T. Rowe Price selected those most appropriate for long-term investors: stocks, bonds, and short-term bonds. T. Rowe Price then chose seven sub-asset classes for the model portfolios: large-cap, small-cap, and international stocks; and short-term, investment-grade, high-yield, and international bonds. The analyses did not consider real estate because of its illiquidity and investors' potential exposure from home ownership. T. Rowe Price believes the selected fixed-income sub-asset classes fairly represent the domestic capital markets. Short-term, investment-grade bonds were chosen for stability, eliminating a cash allocation because investors are best able to decide that according to their near-term needs. The portfolios were built using the complementary behavior of sub-asset classes over long periods of time, enabling more efficient investment mixes through low correlations.

    The initial withdrawal amount is the percentage of the initial value of the investments withdrawn on the first day of the first year. In subsequent years, the amount withdrawn grows by a 3% annual rate of inflation. Success rates are based on simulating 500 market scenarios and various asset allocation strategies. The underlying long-term expected annual return assumptions (without fees) are 10% for stocks; 6.5% for intermediate-term, investment-grade bonds; and 4.75% for short-term bonds. Net-of-fee expected returns use these expense ratios: 1.211% for stocks; 0.726% for intermediate-term, investment-grade bonds; and 0.648% for short-term bonds.