The US Retirement System

American households rely on a combination of resources in retirement, and the role each type of resource plays has changed over time and varies across households. The traditional analogy compares retirement resources to a three-legged stool, with resources divided equally among the legs—Social Security, employer-sponsored pension plans, and private savings. But Americans’ retirement resources are best thought of as a five-layer pyramid.

Retirement Resource Pyramid

The retirement resource pyramid has five layers, which draw from government programs, compensation deferred until retirement, and other savings (Figure 7.1):

  • Social Security
  • homeownership
  • employer-sponsored retirement plans (private-sector and government employer plans, including both defined benefit [DB] and defined contribution [DC] plans)
  • individual retirement accounts (IRAs), including rollovers
  • other assets

Figure 7.1

Retirement Resource Pyramid

 

Figure 7.1

 

Download an Excel file of this data.

Source: Investment Company Institute, The Success of the US Retirement System

Though the importance of each layer differs by household, together they have enabled recent generations of retirees, on average, to maintain their standard of living in retirement.

The construction of each household’s pyramid varies with age and income. Younger households are more likely to save primarily for reasons other than retirement, such as a home purchase, family needs, or education (Figure 7.2). By contrast, older households are more likely to save primarily for retirement, as many already have reached their other savings goals. The tendency of younger workers to focus less on saving for retirement is consistent with economic models of life-cycle consumption predicting that most workers delay saving for retirement until later in their careers. Lower-income households also focus less on saving for retirement, reflecting the fact that Social Security benefits replace a higher share of pre-retirement earnings for workers with lower lifetime earnings.

Figure 7.2

Primary Reason for Household Saving Changes with Age

Percentage of households by age of household head, 2013

Figure 7.2

Download an Excel file of this data.

Source: Investment Company Institute tabulations of the 2013 Federal Reserve Board Survey of Consumer Finances. See ICI Research Perspective, “Supplemental Tables: Who Gets Retirement Plans and Why, 2013.”

Social Security, the base of the US retirement resource pyramid, is the largest component of retiree income and the primary source of income for lower-income retirees. Social Security benefits are funded through a payroll tax equal to 12.4 percent of earnings of covered workers (6.2 percent paid by employees and 6.2 percent paid by employers) up to a maximum taxable earnings amount ($118,500 in 2016). The benefit formula is highly progressive, with benefits representing a much higher percentage of earnings for workers with lower lifetime earnings. By design, Social Security is the primary means of support for retirees with low lifetime earnings and a substantial source of income for all retired workers. For individuals born in the 1960s who claim benefits at age 65, the Congressional Budget Office (CBO) projects that mean first-year Social Security benefits will replace 83 percent of average inflation-indexed lifetime earnings for retired workers in the lowest 20 percent of households ranked by lifetime household earnings (Figure 7.3). The mean replacement rate drops to 64 percent for workers in the second quintile of households, and then declines more slowly as lifetime household earnings increase. Even for workers in the top 20 percent of lifetime earners, Social Security benefits are projected to replace a considerable portion (34 percent) of earnings.

For many near-retiree households, homeownership is the second most important retirement resource after Social Security. Older households are more likely to own their homes; more likely to own their homes without mortgage debt; and, if they still have mortgages, more likely to have small mortgages relative to the value of their homes. Retired households typically access this resource simply by living in their homes rent-free.

Figure 7.3

Social Security Benefit Formula Is Highly Progressive

Average projected Social Security replacement rates for workers in the 1960s birth cohort by quintile of lifetime household earnings, percent

Figure 7.3

Download an Excel file of this data.

Note: For each worker, the replacement rate is the ratio of Social Security benefits net of income tax to average inflation-indexed lifetime earnings. Replacement rates are for workers claiming benefits at age 65. For workers born in the 1960s, the Social Security full benefit retirement age is 67. If these workers claimed benefits at age 67, benefits would increase by about 15 percent.
Source: Congressional Budget Office, The 2016 Long-Term Projections for Social Security: Additional Information

Employer-sponsored retirement plans and IRAs, which complement Social Security benefits and are important resources for households regardless of income or wealth, increase in importance for households for whom Social Security replaces a smaller share of earnings. In 2013, about eight out of 10 near-retiree households had accrued benefits in employer-sponsored retirement plans—DB and DC plans sponsored by private-sector and government employers—or IRAs (Figure 7.4).

Although less important on average, retirees also rely on other assets in retirement. These assets can be financial—including bank deposits, stocks, bonds, and mutual funds owned outside employer-sponsored retirement plans and IRAs. They also can be nonfinancial—including business equity, investment real estate, second homes, vehicles, and consumer durables (long-lived goods such as household appliances and furniture). Higher-income households are more likely to have large holdings of assets in this category.

Figure 7.4

Near-Retiree Households Across All Income Groups Have Retirement Assets, DB Plan Benefits, or Both

Percentage of near-retiree households1 by income quintile,2 2013

Figure 7.4

Download an Excel file of this data.

1 Near-retiree households are those with a head of household aged 55 to 64, and a working head of household or working spouse.
2 Income is household income before taxes in 2012.
3 Retirement assets include DC plan assets (401(k), 403(b), 457, thrift, and other DC plans) and IRAs (traditional, Roth, SEP, SAR-SEP, and SIMPLE), whether from private-sector or government employers.
4 Households currently receiving DB plan benefits and households with the promise of future DB plan benefits, whether from private-sector or government employers, are counted in this category.
Note: Components may not add to the total because of rounding.
Source: Investment Company Institute tabulations of the 2013 Federal Reserve Board Survey of Consumer Finances

Snapshot of US Retirement Market Assets

Employer-sponsored retirement plans (DB and DC plans sponsored by private-sector and government employers), IRAs (including rollovers), and annuities play an important role in the US retirement system, with assets totaling $25.3 trillion at year-end 2016, up 6.1 percent from $23.9 trillion at year-end 2015 (Figure 7.5). The largest components of retirement assets were IRAs and employer-sponsored DC plans, holding $7.9 trillion and $7.0 trillion, respectively, at year-end 2016. Other employer-sponsored plans include private-sector DB pension funds ($2.9 trillion), state and local government DB retirement plans ($3.9 trillion), and federal government DB plans ($1.6 trillion). In addition, annuity reserves outside of retirement plans were $2.0 trillion at year-end 2016.

Figure 7.5

Total US Retirement Assets

Trillions of dollars; year-end, selected years

Figure 7.5

Download an Excel file of this data.

1 Other retirement assets includes private-sector DB plans; federal, state, and local DB plans; and all fixed and variable annuity reserves at life insurance companies less annuities held by IRAs, 403(b) plans, 457 plans, and private pension funds. Because ICI estimates of annuities held in IRAs, 457 plans, and 403(b) plans are netted from the Federal Reserve Board’s financial accounts’ annuities (life insurance pension fund reserves) figure and reported in their respective categories by ICI, ICI reports a lower annuities total than in the financial accounts of the United States. Federal pension plans include US Treasury security holdings of the civil service retirement and disability fund, the military retirement fund, the judicial retirement funds, the Railroad Retirement Board, and the foreign service retirement and disability fund. These plans also include securities held in the National Railroad Retirement Investment Trust.
2 DC plans include private employer-sponsored DC plans (including 401(k) plans), 403(b) plans, 457 plans, and the Federal Employees Retirement System (FERS) Thrift Savings Plan (TSP).
3 IRAs include traditional IRAs, Roth IRAs, and employer-sponsored IRAs (SEP IRAs, SAR-SEP IRAs, and SIMPLE IRAs).
e Data are estimated.
Note: Components may not add to the total because of rounding.
Sources: Investment Company Institute, Federal Reserve Board, Department of Labor, National Association of Government Defined Contribution Administrators, American Council of Life Insurers, and Internal Revenue Service Statistics of Income Division. See Investment Company Institute, “The US Retirement Market, Fourth Quarter 2016.””

Retirement assets include individual account-based savings (e.g., IRAs and DC plans) and assets held in DB plans. Traditional DB plans promise to pay benefits in retirement typically based on salary and years of service, and assets held in those plans represent funding for those promised benefits. Some DB plans do not have sufficient funding to cover promised benefits that households have a legal right to expect; the total unfunded liabilities of DB plans were $4.2 trillion at year-end 2016 (Figure 7.6). Underfunding is more pronounced in government-sector pension plans. As of year-end 2016, private-sector DB plans had $2.9 trillion in assets and $0.5 trillion in unfunded liabilities. On the other hand, state and local government DB plans had $3.9 trillion in assets and $1.9 trillion in unfunded liabilities, and federal DB plans had $1.6 trillion in assets and $1.8 trillion in unfunded liabilities.

Figure 7.6

Total US Retirement Assets and Unfunded Pension Liabilities

Trillions of dollars, year-end 2016

Figure 7.6

Download an Excel file of this data.

e Data are estimated.
Sources: Investment Company Institute and Federal Reserve Board. See Investment Company Institute, “The US Retirement Market, Fourth Quarter 2016.”

Ownership of retirement accumulations is widespread; 61 percent of US households (or 77 million) reported that they had employer-sponsored retirement plans, IRAs, or both in mid-2016 (Figure 7.7). Fifty-five percent of US households reported that they had employer-sponsored retirement plans—that is, they had assets in DC plan accounts, were receiving or expecting to receive benefits from DB plans, or both. Thirty-four percent reported having assets in IRAs, and 28 percent had both IRAs and employer-sponsored retirement plans. The households in this snapshot represent a wide range of ages and so, they are at different points in the life cycle of saving. Focus on retirement savings tends to increase with age (Figure 7.2), and older households are more likely to have retirement resources; for example, about eight out of 10 near-retiree households have retirement accumulations (Figure 7.4).

Figure 7.7

Many US Households Have Tax-Advantaged Retirement Savings

Percentage of US households, mid-2016

Figure 7.7

Download an Excel file of this data.

1 IRAs include traditional IRAs, Roth IRAs, and employer-sponsored IRAs (SEP IRAs, SAR-SEP IRAs, and SIMPLE IRAs).
2 Employer-sponsored retirement plans include DC and DB retirement plans.
Sources: Investment Company Institute and US Census Bureau. See ICI Research Perspective, “The Role of IRAs in US Households’ Saving for Retirement, 2016.”

Ownership of IRA and DC plan assets has tended to increase with each successive generation of workers, although recent data suggest that ownership rates may have stabilized. For example, in 1983, when they were 44 to 53 years of age, 44 percent of households born in the 1930s owned IRAs or DC plan accounts (Figure 7.8). By comparison, households born a decade later had a 54 percent ownership rate when they were 43 to 52 years old in 1992; and, among households born in the 1950s, 60 percent had IRAs or DC plan accounts when they were 45 to 54 years old in 2004. Earlier in their careers, the 1960s birth cohort appeared to be continuing the trend of increased ownership. In 2013, however, when they were 44 to 53 years old, 57 percent of households born in the 1960s owned IRAs or DC plan accounts—slightly lower than the percentage for the 1950s birth cohort at similar ages and slightly higher than the 1940s birth cohort. Recent experience could indicate that long-term growth in ownership has stabilized, or it could reflect a temporary pause in the long-term trend caused by the weak economy.

Figure 7.8

Rates of IRA or Defined Contribution Plan Ownership

Percentage of US households owning IRAs or DC plans by decade in which household heads were born, 1983–2013

Figure 7.8

Download an Excel file of this data.

Note: Age is the average age of the 10-year birth cohort at the time of the survey. The 10-year birth cohorts are defined using the age of the head of household.
Source: Investment Company Institute tabulations of the Federal Reserve Board Survey of Consumer Finances

Defined Contribution Retirement Plans

DC plans provide employees with a retirement account funded with employer contributions, employee contributions, or both, plus investment earnings or losses on those contributions, less withdrawals. Assets in employer-sponsored DC plans have grown faster than assets in other types of employer-sponsored retirement plans over the past three decades, increasing from 26 percent of employer plan assets in 1986 to 46 percent at year-end 2016. At the end of 2016, employer-sponsored DC plans—which include 401(k) plans, 403(b) plans, 457 plans, the federal Thrift Savings Plan (TSP), Keoghs, and other private-sector DC plans—held an estimated $7.0 trillion in assets (Figure 7.9). With $4.8 trillion in assets at year-end 2016, 401(k) plans held the largest share of employer-sponsored DC plan assets. 403(b) plans, which are similar to 401(k) plans, and which allow employees of educational institutions and certain nonprofit organizations to receive deferred compensation, held another $0.9 trillion in assets. In addition, 457 plans—which allow employees of state and local governments and certain tax-exempt organizations to receive deferred compensation—and the Federal Employees Retirement System (FERS) TSP held a total of $0.7 trillion. Other private-sector DC plans without 401(k) features held the remaining $0.6 trillion.

Figure 7.9

Defined Contribution Plan Assets by Type of Plan

Trillions of dollars; year-end, selected years

Figure 7.9

Download an Excel file of this data.

* Other private-sector DC plans includes Keoghs and other private-sector DC plans (profit-sharing, stock bonus, and money purchase) without 401(k) features.
Note: Components may not add to the total because of rounding.
Sources: Investment Company Institute, Federal Reserve Board, Department of Labor, National Association of Government Defined Contribution Administrators, and American Council of Life Insurers. See Investment Company Institute, “The US Retirement Market, Fourth Quarter 2016.”

401(k) and 403(b) Plan Design and Investment Lineup

Plan Design

Employers who sponsor a 401(k) plan have the option to include features in the plan, such as employer contributions, access to plan assets through participant loans, and automatic enrollment of employees into the plan to encourage participation. The most common of these plan features is employer contributions. In 401(k) plans, employers can make contributions without regard to employee contributions or by using a matching structure that gives employees an incentive to contribute to the plan. Analysis of large 401(k) plans found that nearly nine in 10 (87 percent) made employer contributions of some type in plan year 2014 (Figure 7.10). More than three-quarters of large 401(k) plans had loans outstanding* and more than one-quarter included automatic enrollment in 2014. An analysis of large 403(b) plans found that they are similarly likely to have employer contributions but less likely to have loans outstanding or automatic enrollment.

When designing 401(k) plans, employers tend to select a combination of features that their employees are likely to value. In 2014, 47 percent of large 401(k) plans had both employer contributions and participant loans outstanding but no automatic enrollment, making this the most common combination of plan features (Figure 7.10). The next most common plan design combined all three features—employer contributions, automatic enrollment, and outstanding loans—and was offered by 21 percent of plans, followed by 16 percent having employer contributions only. Less than 3 percent of large plans did not offer any of the three features—that is, they did not provide employer contributions, did not have participant loans outstanding, and did not automatically enroll participants.

* Although the availability of a loan feature is not reported on Form 5500, it is possible to determine whether participants have loans by capturing loan use rather than loan offering. See The BrightScope/ICI Defined Contribution Plan Profile: A Close Look at 401(k) Plans, 2014.


Figure 7.10

401(k) Sponsors Use a Variety of Plan Designs

Percentage of plans with selected plan activity combinations, 2014

Figure 7.10

Download an Excel file of this data.

Note: The sample is 50,115 plans with $3.7 trillion in assets. The results exclude 403(b) plans with a 401(k) feature and plans with fewer than 100 participants or less than $1 million in plan assets. A plan was determined to allow participant loans if any participant had a loan outstanding at the end of plan year 2014. Components do not add to 100 percent because of rounding.
Source: BrightScope Defined Contribution Plan Database. See BrightScope and Investment Company Institute, The BrightScope/ICI Defined Contribution Plan Profile: A Close Look at 401(k) Plans, 2014.

Investment Lineup

In addition to choosing how to structure contributions to the 401(k) plan, employers also select the investment options to make available to plan participants. In 2014, domestic equity funds, international equity funds, and domestic bond funds were offered in nearly all large 401(k) plans (Figure 7.11). Although these three fund types are equally likely to be offered, when these funds are available in the plan, employers tend to offer more domestic equity funds (10 funds on average) than domestic bond funds (four funds) or international equity funds (three funds). Target date funds also are common investment choices, with about three-quarters of large 401(k) plans offering nine of these funds on average. In addition, about half of large 401(k) plans offered one money fund on average and seven in 10 offered one guaranteed investment contract (GIC). In total, the average large 401(k) plan offered 28 funds to participants in 2014. Large private-sector 403(b) plans also offer participants a diverse array of investment options to choose from.

Figure 7.11

Incidence of Investment Options Offered in 401(k) Plans by Type of Investment

Percentage of plans with audited 401(k) filings in the BrightScope database, 2014

Figure 7.11

Download an Excel file of this data.

1 A target date fund typically rebalances its portfolio to become less focused on growth and more focused on income as it approaches and passes the target date of the fund, which is usually included in the fund’s name.
2 The Investment Company Institute classifies balanced funds as hybrid in its data.
3 Other includes commodity funds, real estate funds, and individual stocks (including company stock) and bonds.
Note: The sample is 29,958 plans with $3.4 trillion in assets. Participant loans are excluded. Funds include mutual funds, collective investment trusts, separate accounts, and other pooled investment products. BrightScope audited 401(k) filings generally include plans with 100 participants or more. Plans with fewer than four investment options, more than 100 investment options, or less than $1 million in plan assets are excluded from this analysis.
Source: BrightScope Defined Contribution Plan Database. See BrightScope and Investment Company Institute, The BrightScope/ICI Defined Contribution Plan Profile: A Close Look at 401(k) Plans, 2014.

401(k) Participants: Asset Allocation, Account Balances, and Loan Activity

Asset Allocation

The income that 401(k) plan accounts provide in retirement depends, in part, on the asset allocation decisions of plan participants.

On average, younger participants allocate more of their portfolios to equities (which include equity mutual funds and other pooled equity investments; the equity portion of balanced funds,* including target date funds; and company stock of their employers). According to research conducted by ICI and the Employee Benefit Research Institute (EBRI), at year-end 2014, individuals in their twenties had 37 percent of their 401(k) assets in equity funds and company stock; 48 percent in target date funds and non–target date balanced funds; and only 8 percent in GICs and other stable value funds, money funds, and bond funds (Figure 7.12). All told, participants in their twenties had 78 percent of their 401(k) assets in equities. By comparison, at year-end 2014, participants in their sixties had 26 percent of their 401(k) account assets in GICs and other stable value funds, money funds, and bond funds; only 23 percent in target date funds and non–target date balanced funds; and 45 percent in equity funds and company stock. All told, participants in their sixties had 56 percent of their 401(k) assets in equities.

* The Investment Company Institute classifies balanced funds as hybrid in its data.


Figure 7.12

401(k) Asset Allocation Varied with Participant Age

Average asset allocation of 401(k) account balances, percentage of account balances, year-end 2014

Figure 7.12

Download an Excel file of this data.

1 A target date fund typically rebalances its portfolio to become less focused on growth and more focused on income as it approaches and passes the target date of the fund, which is usually included in the fund’s name.
2 The Investment Company Institute classifies balanced funds as hybrid in its data.
3 Equities include equity funds, company stock, and the equity portion of balanced funds.
Note: Funds include mutual funds, bank collective trusts, life insurance separate accounts, and any pooled investment product primarily invested in the security indicated. Percentages are dollar-weighted averages.
Source: Tabulations from EBRI/ICI Participant-Directed Retirement Plan Data Collection Project. See ICI Research Perspective, “401(k) Plan Asset Allocation, Account Balances, and Loan Activity in 2014.”

Portfolio allocation also varies widely within age groups. At year-end 2014, 75 percent of 401(k) participants in their twenties held more than 80 percent of their account in equities, and 9 percent of these participants held 20 percent or less (Figure 7.13). Of 401(k) participants in their sixties, 22 percent held more than 80 percent of their account in equities, and 18 percent held 20 percent or less.

Figure 7.13

Asset Allocation to Equities Varied Widely Among 401(k) Plan Participants

Asset allocation distribution of 401(k) participant account balance to equities, percentage of participants, year-end 2014

Figure 7.13

Download an Excel file of this data.

Note: Equities include equity funds, company stock, and the equity portion of balanced funds. Funds include mutual funds, bank collective trusts, life insurance separate accounts, and any pooled investment product invested primarily in the security indicated. Components do not add to 100 percent because of rounding. The Investment Company Institute classifies balanced funds as hybrid in its data.
Source: Tabulations from EBRI/ICI Participant-Directed Retirement Plan Data Collection Project. See ICI Research Perspective, “401(k) Plan Asset Allocation, Account Balances, and Loan Activity in 2014.”

Target Date Funds

Target date funds, introduced in the mid-1990s, have grown rapidly in recent years. A target date fund (including both target date mutual funds and other pooled target date investments) follows a predetermined reallocation of assets over time based on a specified target retirement date. Typically the fund rebalances its portfolio to become less focused on growth and more focused on income as it approaches and passes the target date, which is usually indicated in the fund’s name.

The share of 401(k) plans that offer target date funds, the share of 401(k) plan participants offered target date funds, and the share of 401(k) participants holding target date funds all have increased. At year-end 2014, target date funds accounted for 18 percent of 401(k) assets, up from 5 percent at year-end 2006 (Figure 7.14). Seventy-two percent of 401(k) plans offered target date funds and 73 percent of 401(k) plan participants were offered target date funds in 2014. The percentage of 401(k) participants with target date fund assets (48 percent) was lower than the percentage of participants who were offered the option because not all plan participants choose to allocate assets to these funds. Similarly, the share of 401(k) assets in target date funds (18 percent) was lower than the share of participants invested in these funds because not all participants with assets in target date funds allocated 100 percent of their holdings to these funds, and because participants with assets in these funds were more likely to be younger or recently hired and have lower account balances.

Figure 7.14

Target Date Funds’ 401(k) Market Share

Percentage of total 401(k) market; year-end, 2006 and 2014

Figure 7.14

Download an Excel file of this data.
Note: Funds include mutual funds, bank collective trusts, life insurance separate accounts, and other pooled investment products.
Source: Tabulations from EBRI/ICI Participant-Directed Retirement Plan Data Collection Project. See ICI Research Perspective, “401(k) Plan Asset Allocation, Account Balances, and Loan Activity in 2014.”

Account Balances

Account balances tended to be higher the longer 401(k) plan participants had been working for their current employers and the older the participant. Participants in their sixties with more than 30 years of tenure at their current employer had an average 401(k) account balance of $274,043 at year-end 2014 (Figure 7.15). Participants in their forties with five to 10 years of tenure at their current employer had an average 401(k) balance of $66,173. The median 401(k) plan participant was 46 years old at year-end 2014, and the median job tenure was eight years.

Figure 7.15

401(k) Balances Tend to Increase with Participant Age and Job Tenure

Average 401(k) account balance by participant age and tenure, 2014

Figure 7.15

Download an Excel file of this data.

Source: Tabulations from EBRI/ICI Participant-Directed Retirement Plan Data Collection Project. See ICI Research Perspective, “401(k) Plan Asset Allocation, Account Balances, and Loan Activity in 2014.”

Loan Activity

Most 401(k) participants do not borrow from their plans, although the majority have access to loans. At year-end 2014, 20 percent of participants eligible for loans had loans outstanding, down slightly from year-end 2013. Not all participants, however, have access to 401(k) plan loans—factoring in all 401(k) participants with and without loan access in the EBRI/ICI 401(k) database, only 17 percent had loans outstanding at year-end 2014. The average unpaid loan balances among participants with loans represented about 11 percent of their 401(k) account balances (net of the unpaid loan balances). In aggregate, US Department of Labor data indicate that outstanding loan amounts were less than 2 percent of 401(k) plan assets in 2014.

Individual Retirement Accounts

The first type of IRA—known as a traditional IRA—was created under the Employee Retirement Income Security Act of 1974 (ERISA). IRAs provide all workers with a contributory retirement savings vehicle and, through rollovers, give workers leaving jobs a means to preserve the tax benefits and growth opportunities that employer-sponsored retirement plans provide. Roth IRAs, first available in 1998, were created to provide a contributory retirement savings vehicle on an after-tax basis with qualified withdrawals distributed tax-free. In addition, policymakers have added employer-sponsored IRAs (SEP IRAs, SAR-SEP IRAs, and SIMPLE IRAs) to encourage small employers to provide retirement plans by simplifying the rules applicable to tax-qualified plans.

Total IRA assets, $7.9 trillion at year-end 2016, accounted for 31 percent of US retirement assets. Mutual funds accounted for $3.7 trillion of IRA assets at year-end 2016, up from $3.5 trillion at year-end 2015 (Figure 7.16). Assets managed by mutual funds were the largest component of IRA assets (representing 47 percent), followed by the other assets category, which includes ETFs, closed-end funds, individual stocks and bonds, and other non–mutual fund securities held through brokerage accounts ($3.2 trillion at year-end 2016).

Figure 7.16

IRA Assets

Trillions of dollars; year-end, selected years

Figure 7.16

Download an Excel file of this data.

1 Other assets includes individual stocks, individual bonds, closed-end funds, ETFs, and other assets held through brokerage or trust accounts.
2 Life insurance company IRA assets are annuities held by IRAs, excluding variable annuity mutual fund IRA assets, which are included in mutual funds.
3 Bank and thrift deposits include Keogh deposits.
e Data are estimated.
Note: Components may not add to the total because of rounding.
Sources: Investment Company Institute, Federal Reserve Board, American Council of Life Insurers, and Internal Revenue Service Statistics of Income Division. See Investment Company Institute, “The US Retirement Market, Fourth Quarter 2016.”

IRA Investors

Approximately one-third of US households, or nearly 43 million, owned at least one type of IRA as of mid-2016 (Figure 7.17). Traditional IRAs—those introduced under ERISA—were the most common type, owned by 32 million US households. Roth IRAs, first available in 1998 under the Taxpayer Relief Act of 1997, were owned by 22 million US households. More than 7 million US households owned employer-sponsored IRAs (SEP IRAs, SAR-SEP IRAs, or SIMPLE IRAs).

Investment returns and rollovers from employer-sponsored retirement plans, rather than new contributions, have fueled the growth of IRAs. For example, the Internal Revenue Service Statistics of Income division reports $435 billion was rolled over to IRAs in tax year 2014, compared with $63 billion that was contributed. Although most US households are eligible to make contributions to IRAs, few do so. Indeed, only 11 percent of US households contributed to traditional or Roth IRAs in tax year 2015 and very few eligible households made “catch-up” contributions.

Figure 7.17

Nearly 43 Million US Households Owned IRAs

 Year createdNumber of US households with
type of IRA
Mid-2016
Percentage of US households with type of IRA
Mid-2016
Assets in IRAs
Billions of dollars,
year-end 2016
Traditional IRA 1974
(Employee Retirement Income Security Act)
32.1 million 25.5% $6,695e
SEP IRA 1978
(Revenue Act)
7.2 million 5.7% $495e
SAR-SEP IRA 1986
(Tax Reform Act)
SIMPLE IRA 1996
(Small Business Job Protection Act)
Roth IRA 1997
(Taxpayer Relief Act)
21.9 million 17.4% $660e
Any IRA   42.5 million 33.8% $7,850e

Download an Excel file of this data.

e Data are estimated.
Note: Households may own more than one type of IRA. SEP IRAs, SAR-SEP IRAs, and SIMPLE IRAs are employer-sponsored IRAs.
Sources: Investment Company Institute and US Census Bureau. See ICI Research Perspective, “The Role of IRAs in US Households’ Saving for Retirement, 2016” and “The US Retirement Market, Fourth Quarter 2016.”

Instead, investment returns and rollovers from employer-sponsored retirement plans have fueled the growth of IRAs. In any given year, a small portion of traditional IRA investors make rollovers, but analysis of The IRA Investor Database—which contains information on more than 16 million IRA investors—finds that, for the most part, the people who make rollovers differ from year to year. Rollovers play an important role in opening traditional IRAs. With the availability of retirement accumulations that can be rolled over when a worker separates from an employer, whether from DC accounts or as lump-sum distributions from DB plans, most (85 percent) new traditional IRAs in 2014 were opened only with rollovers (Figure 7.18). By contrast, in 2014, 12 percent of Roth IRAs were opened only with rollovers; the majority (74 percent) were opened only with contributions.

A substantial share of traditional IRA investors has rolled over assets from an employer plan. In any given year, only a small portion of traditional IRA investors have a rollover, but, for the most part, the people who make rollovers differ from year to year. For example, each year from 2007 through 2014 about one in 10 traditional IRA investors in The IRA Investor Database had a rollover, but nearly half of investors with traditional IRAs at year–end 2014 had a rollover at some point during the eight-year span.

Figure 7.18

New Roth IRAs Often Are Opened with Contributions; New Traditional IRAs Often Are Opened with Rollovers

Percentage of new IRAs opened in 2014 by type of IRA

Figure 7.18

Download an Excel file of this data.

Note: New IRAs are accounts that did not exist in The IRA Investor Database in 2013 and were opened in 2014 by one of the paths indicated. The calculation excludes IRAs that changed financial services firms. The samples are 0.3 million new Roth IRA investors aged 18 or older at year-end 2014 and 0.8 million new traditional IRA investors aged 25 to 74 at year-end 2014. Components may not add to 100 percent because of rounding.
Source: The IRA Investor Database™. See ICI Research Report, “The IRA Investor Profile: Roth IRA Investors’ Activity, 2007–2014.”

Traditional IRA–owning households generally researched the decision to roll over money from their former employer’s retirement plan into a traditional IRA. The most common source of information was professional financial advisers. Advisers were consulted by 60 percent of traditional IRA–owning households with rollovers, with half indicating they primarily relied on financial professionals (Figure 7.19). Older households were more likely to consult professional financial advisers than younger households. Seven percent of traditional IRA–owning households with rollovers indicated their primary source of information was online materials from financial services firms, with younger households more likely to rely on online resources as their primary source of information than were older households. Ten percent of rollover households primarily relied on information from their employers.

Figure 7.19

Multiple Sources of Information Are Consulted for the Rollover Decision

Percentage of traditional IRA–owning households with rollovers, mid-2016

Figure 7.19

Download an Excel file of this data.

1 Multiple responses are included; 67 percent of traditional IRA–owning households with rollovers consulted multiple sources of information.
2 Other responses given included: myself, other online information, bank, and books and magazines.
Source: Investment Company Institute IRA Owners Survey. See ICI Research Perspective, “The Role of IRAs in US Households’ Saving for Retirement, 2016.”

Households owning IRAs generally are headed by middle-aged individuals (median age of 54 years) with moderate household incomes (median income of $90,000). These households held a median of $70,000 in traditional or Roth IRAs. In addition, many households held multiple types of IRAs. For example, 43 percent of households with traditional IRAs also owned Roth IRAs, and 12 percent also owned employer-sponsored IRAs.

IRA Portfolios

At year-end 2014, younger IRA investors tended to have more invested in equities, equity funds, and target date funds, on average, than older investors, according to The IRA Investor Database. Older investors were invested more heavily in non–target date balanced funds and fixed-income investments. For example, traditional IRA investors in their thirties had, on average, 53 percent of their assets in equities and equity funds and another 20 percent in target date funds (Figure 7.20). Traditional IRA investors in their sixties held 53 percent and 6 percent of their traditional IRA assets, respectively, in these two asset categories. Traditional IRA investors in their sixties had 39 percent of their assets in money market funds (10 percent), bonds and bond funds (17 percent), and non–target date balanced funds (12 percent). By contrast, traditional IRA investors in their thirties held 23 percent of their assets in these three asset categories.

Roth IRA investors display a similar pattern of investing by age, although Roth IRA investors of all ages tended to have higher allocations to equities and equity funds compared with traditional IRA investors—for example, Roth IRA investors in their thirties and sixties held about the same portion of their assets (63 and 64 percent, respectively) in equities and equity funds (Figure 7.20). Roth IRA investors in their thirties had, on average, 19 percent of their assets in target date funds, while Roth IRA investors in their sixties had 4 percent. By contrast, Roth IRA investors in their sixties had 30 percent of their assets in money market funds (8 percent), bonds and bond funds (9 percent), and non–target date balanced funds (13 percent). Roth IRA investors in their thirties held 17 percent of their assets in these three asset categories.


Figure 7.20

IRA Asset Allocation Varied with Investor Age

Average asset allocation of IRA balances, percentage of assets, year-end 2014

Figure 7.20

Download an Excel file of this data.

1 Other investments includes certificates of deposit and unidentifiable assets.
2 Bond funds include bond mutual funds, bond closed-end funds, and bond ETFs.
3 Balanced funds invest in a mix of equities and fixed-income securities. The Investment Company Institute classifies balanced funds as hybrid in its data.
4 A target date fund typically rebalances its portfolio to become less focused on growth and more focused on income as it approaches and passes the target date of the fund, which is usually included in the fund’s name.
5 Equity funds include equity mutual funds, equity closed-end funds, and equity ETFs.
Note: Percentages are dollar-weighted averages. Components may not add to 100 percent because of rounding.
Source: The IRA Investor Database™. See ICI Research Report, “The IRA Investor Profile: Traditional IRA Investors’ Activity, 2007–2014” and ICI Research Report, “The IRA Investor Profile: Roth IRA Investors’ Activity, 2007–2014.”

Distributions from IRAs

Withdrawals from IRAs tend to occur later in life, often to fulfill required minimum distributions (RMDs). An RMD is equal to a percentage of the IRA balance, based on remaining life expectancy. Traditional IRA owners aged 70½ or older generally must withdraw at least the minimum amount each year or pay a penalty. In tax year 2015, 71 percent of households that took traditional IRA withdrawals stated they calculated the withdrawal amount based on RMD rules.

In contrast to traditional IRAs, Roth IRAs have no RMDs (unless they are inherited). As a result, withdrawal activity is much lower among Roth IRA investors. In 2014, only 4 percent of Roth IRA investors aged 25 or older made withdrawals, compared with 23 percent of traditional IRA investors (Figure 7.21). Early withdrawal penalties can apply to both Roth and traditional IRA investors younger than 59½, and withdrawal activity is lower among investors younger than 60 compared with investors aged 60 or older.

Figure 7.21

Roth IRA Investors Rarely Take Withdrawals; Traditional IRA Investors Are Heavily Affected by RMDs

Percentage of IRA investors with withdrawals by type of IRA and investor age, 2014

Figure 7.21

Download an Excel file of this data.

Note: The samples are 5.2 million Roth IRA investors aged 25 or older at year-end 2014 and 11.0 million traditional IRA investors aged 25 or older at year-end 2014.
Source: The IRA Investor Database™. See ICI Research Report, “The IRA Investor Profile: Roth IRA Investors’ Activity, 2007–2014.”

Withdrawals from IRAs tend to be retirement related. Of the 25 percent of traditional IRA–owning households who reported taking withdrawals in 2015, 88 percent reported that the head of household, the spouse, or both were retired. Of retired households that took traditional IRA withdrawals in 2015, 45 percent reported using some or all of the withdrawal amount to pay for living expenses (Figure 7.22). Among retired households, other uses included reinvesting or saving in another account (42 percent), buying, repairing, or remodeling a home (16 percent), and paying for a healthcare expense (9 percent).

Traditional IRA–owning households that reported taking withdrawals in 2015 and were not retired indicated a slightly different pattern for the withdrawals. The nonretired households with withdrawals were less likely to indicate using some or all of the money for living expenses (30 percent) or to reinvest or save it in another account (29 percent) than the retired households (Figure 7.22).

Figure 7.22

Traditional IRA Withdrawals Among Retirees Often Are Used to Pay for Living Expenses

Percentage of traditional IRA–owning households by retirement status,1 mid-2016

  Purpose of traditional IRA withdrawalRetired1, 2Not retired3
Took withdrawals to pay for living expenses 45 30
Spent it on a car, boat, or big-ticket item other than a home 6 10
Spent it on a healthcare expense 9 9
Used it for an emergency 6 9
Used it for home purchase, repair, or remodeling 16 14
Reinvested or saved it in another account 42 29
Paid for education 1 4
Some other purpose 10 15

Download an Excel file of this data.

1 The household was considered retired if either the head of household or spouse responded affirmatively to the question: “Are you retired from your lifetime occupation?”
2 The base of respondents includes the 22 percent of traditional IRA–owning households that were retired in mid-2016 and took withdrawals in tax year 2015.
3 The base of respondents includes the 3 percent of traditional IRA–owning households that were not retired in mid-2016 and took withdrawals in tax year 2015.
Note: Multiple responses are included.
Source: Investment Company Institute IRA Owners Survey. See ICI Research Perspective, “The Role of IRAs in US Households’ Saving for Retirement, 2016.”

The Role of Mutual Funds in Retirement Savings

Mutual funds play a major role in employer-sponsored DC plans, such as 401(k) plans, and IRAs. At year-end 2016, mutual funds accounted for 55 percent of DC plan assets and 47 percent of IRA assets (Figure 7.23). Investors held slightly more mutual fund assets in DC plans ($3.9 trillion) than in IRAs ($3.7 trillion). Among DC plans, 401(k) plans held the most assets in mutual funds, with $3.0 trillion, followed by 403(b) plans ($435 billion), other private-sector DC plans ($301 billion), and 457 plans ($102 billion) (Figure 7.24). Combined, the $7.6 trillion of mutual fund assets held in DC plans and IRAs at the end of 2016 accounted for 30 percent of the $25.3 trillion US retirement market.

Figure 7.23

Substantial Amount of Retirement Assets Are Invested in Mutual Funds

Assets, billions of dollars, year-end 2016

Figure 7.23

Download an Excel file of this data.

e Data are estimated.
Source: Investment Company Institute, “The US Retirement Market, Fourth Quarter 2016”

Assets in DC plans and IRAs represent a large share of mutual fund assets overall, and long-term mutual fund assets in particular (Figure 7.23). The $7.6 trillion in mutual fund retirement assets made up 46 percent of all mutual fund assets at year-end 2016. DC plans and IRAs held 53 percent of equity, hybrid, and bond mutual fund assets, but only 14 percent of money market fund assets.

Types of Mutual Funds Used by Retirement Plan Investors

Retirement investors tend to hold equity investments. At year-end 2016, 57 percent of the $7.6 trillion in mutual fund retirement assets held in DC plans and IRAs were invested in domestic or world equity funds (Figure 7.24). By comparison, 52 percent of overall fund industry assets—retirement and nonretirement accounts—were invested in domestic or world equity funds. Domestic equity funds alone constituted about $3.3 trillion, or 44 percent, of mutual fund assets held in DC plans and IRAs.

Figure 7.24

Majority of Mutual Fund Retirement Account Assets Were Invested in Equities

Billions of dollars, year-end 2016

 EquityHybrid1BondMoney
market
Total
DomesticWorld
IRAs2 1,547 466 831 622 243 3,710
DC plans 1,776 519 988 443 129 3,854
    401(k) plans 1,346 432 835 316 86 3,016
    403(b) plans 264 34 81 36 19 435
    457 plans 55 14 21 11 1 102
    Other private-sector DC plans3 109 40 51 79 22 301
Total  3,323  985  1,819  1,065  373  7,565

Download an Excel file of this data.

1 Hybrid funds invest in a mix of equities and fixed-income securities. Most target date and lifestyle funds are counted in this category.
2 IRAs include traditional IRAs, Roth IRAs, and employer-sponsored IRAs (SEP IRAs, SAR-SEP IRAs, and SIMPLE IRAs).
3 Other private-sector DC plans includes Keoghs and other private-sector DC plans without 401(k) features.
Note: Components may not add to the totals because of rounding.
Source: Investment Company Institute, “The US Retirement Market, Fourth Quarter 2016”

Retirement investors also gain exposure to equities and fixed-income securities through hybrid funds. At year-end 2016, 24 percent of mutual fund assets held in DC plans and IRAs were held in hybrid funds, which invest in a mix of equity, bond, and money market securities (Figure 7.24).

The remaining 19 percent of mutual fund assets held in DC plans and IRAs at the end of 2016 were invested in bond funds and money market funds. Bond funds held $1.1 trillion, or 14 percent, of mutual fund assets held in DC plans or IRAs, and money market funds accounted for $373 billion, or 5 percent (Figure 7.24).

Target Date and Lifestyle Mutual Funds

Target date and lifestyle mutual funds, generally included in the hybrid fund category, have grown more popular among investors and retirement plan sponsors over the past decade. A target date fund follows a predetermined reallocation of assets over time based on a specified target retirement date. Typically the fund rebalances its portfolio to become less focused on growth and more focused on income as it approaches and passes the target date, which is usually indicated in the fund’s name. A lifestyle fund maintains a predetermined risk level and generally uses words such as “conservative,” “moderate,” or “aggressive” in its name to indicate the fund’s risk level.

Assets in target date and lifestyle mutual funds totaled $1.3 trillion at year-end 2016, up from $1.1 trillion at year-end 2015 (Figure 7.25). Target date mutual funds’ assets were up 16 percent in 2016, increasing from $763 billion to $887 billion. Assets in lifestyle mutual funds were $372 billion at year-end 2016, unchanged for the year. Most (88 percent) target date mutual fund assets were held in retirement accounts, compared with 44 percent of lifestyle mutual fund assets.

Figure 7.25

Target Date and Lifestyle Mutual Fund Assets by Account Type

Billions of dollars; year-end, 2005–2016

Figure 7.25

Download an Excel file of this data.

1 IRAs include traditional IRAs, Roth IRAs, and employer-sponsored IRAs (SEP IRAs, SAR-SEP IRAs, and SIMPLE IRAs).
2 DC plans include 401(k) plans, 403(b) plans, 457 plans, Keoghs, and other DC plans without 401(k) features.
3 A target date mutual fund typically rebalances its portfolio to become less focused on growth and more focused on income as it approaches and passes the target date of the fund, which is usually included in the fund’s name.
4 A lifestyle mutual fund maintains a predetermined risk level and generally contains “conservative,” “moderate,” or “aggressive” in its name.
Note: Components may not add to the total because of rounding.
Source: Investment Company Institute, “The US Retirement Market, Fourth Quarter 2016”

Target date, lifestyle, and index mutual funds have grown as a share of mutual fund assets in DC plans. Target date mutual funds increased 13 percentage points as a share of DC plans’ mutual fund assets from year-end 2005 to year-end 2016, increasing from 3 percent to 16 percent (Figure 7.26). At year-end 2016, target date, lifestyle, and index mutual funds made up 37 percent of mutual fund assets in DC plans compared with only 15 percent in 2005.

Figure 7.26

Target Date, Lifestyle, and Index Funds Have Risen as a Share of DC Plans’ Mutual Fund Assets

Percentage of mutual fund assets held in DC plans;1 year-end, 2005–2016

Figure 7.26

Download an Excel file of this data.

1 DC plans include 401(k) plans, 403(b) plans, 457 plans, Keoghs, and other private-sector DC plans without 401(k) features.
2 A target date mutual fund typically rebalances its portfolio to become less focused on growth and more focused on income as it approaches and passes the target date of the fund, which is usually included in its name.
3 A lifestyle mutual fund maintains a predetermined risk level and generally contains “conservative,” “moderate,” or “aggressive” in the fund’s name.
4 Index mutual funds are equity, bond, and hybrid funds that target specific market indexes with the general objective of meeting the performance of that index. Equity index funds are the most common type of index fund.
Note: Components may not add to the total because of rounding.
Source: Investment Company Institute, “The US Retirement Market, Fourth Quarter 2016”

The Role of Mutual Funds in Education Savings

Twenty-two percent of households that owned mutual funds in 2016 cited education as a financial goal for their fund investments (Figure 6.9). Nevertheless, the demand for education savings vehicles has been historically modest since their introduction in the 1990s, partly because of their limited availability and investors’ lack of familiarity with them. The Economic Growth and Tax Relief Reconciliation Act (EGTRRA), enacted in 2001, enhanced the attractiveness of Section 529 plans and Coverdell Education Savings Accounts (ESAs)—two education savings vehicles—by making them more flexible and allowing larger contributions. The Pension Protection Act (PPA), enacted in 2006, made the EGTRRA enhancements to Section 529 plans permanent. The Tax Relief, Unemployment Insurance Reauthorization, and Job Creation Act of 2010 extended the EGTRRA enhancements to Coverdell ESAs for two years; the American Taxpayer Relief Act of 2012 made these enhancements permanent.

Assets in Section 529 savings plans increased 9 percent in 2016, with $251 billion at year-end 2016, up from $230 billion at year-end 2015 (Figure 7.27). As of year-end 2016, there were 11.7 million 529 savings plan accounts, with an average account size of approximately $21,400.

Figure 7.27

Section 529 Savings Plan Assets

Billions of dollars; year-end, selected years

Figure 7.27

Download an Excel file of this data.

Note: Data were estimated for a few individual state observations in order to construct a continuous time series.
Sources: Investment Company Institute and College Savings Plans Network. See Investment Company Institute, “529 Plan Program Statistics, December 2016.”

In mid-2016, as a group, households saving for college through 529 plans, Coverdell ESAs, or mutual funds held outside these accounts tended to be headed by younger individuals, with 48 percent younger than 45 (Figure 7.28). Heads of households saving for college had a range of education attainment: 42 percent had less than a college degree and 58 percent had a college degree or more. These households also had a range of incomes: 30 percent earned less than $75,000; 15 percent earned between $75,000 and $99,999; and 55 percent earned $100,000 or more. About six in 10 of these households had children (younger than 18) in the home, and 37 percent had more than one child in the home.

Figure 7.28

Characteristics of Households Saving for College

Percentage of US households saving for college,1 mid-2016

 
Age of head of household2
Younger than 35 21
35 to 44 27
45 to 54 26
55 to 64 14
65 or older 12
Education level
High school diploma or less 16
Associate’s degree or some college 26
Completed college 24
Some graduate school or completed graduate school 34
Household income3
Less than $25,000 6
$25,000 to $34,999 3
$35,000 to $49,999 7
$50,000 to $74,999 14
$75,000 to $99,999 15
$100,000 or more 55
Number of children in home4
None 39
One 24
Two 24
Three or more 13

Download an Excel file of this data.

1 Households saving for college are households that own education savings plans (Coverdell ESAs or 529 plans) or that said paying for education was one of their financial goals for their mutual funds.
2 Age is based on the sole or co-decisionmaker for saving and investing.
3 Total reported is household income before taxes in 2015.
4 The number of children reported is children younger than 18 living in the home.

Back to Top