Measuring Retirement Plan Effectiveness

Transcription

1 T. Rowe Price Measuring Retirement Plan Effectiveness T. Rowe Price Plan Meter helps sponsors assess and improve plan performance Retirement Insights Once considered ancillary to defined benefit (DB) pension plans, defined contribution (DC) plans have become America s primary retirement savings vehicle. In 1979, almost two-thirds of workers participated in defined benefit plans. Thirty years later, in 2009, more than three-quarters (79%) participated in DC plans. 1 One of the outcomes of the shift from pension plans to DC plans is that employees have become more responsible for meeting challenges that once were in the purview of employers who provided DB plans. Today, fewer retirees can expect to receive predictable streams of income provided by their employers through DB plans. Instead, their retirement income largely will depend upon the amounts they ve chosen to save during their lifetimes and the performance of the investments that they have selected. In recognition of these changing circumstances, Congress passed the Pension Protection Act of 2006 (PPA); the most comprehensive legislation to affect the retirement plan industry in decades. It expanded the administrative and plan design options available to retirement plan sponsors, creating a valuable opportunity for DC plan sponsors and other fiduciaries to assess their current plans and consider alternatives. One of the most important components of the PPA was the package of changes designed to expand the use of automatic features in DC plans, such as automatic enrollment, automatic deferral increase, and automatic investment into diversified default investment options including a new safe harbor treatment that may provide nondiscrimination testing relief for plans that incorporate automatic enrollment features and meet certain other design requirements. Clearly, DC plan sponsors and fiduciaries are being encouraged to adopt plan design provisions that help optimize outcomes for employees and improve their retirement preparedness. Since this landmark legislation was passed into law, the number of sponsors embracing automatic provisions has grown and so has the need for new ways to measure the effectiveness of DC plans. Assessing Your Retirement Program s Potential T. Rowe Price developed its proprietary Plan Meter report as an analytical tool to help plan sponsors assess the retirement readiness of their employees and better understand the impact of implementing automated strategies within their plans. 1 Source: Retirement Research Inc. analysis of Form 5500 and Federal Reserve Board Flow of Funds data, May 2010.

2 A changing perspective for DC plan sponsors Traditional DC plan measurements have provided sponsors with plan-level information about specific parts of their plans such as participation rates, deferral rates, and account values without providing a full picture of employees overall retirement preparedness. In fact, in some cases, plan-level measurements have masked issues affecting subgroups within a plan s population. For example, a plan may be meeting its participation rate goal, but have specific population groups that are well below that goal. DC plan sponsors need tools that can help them understand how well the parts of their plan are working together to help employees become better prepared for retirement. This aggregate view is critical and has led to a change in DC plan vernacular. Traditional inquiries about plan metrics have been augmented to include questions such as: Are our employees accumulating enough assets to generate sufficient retirement income? What is an appropriate replacement income target for our retirement plan? What are the appropriate benchmarks for measuring plan effectiveness? How can we quantify the effects of potential changes in plan design? DC plan sponsors are looking beyond the need to help employees accumulate assets; they are focusing on helping employees accumulate enough assets to support the withdrawal amounts needed during a potentially lengthy retirement. As a result, sponsors are establishing different plan goals and looking for new ways to measure performance. Replacement Income: A new standard for measuring DC plan performance The paradigm for measuring retirement plan effectiveness is shifting. Historically, DB plans have been designed to meet specific replacement income goals. It seems reasonable that the new barometer for measuring DC plan effectiveness and employees retirement readiness should be similar. Measuring replacement income from plan assets puts retirement savings into context by focusing on the likelihood that employees projected retirement plan account balances will generate sufficient income to replace a specific percentage of their current paychecks after retirement. what is replacement income? Replacement income is a measure of an employee s long-term financial security. Simply put, it is the stream of money that an employee has the potential to generate from savings and other sources when regular employment comes to an end. When plan sponsors focus on the concept of replacement income, they are looking beyond the need to help employees accumulate assets; they are focusing on helping employees accumulate enough assets to provide an adequate stream of distributions to generate sufficient income during a potentially lengthy retirement. Rather than measuring long-term financial security in terms of a lump sum at retirement, the replacement income model compares employees current income with an estimate of the present value of income that will be replaced with retirement savings. As a result, plan sponsors can assess whether employees are saving adequately to maintain a similar standard of living in retirement. We anticipate that setting replacement income goals will become a standard for retirement plan management. Financial planners suggest that retirees will need to replace, on average, 75% of preretirement income in order to maintain their standard of living after retirement. This amount may include income generated by retirement plans, other savings, pensions, Social Security benefits, and work after retirement. This estimate is based on the idea that some of an employee s expenses will be eliminated once they retire. For example, while employees are working they often are saving up to 15% of their income for retirement (including any company contributions) and paying approximately 8% in FICA taxes. If these expenses end at retirement, a replacement income goal of 75% may be reasonable. As a rule of thumb, financial planners suggest that employees can typically expect to replace roughly 25% of income with Social Security benefits. For employees who will not have pension benefits, or other savings, or where part-time work may neither be reasonable nor desirable, DC plan savings will need to replace approximately 50% of their income. As a result, 50% is often identified as the replacement income goal of many defined contribution retirement programs. Today, however, projections for sponsors who are using Plan Meter suggest that many plans employees are on track to replace significantly less than 50% of their current income. By implementing appropriate plan changes, plan sponsors can help employees improve their replacement income percentages and therefore the lifestyles of plan participants once they retire. 2 T. Rowe Price RETIREMENT insights

3 T. ROWE PRICE PLAN METER Measuring and improving retirement preparedness The Plan Meter report was designed to quantify the ability of plan participants to replace their current income after retirement. Each Plan Meter report is custom-generated for a plan or group of plans (of the same plan sponsor) and is based on the investment options and rules of the sponsor s retirement program. It provides a plan sponsor with information about how a retirement program is currently performing and compares that with the projected outlook assuming the adoption of suggested behaviors that can help employees save more for retirement, including: Employee adoption of age-appropriate hypothetical asset allocation strategies Employee contributions reaching targeted levels, and maximizing catch-up contributions, where applicable Employees delaying retirement for three years Plan-level outcomes are illustrated using a series of meters, like those shown on page 5. These data highlight the amount of income employees can expect to receive in retirement from their assets in the DC plan(s). They also compare current projected income levels with a sponsor s target replacement income. Measuring replacement income Plan Meter measures replacement income by considering the projected stream of distributions from plan assets in current dollars as a percentage of employees current salaries. By focusing on percentages rather than dollar amounts, Plan Meter provides a consistent, plan-level measurement that is relevant to all employees regardless of income level. To create these projections, T. Rowe Price uses a proprietary Monte Carlo simulation methodology. In contrast to tools generating average case outcomes, Monte Carlo analyses produce outcome ranges based on probability thus incorporating future uncertainty. Although the engine cannot predict future investment performance, by simulating 1,000 hypothetical future market scenarios for each retirement planning strategy for each plan participant, it can help plan sponsors realistically assess whether employees are likely to achieve their retirement income goals and what they might do to be more successful. Targeting areas for improvement While the report quantifies how changes to the overall plan could improve the plan s replacement income results, most sponsors are also interested in understanding which segments of their employee populations may be at risk and why. Using the retirement income results of the overall plan as a benchmark, Plan Meter delves deeper within a plan, providing detailed projected replacement income information about key population segments. These populations are defined by demographics (such as age, salary, and tenure) as well as by saving and investing behaviors (such as asset allocation, account balance, and contribution level). This analysis allows sponsors to identify issues with specific plan populations and focus their efforts on areas where action is likely to have the greatest impact. For instance, evaluating projected replacement income by employee age can help plan sponsors determine which age groups currently are on track to reach the plan s replacement income goal and which are not. For example, younger employees tend to have higher replacement income projections because they have longer investment and savings time horizons. Plan Meter reports have identified opportunities for plan sponsors to help specific age groups of employees significantly improve replacement income by adopting systematic features and services that influence employee behavior and encourage them to save more. The keys to success for any retirement program are to save more, use an appropriate asset allocation strategy, invest more conservatively over time, and delay withdrawing savings for as long as possible. The goal is to save as much as possible and keep those savings working for as long as possible. T. Rowe Price offers many possible solutions that reinforce these important messages and may help drive employee behavior. 3

4 Case Study: Building the Case for Retroactive Automatic Enrollment Plan Meter can help sponsors measure the potential of plan changes before they re made. A retirement plan committee was looking to solve the problem of younger employees not taking advantage of the opportunity to participate in the company s retirement plan. The committee was prepared to implement routine automatic services for all new employees, but hadn t considered implementing these features for their current employees. Plan Meter s analysis confirmed that employees in their 20s and 30s about 44% of the total plan population were not currently on track to meet the plan s replacement income goal. In fact, in this plan, the median current replacement income for employees in their 20s was 0% a clear red flag for the committee. Employees in their 30s were only slightly better off with a median current replacement income of 9%, but were still below the plan s replacement income goal of 50%. These data provided a compelling argument for retroactive auto-enrollment. Plan Meter showed that employing auto-increase and auto-enrollment of eligible nonparticipants into an age-appropriate asset allocation strategy had the potential to make a significant difference in replacement income particularly for employees in their 20s. Auto-enrollment: Setting the enrollment default contribution rate at 6% had the potential to increase the median replacement income of employees in their 20s from 0% to 32%. Auto-increase: Automatically increasing participant contributions by 1% in each of the next four years (up to a maximum of 10%) had the potential to further boost median replacement income for those in their 20s from 32% to 44%. Plan Meter helped build the case for the committee to retroactively enroll current employees who were not yet participating in the plan; automatically increase all employees whose deferral rates were lower than the plan s target of 10%; and enhance communications to help all employees understand the benefits of switching to an age-appropriate asset allocation. These changes were projected to have the potential to increase the plan s overall median current replacement income from 7% to 25%, an improvement of more than 250%. The committee decided to implement automatic features retroactively for all eligible employees. Plan Meter analysis can help identify and quantify opportunities to improve replacement income for targeted employee populations. 4 T. Rowe Price RETIREMENT insights

5 Navigating the Plan Meter Report This example reviews some key aspects of a Plan Meter report by illustrating high-level results for a current T. Rowe Price client that used Plan Meter to evaluate its retirement plan against a targeted replacement income goal of 50%. Throughout this paper, case studies delve into further detail. In this example, the four meters reflect current and hypothetical replacement income projections for this plan. The first meter shows the plan s median current projected replacement income percentage. Each of the other three meters displays the projected outcome based on the adoption of various behavioral changes. Current and Projected Income Scenarios % Median Projected Replacement Income: Age-Appropriate Asset Allocation Strategy (AAS) % Median Projected Replacement Income: AAS + Target Contribution Rate + Catch-Up Contributions % 52% 3 5 Median Current Replacement Income Median Projected Replacement Income: AAS + Target Contribution Rate + Catch-Up Contributions + Delayed Retirement Meter 1: Median current replacement income In the example to the left, the first meter shows the plan s median current replacement income percentage, which is 24%. This is well below the plan s targeted goal of 50%. For this report population, the median current total contribution rate was 11%, which included both employer and employee contributions. Meter 2: Projected outlook if employees adopt age-appropriate asset allocation strategies The second meter shows the plan s hypothetical replacement income percentage assuming that employee contribution levels remain constant and that each employee adopts an age-appropriate asset allocation strategy. This measurement can help plan sponsors quantify the potential effect of a default age-appropriate asset allocation strategy on a plan s median replacement income ratio. As you can see in this example, adopting age-appropriate asset allocation strategies moves the projected median replacement income percentage slightly higher from 24% to 26%. Meter 3: Projected outlook if employees adopt age-appropriate asset allocation strategies, contribute at targeted rates, and maximize catch-up contributions The third meter shows how the plan s replacement income percentage might change if employees shifted into age-appropriate asset allocation strategies, contributed at the plan s targeted rate of 20%, and made catch-up contributions when eligible. When these assumptions are combined, the plan s projected replacement income ratio increases to 42% a 75% improvement over the current replacement income level of 24%. This measurement can help plan sponsors quantify the potential effect of combining a default age-appropriate asset allocation strategy with automatic enrollment and an automatic deferral increase on a plan s median replacement income ratio. Meter 4: Projected outlook if employees adopt age-appropriate asset allocation strategies, contribute at targeted rates, maximize catch-up contributions, and delay retirement for three years If, in addition to making the behavioral changes highlighted in meter 3, employees also delay retirement for three years, the plan is projected to surpass its median replacement income goal of 50%. Targeted communications can help employees who are nearing retirement age understand the value of delaying retirement. In this example, delaying retirement for all employees by three years caused median projected replacement income to rise from 42% to 52% a 24% improvement. Highlighted results Another recent Plan Meter report evaluated a large population of employees who had invested over 80% of their assets in stocks. On average, 35% to 40% of their total stock holdings were invested in individual stocks, with the balance in diversified stock funds. Plan Meter s analysis found that this population s median replacement income had the potential to increase from 12% to 18% by reallocating to a diversified, age-appropriate asset allocation an improvement of 50%. Employing age-appropriate asset allocation techniques that allows employees to manage both short-term volatility and longevity risk has the potential to produce higher aggregate replacement income. 5

6 Highlighted results Savings rates really do matter In addition to the way employees savings are invested, research has shown that the amount employees save for retirement has a very important impact on retirement readiness. Financial experts suggest that by setting aside 15% to 20% of income annually considering both employee and employer contributions an employee starting at age 25 should be able to accumulate enough savings to replace at least 50% of his or her current income (adjusted every year for inflation) during a 30-year retirement. When combined with income from other sources personal savings, pensions, Social Security benefits, part-time work, etc. employees saving at this rate have the potential to replace the recommended 75% of current income in retirement. While saving 15% to 20% of income (including employer contributions) may seem daunting to employees, Plan Meter data help to reinforce the correlation between higher savings rates and projected replacement income levels of at least 50% or higher. The results of an analysis based on 167 Plan Meter reports are illustrated in chart 1. For over 98% of the reports, employees with current projected replacement income of 50% or more had median total contribution rates of at least 15% or higher. Plan Meter data further reinforce that employees who save more generally have higher projected replacement income. Based on the same report sample, employees saving 20% or more of their income had median projected replacement income, as a group, that was almost twice that of the general plan population. Median current replacement income across the entire report sample was 18%, while median current replacement income for employees saving 20% or more was 35% almost double the income. As you can see highlighted in chart 2, starting early matters younger employees who save 20% or more of their income have significantly higher median projected replacement income than older employees who are saving at the same rates. While older employees have greater motivation to save more, encouraging high levels of saving among younger employees has the potential to dramatically improve replacement income levels. Targeted communications can help younger employees understand the importance of starting early. Chart 1: Replacement income of 50% or more typically requires contributions of 15% or more Median Total Contribution 30% Results are based on a 167-report sample, analyzing total contribution rates for employees with current projected replacement income of at least 50% or higher. The scatter graph shows the relationship between the median total contribution rate and median age for each of the reports analyzed. These data highlight that with the exception of younger populations, median total contribution rates of at least 15% or higher are typically necessary to generate replacement income of 50% of more. But even the youngest populations, with the median age ranging from mid- to late-20s, had median total contribution rates of at least 13% or higher just two percentage points below the recommended 15%. The median total contribution rate across the full population sample was 18%, and the median age was 34. Chart 2: Starting Early Matters Median Replacement Income 80% Participants with Current Replacement Income > = 50% Median Age Participants with Total Contribution Rates > = 20% Median Age Results are based on the same 167-report sample as chart 1, analyzing median replacement income employees with total contribution rates of at least 20% or more. The scatter graph demonstrates the relationship between median projected replacement income and median age for each of the reports analyzed. These data highlight that employees saving 20% or more tend to be older; however, older populations also trend toward having relatively lower replacement income as compared with younger groups. The youngest populations in this case (median age 33) have a median projected replacement income as a percentage of current salary of 70%, as compared with 23% for the oldest population (median age 63) three times more income. The median replacement income across the full population sample is 35%, and the median age is T. Rowe Price RETIREMENT insights

7 Improving retirement readiness From February 2006 through September 2008, T. Rowe Price introduced Plan Meter reports to more than 140 clients representing more than 160 unique employee populations. Feedback suggests that what sponsors find particularly valuable is Plan Meter s ability to provide a comprehensive analysis of their defined contribution retirement program by simultaneously aggregating multiple measures participation rates, savings rates, asset allocation into a single data point: the replacement income ratio. They also welcome the opportunity to try on alternative plan designs and quantify the potential impact of new features and services before making changes potentially saving the company both time and money. The opportunity for plan sponsors to help employees improve retirement readiness is remarkable. To date, analysis of the same 167 report sample referred to on page 6 shows that plans have the potential to more than double their median current projected replacement income from 18% to 40%. Additionally, approximately one-third of the sponsors receiving Plan Meter reports have implemented changes designed to take advantage of opportunities identified in the reports. These changes have included adding automatic services, selecting investments with age-appropriate asset allocations as default investment options, using targeted communications, and increasing employer matching contribution formulas. Many sponsors have used Plan Meter to validate plan design changes, communicate more effectively with plan decision-makers, and identify key populations that need assistance. It is a tool that provides plan sponsors with the information they need to optimize plan performance. Highlighted Results: Looking at subgroup populations also can help sponsors see the value of services that may not appear to have significant worth when viewed at the plan level. For example, catch-up contributions do not have a significant impact on replacement income for employees in aggregate because most employees don t contribute the amount required to even be eligible to make additional catch-up contributions. However, they can help improve replacement income levels for older employees, particularly highly compensated employees who are more likely to be limited on how much they can contribute due to nondiscrimination testing limits or the 402(g) annual contribution limit. Plan Meter data have shown that catch-up contributions have the potential to help employees age 50 and over realize improvements in projected replacement income of 13% or more. For employees in their 60s, projected improvements have ranged as high as 25%. Case Study: Helping Employees Save 15% to 20% of Income Plan Meter can help sponsors evaluate options for reaching their plan s replacement income goals most effectively. A large firm wanted to identify ways to make its plan more effective. The company s retirement program included a 401(k) plan with a match of 50% on the first 6% (3% for all employees who contribute 6% or more) and a money purchase pension plan with a 4% employer contribution. While the investment committee already had implemented auto-enrollment (at a 3% default rate) and opt-out auto-increase (1% each year up to 6%) services, Plan Meter projections found that a 6% auto-increase cap was too low to reach the company s goal of helping employees replace 50% of their income from plan savings after retirement. As a result of the Plan Meter analysis, the investment committee decided to take a four-pronged approach to improving replacement income levels. First, it raised the plan s automatic deferral increase default from 6% to 10% of salary. Second, the company continued to make a 4% employer contribution to the money purchase pension plan, which pushed employees overall savings rates, including employer contributions, into the range that experts recommend (15% to 20% of income). Third, the company encouraged employees who were age 50 and older to maximize catch-up contributions. Finally, the committee recommended using targeted communications to help participants understand the benefits of delaying retirement for three years. If employees embraced all of the changes, Plan Meter showed that the company s new strategy had the potential to produce a 62% median projected replacement income ratio. The Plan Meter report shows how a retirement program is performing as a whole and how the entire plan population can benefit when a sponsor systematically encourages behaviors that help employees prepare for retirement more effectively. 7

8 Case Study: Modeling and Evaluating Plan Design Changes Plan Meter can help sponsors determine how to improve projected retirement income while controlling costs. A company s management wanted to explore options that could help employees become better prepared for the future without significantly increasing plan-related expenses. Its goal was to improve the plan s 75% participation rate and increase its 6% median employee contribution rate. A key issue was whether the company could afford to retroactively auto-enroll current eligible nonparticipants as well as newly hired employees. The company s benefits committee was charged with identifying cost-contained solutions to help meet management s goals. The committee analyzed a variety of options using Plan Meter. It considered different autoenrollment default rates (ranging from 3% to 6%) and various target contribution rates (ranging from 11% to 20%). These scenarios provided a range of possible outcomes, but one point was clear: The auto-enrollment default rate, while relevant on an individual basis, did not significantly improve projected replacement income at the plan level. However, increasing target contribution rates did. The committee determined that by combining a 3% auto-enrollment default with automatic contribution increases of 1% each year (up to 10%) it could significantly improve overall replacement income projections. When the 5% employer match was included in the calculation, employees target contribution rates could reach 15%. By selecting an auto-enrollment default rate of 3% rather than 5%, the committee would be able to limit near-term matching contribution cost increases. Automatically enrolled employees who did not opt out would receive full matching contributions after three years of plan participation as they gradually increase their contribution rates. By incorporating these changes along with encouraging employees to move to age-appropriate asset allocations, Plan Meter projected that median replacement income had the potential to improve from 13% to 25% more than 90% improvement for the plan in aggregate. And since most current participants already were contributing above the match threshold of 5%, the only direct cost to the company would be the match costs associated with newly enrolled participants. There is no one-size-fits-all solution when it comes to retirement plan design. Plan Meter provides sponsors with an opportunity to try on plan changes and assess their impact on projected retirement income before committing to a course of action. 8 T. Rowe Price RETIREMENT insights

9 Case Study: Using Plan Meter to Effect Change and Validate the Need for QDIAs Plan Meter provides quantitative data that can inform and support plan sponsor decisions. Nearly one-half of the employees in a company s retirement plan had, on average, more than 50% of their assets invested in cash. The company s retirement plan default investment was a money market fund and many automatically enrolled employees had never reallocated their assets. The company s benefits committee was concerned about fiduciary responsibility, as well as employees chances of having adequate resources in retirement. Before the benefits committee could recommend a change to the plan s investment committee, they needed quantitative data that would support their recommendation. Plan Meter analysis provided the supporting data. Median total contribution rates for the plan were relatively high at 13%; however, the Plan Meter report found that employees median current replacement income was only 25%. The report also showed that, for the plan as a whole, moving employees to ageappropriate asset allocations increased projected replacement income from 25% to 29% a modest improvement of 16%. However, Plan Meter s asset allocation analysis highlighted the fact that employees across all age groups who were investing more than 80% of their account balances in cash had the potential to increase their median replacement income from 15% to 23% a 53% improvement simply by adopting an age-appropriate asset allocation. In fact, by adopting this strategy, the employees under age 30 who had more than 80% allocated to money market funds increased their median replacement income from 19% to 38% a 100% improvement. As a result, the committee elected to implement a target-date investment strategy. Plan Meter provides quantitative data that can inform and support design change decisions, as well as facilitate communication between retirement plan decision makers. 9

10 Plan Meter Best Practice Considerations Traditional DC plan measurements have provided sponsors with plan-level information about specific aspects of their plans such as participation rates, deferral rates, and account balances. Customized Plan Meter reports provide sponsors with a full picture of employees overall retirement preparedness by evaluating all aspects of current plan performance and measuring the result in terms of projected replacement income. Instead of making an educated guess about the best course of action for a plan to take, sponsors can make informed decisions by trying on different plan features and services. Plan Meter is a tool that provides fresh insight into the relationship between plan design and replacement income. To maximize the benefits of Plan Meter reports, T. Rowe Price recommends that plan sponsors consider the following: 1. Define target replacement income goals Replacement income is a valuable standard for measuring plan performance because it puts retirement savings into context comparing the amounts employees are likely to be able to withdraw in retirement from their savings to the amount of income they currently earn. With Plan Meter, sponsors and fiduciaries have an effective and consistent means of measuring replacement income. To set appropriate replacement income goals, DC plan sponsors should consider company goals and their overall retirement program. For instance, employers that sponsor DB plans may be able to set lower DC plan replacement income goals because their DC plans will not be the primary source for employees income after retirement. Conversely, employers that do not have DB plans may wish to set replacement income goals of 50% or higher. 2. Identify opportunities to improve plan effectiveness Plan Meter provides detailed information about specific plan populations, shedding light on current saving and investing behaviors. This information can help sponsors identify areas for improvement and take meaningful action designed to facilitate change. Plan Meter is a tool that can help sponsors build retirement plans that specifically meet the needs of their own employees and unique demographics. 3. Share the results with key decision makers Plan Meter reports provide sponsors with the quantitative data they need to effectively analyze and appropriately communicate plan value to management and other decision-makers. The charts and graphs included in the report clearly illustrate the impact of potential changes in plan design. They can be used to facilitate discussions about opportunities for improving employees retirement preparedness. 4. Document fiduciary processes With respect to plan management and administration, the Department of Labor evaluates ERISA compliance by reviewing the process fiduciaries apply when making their decisions. Plan Meter provides an excellent way for company management to document their decision-making process. 5. Revisit annually Plan Meter reports can be created as needed, although T. Rowe Price believes that an annual review, as part of overall strategic planning, is most appropriate. Plan design and administration goals can vary greatly from one company to the next. Each year, plan sponsors should evaluate replacement income goals and results, employees saving and investing behaviors, and plan communications. When needed, they should consider adjustments that will improve overall plan effectiveness. Sponsors also should review their plans whenever business goals change or new compliance and/or legislation is introduced. A Powerful Tool for Plan Sponsors and Fiduciaries DC plans are expected to become many employees primary source for income after retirement. As a result, DC plan sponsors and fiduciaries are being encouraged to adopt features that help optimize outcomes for employees. Rather than engage in trial-and-error attempts to design plans that help employees accumulate enough assets to generate significant retirement income throughout a potential 30-year retirement, plan sponsors can leverage T. Rowe Price Plan Meter. The report helps quantify the impact of potential plan design changes on projected replacement income before the changes have been implemented, leading to fewer potentially costly mistakes particularly in today s environment where every dollar counts. If you would like to learn more about T. Rowe Price Plan Meter or have a Plan Meter report created for your plan, please contact your T. Rowe Price representative by calling Note All results are impacted by demographics, plan design, and specific report assumptions. With that in mind, they should be viewed as general observations, but directionally valid. All references to employees throughout the paper refer to eligible employees. 10 T. Rowe Price RETIREMENT insights

11 RETIREMENT INCOME PROJECTIONS The future is uncertain; therefore, we predict many futures To create our projections and model future uncertainty, we use a proprietary Monte Carlo simulation. Monte Carlo simulation is an analytical tool for modeling future uncertainty. In contrast to deterministic tools (e.g., expected value calculations) that model the average case outcome, Monte Carlo simulation generates ranges of outcomes based on our underlying probability model. Thus, outcomes generated via Monte Carlo simulation incorporate future uncertainty, while deterministic methods do not. Although the engine cannot predict future investment performance, by simulating thousands of hypothetical future market scenarios, it can help plan sponsors to more realistically assess whether their employees are likely to achieve their retirement income goals. Material assumptions The investment results shown in the various Plan Meter charts were developed with Monte Carlo modeling using the following material assumptions, as well as those outlined in the Plan Meter Report Appendix. The underlying long-term expected annual return assumptions for the asset classes indicated in the charts are not historical returns. Rather, these are based on our best estimates for future long-term periods. Our annual return assumptions take into consideration the impact of reinvested dividends and capital gains. We use these expected returns along with assumptions regarding the volatility for each asset class and the intra-asset class correlations to generate a set of simulated, random monthly returns for each asset class over the specified period of time. These monthly returns are then used to generate 1,000 simulated market scenarios. These scenarios represent a spectrum of possible performance for the asset classes being modeled. The success rates are calculated based on these scenarios. We take taxes and required minimum distributions (RMDs) into consideration, as described in the Appendix, but we assume no early withdrawal penalties. Investment expenses in the form of an expense ratio are subtracted from the expected annual return of each asset class. These expenses are intended to represent the average expenses for a typical actively managed, no-load fund within the peer group for each asset class modeled. The analysis does include all of a participant s assets in the defined contribution plan(s), but categorizes them simply as individual stocks, diversified stock funds, bonds, and short-term investments. Other asset classes not considered or modeled may have characteristics similar or superior to those being analyzed. IMPORTANT: The Plan Meter projections or other information generated by a T. Rowe Price investment analysis tool regarding the likelihood of various investment outcomes are hypothetical in nature, do not reflect actual investment results, and are not guarantees of future results. The simulations are based on a number of assumptions. There can be no assurance that the projected or simulated results will be achieved or sustained. The charts present only a range of possible outcomes. Results may vary with each use and over time, and such results may be better or worse than the simulated scenarios. Clients should be aware that the potential for loss (or gain) may be greater than demonstrated in the simulations. The replacement income (in current dollars) is the percentage of the employee s current annual salary withdrawn in the first year of retirement; in each subsequent year, the amounts withdrawn are adjusted to reflect a particular annual rate of inflation. The underlying long-term expected annual return assumptions (gross of fees) used in each of the Monte Carlo simulations are 10% for large-cap individual stocks; 11% for mid-/small-cap individual stocks; 10% for stock funds; 6.5% for intermediate-term, investment-grade bonds; and 4.75% for money market/stable value investments. The following expense ratios are then applied to arrive at net-of-fee expected returns: 0% for individual stocks; 1.211% for stock funds; 0.726% for intermediate-term, investment-grade bonds; and 0.648% for money market/stable value investments. The simulation success rate of each employee s retirement planning strategy is identified for a sponsor s plans in the Rules and Assumptions section of the Plan Meter Report. Simulation success is defined as having at least one dollar remaining in the portfolio at the end of retirement. (The retirement period in the simulations is assumed to end at age 95.) The simulation success rate of a particular retirement strategy is determined by counting the number of simulation scenarios that result in at least one dollar remaining, and dividing this figure by the total number of simulation scenarios of that strategy used. Limitations of the model Material limitations of the investment model include: Extreme market movements may occur more frequently than represented in our model. Some asset classes have relatively limited histories. While future results for all asset classes in the model may materially differ from those assumed in our calculations, the future results for asset classes with limited histories may diverge to a greater extent than the future results of asset classes with longer track records. Market crises can cause asset classes to perform similarly over time, reducing the accuracy of the projected portfolio volatility and returns. The model is based on the long-term behavior of the asset classes and therefore is less reliable for short-term periods. The model assumes that there is no correlation between asset class returns from month to month. This means that the model does not reflect the average periods of bull and bear markets, which can be longer than those modeled. Inflation is assumed to be constant; variations in inflation levels are not reflected in our calculations. These results are not predictions, but they should be viewed as reasonable estimates. 11

12 These materials have been prepared by T. Rowe Price Retirement Plan Services, Inc., for informational purposes only. T. Rowe Price Retirement Plan Services, Inc., its affiliates, and its associates do not provide legal or tax advice. Any tax-related discussion contained in this communication, including any attachments, is not intended or written to be used, and cannot be used, for the purpose of (i) avoiding any tax penalties or (ii) promoting, marketing, or recommending to any other party any transaction or matter addressed herein. Please consult your independent legal counsel and/or professional tax advisor regarding any legal or tax issues raised in this communication. RETIRE WITH CONFIDENCE T. Rowe Price Retirement Plan Services, Inc., is a recognized industry leader dedicated to helping your employees prepare for a more financially secure retirement. With extensive research and development efforts, we anticipate emerging trends and provide innovative solutions that transform participant behavior. With world-class service and award-winning technology and education, we seek to provide participants with the best possible plan experience. In short, our priority is the success of your participants /