Our retirement system is broken. The average American isn’t saving enough to comfortably retire, and the fault lies in our reliance on defined-contribution (DC) plans, such as 401(k)s. Tinkering with DC plans won’t solve the problem, and the other extreme – a federally mandated guarantee – isn't likely to gain support. But a number of compromises that lie between those approaches offer a better way forward for future generations.
Robert Huebscher weighed in on this dilemma In the September 11, 2012 issue of Advisor Perspectives. His article focused on proposals from Teresa Ghilarducci, a professor at the New School in New York, who advocates scrapping the 401(k) system in favor of guaranteed retirement accounts (GRAs), for which the federal government would guarantee a minimum 3% real return. Huebscher argued that we should instead focus on mandating certain changes to 401(k) plans, such as requiring that all plans offer, as their default option, very low-cost, passively managed target-date funds.
I agree with Huebscher that we shouldn’t abandon 401(k)s, and his recommended changes to that system have merit. But asking workers to shoulder all of the investment risk won't do the job. If we truly want to mend the retirement system, we need another layer of retirement security that falls between Social Security and 401(k)s.
I'll use an example to illustrate the problems with relying on 401(k)s alone, but first let me offer some background on government initiatives to make retirement savings programs work better.
Federal government involvement
In recent years, when the federal government has turned its attention to retirement matters, it has focused on improving 401(k)s. Ghilarducci and some other economists have advocated more extensive changes, but that has not been the government’s priority.
Recently, the federal government has tried to encourage automatic enrollment to increase participation, to require reporting of fees to make costs more transparent, and to make it easier for participants to convert retirement savings into annuities.
Perhaps the biggest challenge has been finding a way to increase participation at the corporate level; only about half the American labor force works for an employer that offers a retirement plan. In response, the Obama administration has proposed the "Auto-IRA," which would require employers who do not currently offer a retirement plan to automatically enroll employees in payroll-deduction IRAs. The administration has been promoting Auto-IRAs since 2009, but the mandate for employers has been a sticking point, and the legislation has languished in Congress with few supporters.
States getting into the act
Even if proposed federal programs get blocked, states may be able to enact employer mandates, and some states are already considering getting into the business of offering retirement savings plans for private sector employees. On September 28, for example, Governor Jerry Brown of California signed into law The California Secure Choice Retirement Savings Program, which is described in this New York Times editorial. This law will be implemented by a two-step process that requires a feasibility study and an additional round of legislative approval.
Similar to the Auto-IRA, the California law requires that employers who do not currently offer a retirement plan implement an automatic payroll-deduction IRA program. Employee contributions of at least 3% will be required, although workers can opt out. By the Times’ count, 11 other states have at least considered similar plans.
But it will not be easy for the states or the federal government to enact the legislation necessary to build a truly robust retirement savings system. For one thing, those proposing legislation have so far been reluctant to require employer contributions. The California program and the Auto-IRA proposal require contributions only from employees, and a 3% minimum contribution is modest. What’s more, any attempts to require that plans offer low-cost funds – especially as the default option – will likely face stiff resistance from the financial industry.
Looking to the UK
Other countries offer a glimpse of what a more robust retirement system might look like. The UK, for instance, just this month put into effect a national policy requiring all employers to automatically enroll employees in a retirement savings program. The program will be phased in over six years, beginning with larger companies, but, once fully in place, it will require contributions of at least 4% from employees, 3% from employers and 1% from the government. Employees will be allowed to opt out.
Employers may establish a program with a private-market provider, if they have not already done so, or they can turn to the newly established National Employment Savings Trust (NEST), which will offer a "public option" of low-cost target-date funds.
This article in the British newspaper Daily Mail describes the program in a detailed but consumer-friendly way.
A 401(k) example
Even if we can build a more robust system for retirement savings, however, I believe we will still fall short of fixing the retirement system.
Consider an individual who works for 35 years, accumulates funds in a 401(k), and, at retirement, purchases an inflation-adjusted immediate annuity. Her objective is to generate enough retirement income to replace 30% of pre-retirement pay. (I assume that Social Security will replace 40%, so the total replacement rate will be 70%.) This chart, which is based on Monte Carlo simulations for various asset allocations, shows the percent-of-pay contributions that would be required during the working years to hit the 30% replacement target for the 401(k). To get to the heart of the matter, the chart shows both the median contribution someone can expect to need and the contribution that would be necessary to ensure, factoring in risk and volatility, that 90% of the Monte Carlo simulations meet or exceed the income replacement goal.
|
401(k) example--Required contribution rates
|
|
Asset Allocation |
Average Return |
Median Contribution |
90% Confidence
Contribution |
0% Stocks, 100% Bonds |
3.20% |
17.0% |
21.8% |
25% Stocks, 75% Bonds |
4.58% |
13.3% |
18.2% |
50% Stocks, 50% Bonds |
5.95% |
10.2% |
16.5% |
75% Stocks, 25% Bonds |
7.33% |
7.6% |
15.8% |
100% Stocks, 0% Bonds |
8.70% |
6.0% |
15.2% |
Source: Author's calculations
This chart illustrates the dilemma that individuals face if they have to rely on savings programs to provide for retirement spending needs. If they want to reduce volatility, they can invest mostly in bonds, but the contribution rates that requires are very high. They can invest more in stocks, in which case the contribution that should achieve their goal will go down, but they will still need to contribute more than 15% of pay if they want to be more or less assured of having enough. For most Americans, contributing at such levels would mean significant sacrifice in the present.
Based on data from Vanguard's annual "How America Saves" report, the median combined contribution rate for participants and employers was 9.6% for 2011, with an average asset allocation to equities of 65%. In my simulations, the median would have been covered, but not the 90% confidence level. A sad irony is that those who can least afford to save are also the least able to adjust spending targets if investments perform poorly. The 401(k) system works quite well for those with high incomes, but not for average Americans.
The above analysis, it’s worth noting, only applies to workers who are covered by a plan and participate. Alicia Munnell and her associates at the Center for Retirement Research estimate that only 42% of workers fit that description.
Relying on the 401(k) system, no matter how much we improve it, will not get us the kind of retirement system we need. We need to build a layer of retirement security that fits between Social Security and 401(k) plans. There are at least three possible options – guarantees, risk sharing, and product innovation.
Guarantees
Ghilarducci has argued for GRAs that would provide at least a 3% minimum real rate-of-return. Huebscher and others have questioned how the federal government could prudently guarantee 3%, with TIPS rates currently negative except for very long durations.
But a study by the Center for Retirement Research, What Does It Cost To Guarantee Returns?, looked at this issue from two different perspectives, leading to different conclusions. Using stock return data going back to 1883, the researchers determined that, for any working career from age 22 to 65 ending between 1925 and 2008 with 100% of savings invested in stocks, a 3% guarantee would never have required any subsidy payments – good news for guarantees and for Ghilarducci.
The study, however, took another view, based on option theory, to demonstrate that, unless the guarantor is less risk-averse than markets on average, there is no feasible way to guarantee more than the risk-free rate, which would be the near-zero TIPS rate in today's interest rate environment. An argument can be made, however, that the federal government should be less risk-averse than the average investor, given that the long time horizon over which it operates permits intergenerational sharing of risk and access to capital markets at attractive rates.
So the picture for guarantees is mixed. Regardless of their merits, however, government guarantees are likely to prove politically toxic in the current environment.
Risk-sharing
Instead of forcing workers to shoulder all investment risk, perhaps there are ways to share the burden among employees, employers and the government. In the private sector, we have gone from employers assuming all the investment risk, with defined-benefit plans, to dumping all the risk on workers, without pausing along the way to consider if both sides could share the risk. But in the public sector – where defined-benefit pension plans still dominate, but states are looking for ways to reduce their risk – a shared approach may be viable. For example, Utah offers a choice for public employees of a DC plan or a hybrid plan where employees are required to contribute only if investment performance is subpar. Some states are also considering plan variations that put both floors and ceilings on credited investment returns. If states successfully implement risk-sharing for public employees, it is at least conceivable that private employers could begin to follow suit.
At the Federal level, Senator Tom Harkin of Iowa has recently introduced a proposal for a hybrid pension plan to be offered to American workers, which may lead to discussions about national approaches that go beyond just improving the 401(k) system.
More-suitable products
401(k) investors might also benefit from having access to products with less risk than standard stocks and bonds. For example, stable value products – portfolios of bonds protected from interest-rate volatility by wrap contracts – can provide stability of principal like money market funds, but with enhanced yields. These products have lost popularity as interest rates have declined, but they could make a comeback if the rate environment changes. Retail products with guarantees, like variable annuities and fixed-index annuities, may find a place in 401(k)s as well, if lower cost versions can be introduced.
Conclusion
When retirement issues come up these days, the focus is usually on promoting DC plans and finding ways to make the DC system work better. DC plans cannot do the job alone, however, and more effort needs to go into examining other options. If we continue to take the easy way out and dump all the risk on employees, merely tinkering with the DC model as we go, we will end up with a broken retirement system. Now’s the time to start down a different path.
Joe Tomlinson is an actuary and a financial planner and is managing director of Tomlinson Financial Planning, LLC in Greenville, Maine. His practice focuses on retirement planning. He also does research and writing on financial planning and investment topics.
Read more articles by Joe Tomlinson
