Helping Working Americans Achieve a Financially Secure Retirement
How the 401(k) System Is Succeeding

Ayco Summer InnerCircle
Benefits and Compensation Conference

Paul Schott Stevens, President and CEO
Investment Company Institute

July 28, 2011
Saratoga Springs, NY

Download slides (pdf)

Thank you for including me in this conference, which has a great history of bringing together human resources professionals to discuss how their work affects the well-being and financial prospects of America’s workers.

Those of us in the mutual fund industry feel we share that same mission. Our products are designed and managed to help American families realize their financial goals. And we know that the primary financial goal for many Americans is a secure retirement.

I was charged with a few assignments for my talk today.

First, as the title suggests, I am going to speak to the U.S. retirement system and the role of 401(k) plans in American’s retirement planning. We strongly believe that the 401(k) system is a success, so I want to arm you with some facts and perspective to support that point of view.

I was also told that this audience would be interested in insights into investors’ attitudes and actions, particularly in the wake of the financial market crisis and what has now been dubbed the Great Recession.

We have some research on that question that I believe you’ll find thought-provoking, both for retirement savers and for households generally.

ICI and its mutual fund members feel great pride and great responsibility for our role in the retirement system. It starts with the investment services our members provide, with mutual funds managing roughly half of the assets in defined contribution plans and individual retirement accounts. Our funds are used in plans of all sizes as cost-effective means of bringing the markets to retirement savers.

But our industry does more than just manage assets. We’ve enjoyed a healthy and productive partnership with employers, other service providers, and policymakers in building and strengthening products and plans that serve Americans in their quest for retirement security.

At ICI, we also feel that we provide leadership in advocacy for retirement savers and plans. That advocacy is rooted in an extensive research program—as you’ll see here today.

As I mentioned, mutual funds manage about half of the assets in DC plans and IRAs. In DC plans, the mutual fund share at the end of March was 55 percent; in IRAs, mutual funds commanded 47 percent of assets.

By the way, mutual fund’s share of these markets are at or near the all-time high—despite what you might have heard about the rising competition from other 401(k) investment products.

What accounts for mutual funds’ strong share of the retirement market? We believe our funds are uniquely well-suited to serve retirement savers.

The regulatory system that governs mutual funds was forged in the aftermath of our nation’s previous great financial crisis—the Great Depression. For seven decades, that system has put investors first.

When plan participants invest in a mutual fund, they receive a diversified, professionally managed portfolio that’s transparent and liquid.

That fund has a simple capital structure and limited use of leverage—a strong structure that kept funds out of the traps that caught many other financial products in the crisis.

The fund’s assets are held under safe custody, and their value is marked-to-market every day, following strict pricing discipline.

These protections are backed by comprehensive disclosure requirements and rules to manage conflicts of interest. Every fund is subject to strong governance, with independent directors serving as watchdogs for the interests of investors.

Overarching all this is the fiduciary culture of our industry. We must never forget to put shareholders and their interests first.

Mutual funds are also cost-effective. The average expense ratio for stock funds in 401(k) plans in 2010 was 71 basis points—less than three-quarters of 1 percent. In addition, nearly all 401(k) mutual fund assets are in no-load funds—funds that have no sales or redemption charges—or funds that waive their loads for retirement plans. That means that investors don’t face sales charges when they invest new contributions—a crucial feature for the dollar-cost averaging approach that is central to 401(k) investing.

It’s also worth noting that mutual funds serve retirement plans of all sizes. There have been some lawsuits recently premised on the notion that large plans shouldn’t be using mutual funds, and that plan fiduciaries are in error if they select so-called “retail” share classes of funds for their plans. We have vigorously countered those arguments in “friend of the court” briefs and have demonstrated that a wide range of plans with assets exceeding $1 billion offer their participants mutual funds as investment options—with great success.

In addition to providing investment options, mutual fund companies have helped promote and develop 401(k) plans, in partnership with employers.

For example, mutual funds have developed new ways of investing, innovations well-suited to 401(k)s. Today’s retirement savers can take advantage of index funds, in addition to a wide range of active management options. They can use lifestyle funds, which are designed to maintain a certain level of risk—conservative, moderate, or aggressive. They can also invest in target date (or lifecycle) funds that rebalance the fund’s portfolio to be more focused on income and less focused on growth over time as the fund approaches and passes the target date.

Many fund companies also perform recordkeeping for the 401(k) plans—including many that offer open-architecture platforms that include investments outside of their own fund companies.

Fund sponsors have also been pioneers in developing communications and education for 401(k) savers. They provide educational materials on the importance of participating and contributing; investment principles; and online calculators to help people assess whether their saving is on target.

And the mutual fund industry is a strong advocate for the 401(k) system because we believe—based on the research conducted by both ICI and its members—that the 401(k) system has been a tremendous success, helping millions of American workers prepare for retirement.

ICI’s research program on retirement issues is quite extensive and has grown considerably in the last few years.

It all begins with data. Every quarter, ICI compiles the U.S. total retirement assets—figures used by the Fed and other government agencies to size the retirement market.

We have an extensive research effort analyzing 401(k) plan participants’ asset allocation, account balances, and loan activity, through two data sets: A very detailed annual update produced since 1996 in collaboration with the Employee Benefit Research Institute, and a quarterly survey of recordkeepers which ICI launched in the wake of the financial crisis in 2008. These data help us understand individual participants’ behaviors in 401(k) plans.

ICI also studies fees and expenses in 401(k) plans. We’ve looked at total expenses paid by 401(k) participants on their mutual funds since 1996. And in 2008, we first published a study conducted by Deloitte Consulting that surveys plan sponsors to develop an “all-in” fee for all types of investments and all services in 401(k) plans. Deloitte is now analyzing data for the second edition of that study.

We recently decided to bring the same in-depth approach to data on IRA investors by developing The  IRA Investor Database™. We use recordkeeper data to study contributions, rollovers, asset allocation, account balances, and withdrawal activity in IRAs. IRAs are important in their own right, as the largest single component of retirement assets. But the insights we gain are also important to the employer-based retirement system, because much of the money in IRAs originated as rollovers from employer-sponsored plans.

Lastly, our household surveys help us understand how households use and view their mutual funds, DC accounts, and IRAs, as well as their views on risks and investing. We do annual surveys and special studies, such as our two surveys on the distribution decisions that DC participants make at retirement.

Let me now share with you some of what we have learned through our extensive research.

Let’s start with an overall picture of the U.S. retirement system—its size and coverage, its role in household balance sheets, the growing importance of DC plans and IRAs, and some surprising findings on trends in retirement income.

A large majority of American households are saving for retirement, either in employer plans—defined benefit or defined contribution—or IRAs. That 70 percent represents 82 million U.S. households.

As you all know, there’s a lot of concern about Americans’ lack of preparedness for retirement—and there’s a case to be made that many households would benefit from greater saving.

But to keep that in perspective, it’s important to know that at the end of the first quarter of this year, American households had $18.1 trillion earmarked for retirement, an all-time high.

This figure also highlights the growing importance of our contributory retirement savings plans. The combination of 401(k) plans, other DC plans, and IRAs accounted for 53 percent of retirement assets in March—compared to 48 percent at year-end 2000 and just 39 percent in 1990. All told, we see more than $9 trillion in assets in those categories.

This measure of retirement assets does not include Social Security. As you know—and as we’ll see in a second—Social Security is a very important contributor to Americans’ retirement security.

The next couple of slides may come as something of a surprise, because we’ve all heard so much about the decline of pensions.

In fact, as retirement assets have risen as a share of households’ financial wealth, pension income has also risen as a share of retiree income.

This slide shows the components of retiree income since 1975—the year after ERISA was passed. Then and now, Social Security—the blue portion of the bars—accounts for more than half of retirees’ income.

But I want to draw your attention to the pension income, both DB and DC, from the private and government sectors. In 1975, pension income accounted for 19% of retirees’ income. In 2009, pension income has risen to 26% of retirees’ income.

Why? Several factors are at work. ERISA protections strengthened DB plans, shored up funding, and shortened vesting periods. Also, DC plans and IRAs have grown rapidly. In fact, we know that the survey this figure is drawn from tends to undercount IRA and DC withdrawals—so the pension income, and the share of that income from DC plans, could be even bigger.

So—as you’ve all heard many times—“everyone knows” that pension income in the private sector is rapidly diminishing. But in this case, “everyone” is wrong.

In fact, retirement income generated by private-sector retirement plans has become more prevalent—not less prevalent—since the passage of ERISA.

After I sent these slides to Ayco, I looked at this one and realized that we left a word out of the title. The title really should say, “More Retirees Receive More Income from Private-Sector Pensions.”

The bars show you that the median income for retirees from private-sector pensions has risen, in terms of inflation-adjusted dollars, by about one-third since the passage of ERISA, to $6,000 annually.

The figures across the bottom tell the rest of the story: A larger share of the retiree population is receiving income from private-sector pensions. That number is up by half—from 20 percent of retirees in 1975 to almost 31 percent in 2009.

So far from diminishing, private-sector pensions are helping more people, and providing more income, than they were back in what some people consider “the golden age of the gold watch.”

Now, we can’t attribute this trend fully to 401(k)s and DC plans. As I noted before, shorter vesting and better funding for DB plans under ERISA greatly improved the prospects that a private-sector worker who was offered a DB plan would actually collect meaningful retirement benefits from it.

But these figures are based on the same survey as the previous slide—and we know that survey undercounts DC and IRA withdrawals. That fact strongly suggests that the growth of DC plans has actually improved the private-sector pension income picture beyond what we see here.

Clearly, we believe that 401(k) plans are a success. But, as I said, we base our advocacy on research—and so you will expect me to prove that statement.

Part of the proof is in the figures I’ve already shown you:

  • Retirement assets are growing;
  • DC plans and IRAs provide a growing share of those assets;
  • Pension income is increasingly important to retirees; and
  • Private-sector pensions are providing more retirement income than ever before.

Now I’d like to discuss two other ways of gauging 401(k)’s success—the attitudes of plan participants and other Americans, and the evidence that 401(k) and other DC plans will improve Americans’ retirement security.

In the wake of the financial crisis, there was widespread skepticism among media and other commentators about the value of the 401(k) system. In response, we launched a series of annual household surveys to gauge public attitudes.

What our 2010 survey found—consistent with our findings in 2008 and 2009—was that DC plan participants appreciate the features of their plans:

  • Those plans help them think about the long term;
  • Tax savings provide “a big incentive” to save; and
  • Choice and control in investments are important.

Please note the figures above these bars—92 percent, 82 percent, 96 percent. You don’t see majorities like that for most propositions that are tested in surveys—particularly not for a system that has taken some very hard knocks from the press and some policymakers.

Another set of questions we’ve asked since 2008 has become increasingly relevant: what do Americans think about the tax features of DC plans?

As you can see in the top bar here, 88 percent of all households surveyed disagree or strongly disagree with the statement that “The government should take away the tax advantages of DC accounts.”

Even the notion that “The government should reduce the amount that individuals can contribute to DC accounts” meets with strong opposition—82 percent of households disagree.

What’s striking about this slide is the bottom bar in each of these sets. We broke out the responses from households that do not own DC accounts or IRAs. Even those who don’t benefit currently from these tax advantages oppose changing or eliminating them—by margins of three or four to one. In a climate where all tax and revenue ideas are on the table, these are powerful figures.

The ultimate test of trust in the 401(k) system is the confidence that households have that this system will help individuals meet their retirement goals. Consistently since 2008, three-quarters of all households—including those that don’t have such plans—have expressed confidence that DC plans can deliver on those goals.

Let me remind you that we launched this research in the wake of the worst financial crisis since the Great Depression, when commentary on the 401(k) system has been largely negative. All of which makes this confidence in 401(k)s even more remarkable.

Is that confidence warranted?

Working with EBRI, we conducted a detailed modeling exercise to find out. The crucial fact to remember is that the 401(k) system is still very young. Although 401(k)s were launched in 1981, the first cohort of workers who are likely to spend their full career in a 401(k)-based system are those who are around 40 today.

In the EBRI/ICI modeling exercise, we walked this cohort through their full careers. We projected their job changes and their income growth, based on real-life observations on contributions, asset allocations, plan loans, and whether to roll over or withdraw plan assets when changing jobs.

We projected their 401(k) and IRA account balances as they aged. This cohort will turn 65 in the decade between 2030 and 2039. At that point, we calculated how much of their pre-retirement income they could replace based on their account balances. The bottom line: for more than 60 percent of this cohort, their 401(k) accumulations are projected to replace more than half of their salaries. Combined with their Social Security benefits, the median replacement rates for this cohort ranged from 84 percent of pre-retirement income to 106 percent.

We are not the only analysts who have reached this conclusion. James Poterba, Steven Venti, and David Wise conducted a careful study of “the changing landscape of pensions” and concluded that “the advent of personal account saving will increase wealth at retirement for future retirees across the lifetime earnings spectrum.” A strong endorsement from three of the most distinguished scholars in retirement research.

What features account for the success of the 401(k) system?

Start with the roles that employers and service providers, including investment providers, play. Employers, as fiduciaries, and providers all have a stake in 401(k)’s success, and we work together to make the system stronger. As just one example, we provide education and tools to promote the importance of saving and investing.

The basic design of DC plans—with consistent saving, dollar-cost-averaging, and portability—fits the needs of today’s workforce, pairing savings discipline with flexible investment opportunities.

And the 401(k) system is constantly evolving, as sponsors and service providers create new innovations in plan design—as well as the investment innovations I mentioned earlier. Let me just discuss a couple of those innovations—automatic enrollment and default investments.

Automatic enrollment is a clear example of innovation that has benefited millions of American workers. In Vanguard’s data, the share of plans that provide auto enrollment has grown almost five-fold since 2005—the year before enactment of the Pension Protection Act.

The significance of this is clear, because auto-enrollment removes the first barrier to plan participation—signing up for the plan. In Vanguard’s experience, auto-enrollment plans have participation rates of 82 percent of workers. That’s significantly higher than similar plans with voluntary enrollment.

Vanguard’s data also show that large plans are far more likely to have adopted auto enrollment than are plans with fewer than 1,000 participants. Getting these smaller plans to consider automatic enrollment is a desirable goal.

My enthusiasm for automatic enrollment may not match some of the press this feature has gotten recently. You might have read on the front page of the July 7 Wall Street Journal that the Pension Protection Act had reduced retirement savings, based on average contribution rates of plan participants.

This article falls into a mathematical trap. Let’s look at what auto enrollment has really meant.

First, with the spread of auto enrollment, the default contribution rates at which savers are enrolled have increased. The share of plans that start out with 1 or 2 percent contribution rates has fallen, while the share with 3 percent contributions has risen.

Second, automatic annual increases in contribution rates have become far more common—in fact, the share of auto-enrollment plans with auto-increase has more than doubled, to three-quarters of all such plans.

And, as I already mentioned, auto enrollment increases participation rates.

The Journal article presented what you see in the left pair of bars—the average contribution rate among plan participants is lower in auto-enrollment plans than in voluntary plans. As arithmetic, that’s true.

The article totally ignored the fact that more eligible workers are participating—that automatic enrollment turns nonparticipants into savers. That’s how the Journal treats an amazing success—getting people who previously saved nothing for their retirement to start putting away 3 percent of their earnings—as a “failure.”

The right panel shows the full effect of automatic enrollment—the average contribution rate for all eligible employees. It takes into account the fact that some portion of the eligible workforce is contributing nothing—zero.

With the higher participation rate in automatic enrollment plans, there are fewer of those zero-contribution workers. The result: the average contribution rate for all eligible employees is a full percentage point higher in auto-enrollment workplaces than in companies with voluntary enrollment.

And don’t forget—those new 401(k) participants may start saving just 3 percent of their pay, but they’re likely to see their contribution rate increase. Their employer may auto-escalate their contributions. Or, once these employees are in the 401(k) plan, they’re more likely to heed the savings message that their employers is sending and increase contributions on their own.

Put it all together, and automatic enrollment is benefiting millions of Americans—especially those who otherwise wouldn’t be saving at all.

Let’s talk about another innovation—default investment options and the spread of target date funds.

With the growth of automatic enrollment, policymakers and others grew concerned with ensuring that these workers’ contributions were properly invested for the long term. So the Department of Labor adopted regulations that effectively encouraged employers to adopt default investment options that are oriented toward growing a worker’s nest egg—not just preserving it.

The results have been remarkable. The share of auto-enrollment plans that put workers into capital-preservation investments—stable value and money market funds—was decimated in just four years, falling from 40 percent to 4 percent. Those have been replaced with default investment options focused on growth and diversification—target date funds, lifestyle funds, or managed accounts. More than half of auto-enrollment plans are now using target date funds as their default investment option.

Indeed, target date funds are claiming a growing share of the 401(k) market. As a growing majority of plans offer these funds—more than three-quarters in 2009—they’re accessible to more participants, and more workers are embracing them. They’ve grown to be about 10 percent of 401(k) plan assets.

I want you to keep the growing role of target date funds—vehicles that are diversified, with an age-appropriate asset allocation—in mind as we discuss Americans’ reaction to the financial crisis.

The financial crisis of 2007–2009 was the worst since the Great Depression, and so it’s not surprising that Americans have reacted strongly to it. Let’s look at some of the evidence on how it has affected their attitudes, their views about risk, and their actions.

Both ICI and the Federal Reserve Board have long asked households to express their tolerance for financial risk, in terms of whether how much risk they’re willing to take in return for commensurate financial gain. On this slide, we’re looking at willingness to take above-average or substantial financial risk.

Among all U.S. households, that risk tolerance peaked between 1998 and 2001, and fell in the wake of the bear market after the dot-com bubble. It’s down from 23 percent of households to 20 percent.

We also broke out risk tolerance for households that own DC plans or IRAs. As you might expect, because these households are investors, they express more willingness to take above-average or substantial financial risk. But the decline in their risk tolerance since the peak in 1998 has been even sharper—from 33 percent of households then to 25 percent in last year’s survey.

We call this the “worm chart.”

Each line segment, or “worm,” shows how risk tolerance has changed over time for a particular age cohort, identified by their birth years. Each line segment covers a series of surveys from 1989 to 2010. And each cohort is plotted against its average age at the time of each survey.

These “worms” demonstrate two things:

  • First, as you’d expect, older investors tend to be less tolerant of risk.
  • Second, the recent decline in risk tolerance has occurred in almost every age group. You’ll see that in most cases, the right-hand end of the line segment reflects a drop-off as our recent surveys find less risk tolerance.

Now I want to focus your attention on one particular age cohort—what we call the “30-Somethings,” folks born in the 1970s. This is the group that came of investing age in the first decade of this century. What they’ve seen in their investing life is two major bear markets and a decade in which the S&P 500 has been flat.

As you’ll see in that white oval, the 1970s cohort—the blue line segment—had a sharp decline in their willingness to take risk in their 30s. And they are far less willing to take risk than the next-older cohort—those born in the 1960s, the red line segment—was when they were the same age. Just look at the gap between the red and blue lines.

This reluctance to take risk is emerging in the behavior of these 30-somethings as well. The 1970s cohort—the ones whose average age was around 35 in 2010—is much less likely to own equities than the previous cohort when they were that same age, in 2000. We see that in a number of accounts—stock funds in their IRAs and DC plans, stock funds held outside those plans, and holdings of individual stocks or employer stocks.

In fact, the only area in which this ‘70s cohort is more likely to gain equity exposure is in their holdings of hybrid funds in their DC plans or IRAs. As we’ll discuss more in a minute, we see this as evidence that 401(k) plan features are helping this financially shaken age group stay on track in their investing.

There are other ways to show that households are acting on their reduced tolerance for risk. In our household survey at the end of 2010, we asked about specific actions that households may be taking in their investing and retirement planning. We discovered that a majority of households with financial assets—58 percent—had made conservative financial adjustments in their regular savings rate (increasing it), their allocation to stocks (reducing it), or their retirement age (increasing it). Many households made two or all three of these changes.

In my last few slides, let’s look at how the financial crisis affected 401(k) participants. What I will demonstrate is that the features of 401(k) plans—including regular contributions, dollar-cost averaging, employer involvement, and diversified and balanced asset allocations—have served your workers well through this episode.

It will come as no surprise to anyone that 401(k) balances fell in 2008. You might be surprised at how quickly they bounced back. Part of that rebound can be attributed to regular contributions, of course. But it’s also important to note that 401(k)s tend to suppress the sort of bad investor behavior—such as trying to time the market—that can really damage long-term returns.

Just one note on this figure—we’re looking here at what we call “consistent participants,” a longitudinal sample of workers who have had 401(k) accounts at the same employer every year from year-end 2003 through 2009. We focus on this group to screen out the noise—all the workers who take jobs and leave jobs, join plans or roll over their assets from plans—and zero in on how accounts grow over time.

As I said, we launched our quarterly recordkeeper survey at the end of 2008, when the press was full of stories about workers fleeing from their 401(k) plans in panic. We didn’t think that was really happening—and the data have borne out our confidence. In a nutshell, fewer than one in 25 DC plan participants took any withdrawal from their plan, or stopped contributing, in any of the three years in which we’ve gathered this data. Fewer than one in 50 has taken a hardship withdrawal. And fewer than one in seven adjusted asset allocations in 2008—a figure that’s declined as the markets have recovered.

We named our latest report on this data “Commitment to Retirement Security.” I think you can agree that’s what we’re seeing.

At the same time, as I showed earlier, retirement savers have not been immune to the general decline in risk tolerance. As this chart shows, 401(k) participants in most age groups have reduced the extremes in their asset allocations.

If you look at the bottom of each bar, that’s the share of participants in each age group that holds no equities. That’s remained stable—so retirement savers aren’t abandoning stocks.

Where we see change for most age groups is in the top portion of each bar, the green. These are the savers who have allocated 80 percent or more of their accounts to stocks. With the exception of the very youngest workers, fewer workers at any given age are in this extreme group in 2009 than were in 2000. The decline is greatest among investors in their 60s, but we see it even among participants in their thirties. Retirement savers have tended to temper their allocations to stocks.

But they haven’t gone overboard—and, as I’ve indicated a couple of times before, we can credit the features of 401(k)s and the growth of target date funds for that.

This slide focuses on workers who were in their 30s and on their current job less than two years at the time of each survey. As you can see from the top segment of each bar, recently hired 30-somethings have steadily increased their investments in balanced investments, from 8 percent of their total assets for the recent hires in 1998 to 34 percent among the recent hires in 2009. The growing role of target date funds—which accounted for one-quarter of this group’s 401(k) plan assets in 2009—is fuelling that trend.

I’ve thrown a lot of data at you. Ayco has provided you with my slides and with the references we used to put them together. I hope you’ll find them useful and will contact us if you want to pursue any of these issues in our research.

Just to recap, my messages to you today are:

  • 401(k) plans are succeeding, in part because of the productive partnership between employers like yourselves and ICI’s member mutual fund companies.
  • The financial crisis has had a major impact on Americans’ attitudes toward financial risk.
  • But—that crisis has not undermined Americans’ confidence in 401(k)s, nor has it driven 401(k) savers off course.
  • In fact, there’s strong evidence that the features of 401(k) plans are helping Americans avoid over-reaction to the crisis and are countering extremes in investor behavior that might have resulted from the hard times we’ve been through.

With that, I’m happy to devote my remaining time to your questions.