401(k) plan asset allocation, account balances, and loan activity in 2010

THE BULK OF 401(K) ASSETS CONTINUED TO BE INVESTED IN STOCKS: On average, at year-end 2010, 62 percent of 401(k) participants' assets were invested in equity securities through equity funds, the equity portion of balanced funds, and company stock. Thirty-three percent were in fixed-income securities such as stable value investments and bond and money funds. SEVENTY PERCENT OF 401(K) PLANS INCLUDED TARGET-DATE FUNDS IN THEIR INVESTMENT LINEUP AT YEAR-END 2010: At year-end 2010, 11 percent of the assets in the EBRI/ICI 401(k) database were invested in target-date funds and 36 percent of 401(k) participants held target-date funds. Also known as lifecycle funds, they are designed to offer a diversified portfolio that automatically rebalances to be more focused on income over time. MORE NEW OR RECENT HIRES INVESTED THEIR 401(K) ASSETS IN BALANCED FUNDS, INCLUDING TARGET-DATE FUNDS: For example, at year-end 2010, 44 percent of the account balances of recently hired participants in their 20s were invested in balanced funds, compared with 42 percent in 2009, and about 7 percent in 1998. A significant subset of that balanced fund category is target-date funds. At year-end 2010, 35 percent of the account balances of recently hired participants in their 20s were invested in target-date funds, compared with 31 percent at year-end 2009. 401(K) PARTICIPANTS CONTINUED TO SEEK DIVERSIFICATION OF THEIR INVESTMENTS: The share of 401(k) accounts invested in company stock continued to shrink, falling by more than a percentage point (to 8 percent) in 2010, continuing a steady decline that started in 1999. Recently hired 401(k) participants contributed to this trend: They tended to be less likely to hold employer stock. PARTICIPANTS' 401(K) LOAN BALANCES DECLINED SLIGHTLY IN 2010: In 2010, 21 percent of all 401(k) participants who were eligible for loans had loans outstanding against their 401(k) accounts, unchanged from year-end 2009, and up from 18 percent at year-end 2008. Loans outstanding amounted to 14 percent of the remaining account balance, on average, at year-end 2010, compared with 15 percent at year-end 2009. Loan amounts outstanding declined slightly from those in the past few years. THE YEAR-END 2010 AVERAGE ACCOUNT BALANCE IN THE DATABASE WAS 3.4 PERCENT HIGHER THAN THE YEAR BEFORE, BUT MAY NOT ACCURATELY REFLECT THE EXPERIENCE OF TYPICAL 401(K) PARTICIPANTS IN 2010: To understand changes in 401(k) participants' average account balances, it is important to analyze a sample of consistent participants. As with previous EBRI/ICI updates, analysis of a sample of consistent 401(k) participants (those that have been in the same plan since 2003) is expected to be published in 2012.     

401(k) plan asset allocation, account balances, and loan activity in 1998

The Employee Benefit Research Institute (EBRI) and the Investment Company Institute (ICI) have been collaborating for the past three years to collect data on participants in 401(k) plans. This effort, known as the EBRI/ICI Participant-Directed Retirement Plan Data Collection Project, has obtained data for 401(k) plan participants from certain of EBRI and ICI members serving as plan record keepers and administrators. The report includes 1998 information on 7.9 million active participants in 30,102 plans holding nearly $372 billion in assets. The data include demographic information, annual contributions, plan balances, asset allocation, and loans, and are broadly representative of the universe of 401(k) plans. The database also includes three years of longitudinal information on approximately 3.3 million participants. Key findings include: For all 401(k) participants in the 1998 EBRI/ICI database, almost three-quarters of plan balances are invested directly or indirectly in equity securities. Specifically, 49.8 percent of total plan balances are invested in equity funds, 17.7 percent in company stock, 11.4 percent in guaranteed investment contracts (GICs), 8.4 percent in balanced funds, 6.1 percent in bond funds, 4.7 percent in money funds, and 0.3 percent in other stable value funds. Participant asset allocation varies considerably with age. Younger participants tend to favor equity funds, while older participants are more disposed to invest in GICs and bond funds. On average, participants in their 20s have 62.1 percent of their account balances invested in equity funds, in contrast to 39.8 percent for those in their 60s. Participants in their 20s invest 4.7 percent of their assets in GICs, while those in their 60s invest 20.6 percent. Bond funds, which represent 4.7 percent of the assets of participants in their 20s, amount to 9.0 percent of the assets of participants in their 60s. Investment options offered by 401(k) plans appear to influence asset allocation. For example, the addition of company stock substantially reduces the allocation to equity funds and the addition of GICs lowers allocations to bond and money funds. The average account balance (net of plan loans) for all participants was $47,004 at year-end 1998, which is 26 percent higher than the average account balance at year-end 1996. The median account balance was $13,038 at year-end 1998. The balances, however, represent only amounts with current employers and do not include amounts remaining in the plans of prior employers. The average balances of older workers with long tenure indicate that a mature 401(k) plan program will produce substantial account balances. For example, individuals in their 60s with at least 30 years of tenure have average account balances in excess of $185,000. The ratio of account balance to 1998 salary varies with salary, increasing slightly as earnings rise from $20,001 to $80,000, and falling a bit for salaries greater than $80,000. The increase in ratio likely reflects a greater propensity of higher-income participants to save, whereas the decline after $80,000 results from contribution and nondiscrimination rule constraints.     

Contribution behavior of 401(k) plan participants

This Issue Brief examines the 1999 contribution behavior of 1.7 million 401(k) plan participants drawn from the EBRI/ICI Participant-Directed Retirement Plan Data Collection Project. The findings in this paper build on previous academic research examining the contribution activity of 401(k) participants, by using a large sample of participants in a wide range of plan sizes and by examining in detail the factors that influence contribution activity. Eighty-five percent of participants in the sample only made before-tax contributions to their plans, and 97 percent of all dollars contributed by employees were contributed on a before-tax basis. On average, participants contributed 6.8 percent of their salaries on a before-tax basis. Before-tax contribution activity varied among participants. About 61 percent of participants contributed more than 5 percent of their salaries on a before-tax basis and about 21 percent set aside more than 10 percent of their salaries on a before-tax basis. Eleven percent of participants analyzed in this study earning more than $40,000 a year contributed at the $10,000 before-tax IRC limit in 1999. Thirteen percent of participants with salaries between $70,000 and $80,000 contributed at the cap, and 18 percent of those with salaries between $80,000 and $90,000 were at the limit. However, it appears that among participants not contributing at the IRC limit, 52 percent could not have done so because of formal plan-imposed contribution limits below the IRC limit. Older participants tended to contribute a higher percentage of their salaries to plans than did younger participants, even after factoring out differences in salary and job tenure. Participants tended to increase the share of their salary (and amounts) contributed to their 401(k) plan as their salaries rose until salaries reached $80,000. For individuals with salaries above $80,000, before-tax contribution rates (though not the amounts contributed) tended to fall as salaries rose because IRC, and possibly plan sponsor, contribution limits became binding for some participants. Giving employees the option of borrowing from their 401(k) accounts increased participant contribution rates. On average, a participant in a plan offering loans appeared to contribute 0.6 percentage point more of his or her salary to the plan than a participant in a plan with no loan provision. Total contributions--the sum of employee and employer contributions--were higher for participants who received an employer contribution as part of their 401(k) plan than for those who did not. The average total contribution rate was 10 percent of salary for employees in plans offering an employer contribution, compared with 7.4 percent for those in plans not offering an employer contribution.     

Misinformed and informed asset allocation decisions of self-directed retirement plan members

Most defined contribution pension plan members misunderstand asset allocation, but those with higher levels of wealth managing their own money are less likely to be confused. Younger, more-educated, higher-earning advice-receiving males with a planner mindset hold more equity. Notably, an understanding of asset allocation accentuates the impact of the key factors age, income and a planner mindset.     

Individual retirement account balances, contributions, and rollovers, 2010: the EBRI IRA database

In 2010, IRA owners were more likely to be male, especially those whose accounts originated from a rollover or were a SEP/SIMPLE. Among all IRA owners in the database, nearly one-half (45.8 percent) were ages 45-64. The average and median IRA account balance in 2010 was $67,438 and $17,863, respectively, while the average and median IRA individual balance (all accounts from the same person combined) was $91,864 and $25,296. Individuals with a traditional-originating from rollovers had the highest average and median balance of $123,426 and $38,138, respectively. Roth owners had the lowest average and median balance at $22,437 and $11,471. The average and median individual IRA balance increased with age through age 70. The average amount contributed to an IRA in the database was $3,335 in 2010. The average contribution was highest for accounts owned by those ages 65-69, and more contributions were made to Roth accounts than to traditional accounts (both those originating from contributions and rollovers). However, the average contribution to a traditional account was higher, at $3,517, compared with $3,240 to a Roth account. Yet, a higher overall amount was contributed to Roths ($2.3 billion for Roths compared with $1.3 billion for traditional accounts). Focusing on those owning traditional or Roth IRAs, 9.3 percent of the accounts received contributions, and 12.1 percent of the individuals owning these IRA types contributed to them in 2010. Among traditional IRA owners, 5.2 percent contributed, while 24.0 percent of those owning a Roth contributed to it during 2010. Of those individuals contributing to an IRA, 43.5 percent contributed the maximum amount. Of those contributing to a traditional IRA, 48.7 percent maxed out their contribution, while 39.3 percent did so with a Roth. The average and median account balances increased from $54,863 and $15,756 respectively in 2008 to $67,438 and $17,863 in 2010. This represents an increase of 22.9 percent in the average account balance and 13.4 percent in the median balance. The total individual balances also increased for both the average (32.2 percent) and the median (26.2 percent).     

Health savings accounts and health reimbursement arrangements: assets, account balances, and rollovers, 2006-2011

ASSET LEVELS GROWING: In 2011, there was $12.4 billion in health savings accounts (HSAs) and health reimbursement arrangements (HRAs), spread across 8.4 million accounts, according to data from the 2011 EBRI/MGA Consumer Engagement in Health Care Survey, sponsored by EBRI and Matthew Greenwald & Associates. This is up from 2006, when there were 1.3 million accounts with $873.4 million in assets, and 2010, when 5.4 million accounts held $7.3 billion in assets. AFTER LEVELING OFF, AVERAGE ACCOUNT BALANCES INCREASED: After average account balances leveled off in 2008 and 2009, and fell slightly in 2010, they increased in 2011. In 2006, account balances averaged $696. They increased to $1,320 in 2007, a 90 percent increase. Account balances averaged $1,356 in 2008 and $1,419 in 2009, 3 percent and 5 percent increases, respectively. In 2010, average account balances fell to $1,355, down 4.5 percent from the previous year. In 2011, average account balances increased to $1,470, a 9 percent increase from 2010. TOTAL AND AVERAGE ROLLOVERS INCREASE: After declining to $1,029 in 2010, average rollover amounts increased to $1,208 in 2011. Total assets being rolled over increased as well: $6.7 billion was rolled over in 2011, up from $3.7 billion in 2010. The percentage of individuals without a rollover remained at 13 percent in 2011. HEALTHY BEHAVIOR DOES NOT MEAN HIGHER ACCOUNT BALANCES AND HIGHER ROLLOVERS: Individuals who smoke have more money in their accounts than those who do not smoke. In contrast, obese individuals have less money in their account than the nonobese. There is very little difference in account balances by level of exercise. Very small differences were found in account balances and rollover amounts between individuals who used cost or quality information, compared with those who did not use such information. However, next to no relationship was found between either account balance or rollover amounts and various cost-conscious behaviors. When a difference was found, those exhibiting the cost-conscious behavior were found to have lower account balances and rollover amounts. DIFFERENCES IN ACCOUNT BALANCES: Men have higher account balances than women, older individuals have higher account balances than younger ones, account balances increase with household income, and education has a significant impact on account balances independent of income and other variables. DIFFERENCES IN ROLLOVER AMOUNTS: Men rolled over more money than women, and older individuals had higher rollover amounts than younger individuals. Rollover amounts increase with household income and education, and individuals with single coverage rolled over a slightly higher amount than those with family coverage.     

Health savings accounts and health reimbursement arrangements: assets, account balances, and rollovers, 2006-2010

ASSET LEVELS GROWING: In 2010, there was $7.7 billion in health savings accounts (HSAs) and health reimbursement arrangements (HRAs), spread across 5.7 million accounts. This is up from 2006, when there were 1.2 million accounts with $835.4 million in assets, and 2009, when 5 million accounts held $7.1 billion in assets. AFTER LEVELING OFF, AVERAGE ACCOUNT BALANCE DROPS SLIGHTLY: Increases in average account balances leveled off in 2008 and 2009, and fell slightly in 2010. In 2006, account balances averaged $696. They increased to $1,320 in 2007, a 90 percent increase. Account balances averaged $1,356 in 2008 and $1,419 in 2009, 3 percent and 5 percent increases, respectively. In 2010, average account balances fell to $1,355, down 4.5 percent from the previous year. AVERAGE ROLLOVER DECLINES, WHILE TOTAL ROLLOVERS INCREASE: Despite a decline in the average rollover amount in 2010, total assets being rolled over have been increasing. $4.2 billion was rolled over in 2010, up from $4 billion in 2009. The average rollover increased from $592 in 2006 to $1,295 in 2009, and fell to $1,029 in 2010. The percentage of individuals without a rollover decreased from 23 percent in 2006 to 10 percent in 2009 and increased slightly to 13 percent in 2010. HEALTHY BEHAVIOR MEANS HIGHER ACCOUNT BALANCES AND HIGHER ROLLOVERS: Individuals who exercised, those who did not smoke, and those who were not obese had higher account balances and higher rollovers than those with less healthy behaviors. It was also found that individuals who used cost or quality information had higher account balances and higher rollovers compared with those who did not use such information. However, no relationship was found between either account balance or rollover amounts and various cost-conscious behaviors such as checking pricing before getting services or asking for generic drugs instead of brand names, among other things. DIFFERENCES IN ACCOUNT BALANCES: Men have higher account balances than women, older individuals have higher account balances than younger ones, account balances increase with household income, and education has a significant impact on account balances independent of income and other variables. DIFFERENCES IN ROLLOVER AMOUNTS: Men rolled over more money than women, and older individuals had higher rollover amounts than younger individuals. Rollover amounts increase with household income and education, and individuals with single coverage rolled over a higher amount than those with family coverage.     

Lessons from the evolution of 401(k) retirement plans for increased consumerism in health care: an application of behavioral research

Employment-based health and retirement benefit programs have followed a similar path of evolution. The relative decision-making roles of the employer and the worker have shifted from the employer to the worker, and workers are more responsible than perhaps they ever have been for their well being--both in terms of their health in general and their financial security during retirement. This shift has been supported, in part, by legislation--namely ERISA, the HMO Act of 1973, the Revenue Act of 1978, and most recently, the Pension Protection Act. This Issue Brief does not pass judgment on this development or address who should bear the responsibilities of preparing workers for retirement or of rationing health care services. The current trend in health care design is toward increased "consumerism." Consumer-driven health is based on the assumption that the combination of greater cost sharing (by workers) and better information about the cost and quality of health care will engage workers to become better health care decision makers. It is hoped that workers will seek important, necessary, high-quality, cost-effective care and services, and become less likely to engage providers and services that are unnecessary and ineffective from either a quality or cost perspective. As employers look ahead toward continually improved plan design, there may be benefits in considering the lessons learned from studying worker behaviors. Specifically, there is evidence about the effects of choice, financial incentives, and information on worker decision making. As a result of research in this area, many retirement plan sponsors have moved toward plan designs and programs that recognize the benefits of well-designed defaults, simplified choices, required active decision making, framing, and commitment to future improvements. With respect to choice, it is now known that more is not always better and may even be worse in some cases. Just as fewer shoppers actually bought a jar of jelly when it was one of 24 as opposed to one of six, evidence has shown that people tend to be less likely to join a company-sponsored retirement plan when more investment options are offered. More choice can also lead to lower satisfaction. It is also known that workers may not be able to appropriately sort through many complex alternatives and that education is not always as effective as employers would hope. Decision complexity often forces people to find a way to simplify, and one of the easiest rules of thumb is to pick the option with the lowest short-term cost, even when that alternative is more costly in the longer run. It is also known that, for good or for bad, choices are constructed on the fly; preferences are dynamic, and logic does not always apply. Financial incentives are helpful in motivating behavior, but they do not affect everyone's decisions. Despite significant financial incentives to participate in 401(k) plans, many workers choose not to. Similarly, despite many of the financial incentives embedded in health care plan design, it can be expected that these incentives will not effectively motivate and engage all workers. One seemingly rational approach to improve workers' decision making is to provide education and guidance to help them sort through complex alternatives and to demonstrate the value of financial incentives. Certainly, providing education and guidance in the form of decision support tools may be an employer's responsibility. However, some studies have shown that, even when "educated" workers have the intent to make improved decisions, they often lack follow-through and fail to take action. In short, education and guidance may not be enough to foster improved health care consumerism. Some employers have begun to design benefit programs with a view toward overcoming behavioral tendencies that negatively affect workers' well-being. Newer retirement plan designs involve careful consideration of default choices. These defaults apply unless workers actively choose a different alternative. Typically, the default attempts to "nudge" workers toward optimal behavior. In the case of 401(k) retirement plan design, more employers are moving toward a default of automatic enrollment in the plan, with automatic investment in a diversified portfolio. Still, additional empirical research and experimentation may be needed to further understand the effects of new retirement plan design features. Future work may also precisely illuminate how the lessons discussed in this Issue Brief may apply to health care plan design that results in improved health-related behaviors. Given the impressive preliminary results in improving retirement planning behaviors, such research and experimentation are likely to be worthwhile.