In our first article, we explored the implications of behavioral finance associated with joining DC plans. In the second article, we focused on the accumulation phase. In both, we included some measures that trustees and sponsors can establish to counter the impact of a lack of financial understanding among their members. This article focuses on the final phase, retirement, and in particular, on how an understanding of member behavior can help sponsors and trustees structure their DC plans to support participants during the payout, or decumulation, phase of their retirement.

Decumulation: Big decisions that affect the rest of our lives

There is a growing emphasis on decumulation globally, but it remains a complex and evolving area. Participants are likely to be more concerned about a secure income stream in light of the economic crisis. In addition, it is being recognized that longevity brings a more varied spending pattern in retirement² with three distinct phases:

In the earlier articles, the focus of our discussion came down to encouraging members to maximize savings and the potential growth of retirement assets subject to realistic levels of risk. These were common themes across all regions. In retirement, people tend to have similar needs, such as providing an income throughout their and their spouse’s retirement, paying for any additional care²  and wishing to leave a bequest. 

However, varied legislation in different regions means that the way members are able to access their funds to meet their needs can vary substantially. As a result, members may be prone to very different behaviors when making decisions. Fortunately, the options available can be grouped into just three key types, decreasing the complexity of the decision-making process, as follows:

Source: OECD, “Forms of Benefit Payment at Retirement” Pablo Antolin, Colin Pugh and Fiona Stewart, September 2008.

Each of these options has its own merits in relation to meeting members’ needs and has its own corresponding behaviors. This is particularly obvious where more than one option is available simultaneously. Some decisions are irrevocable: For example, once an annuity is purchased it is usually locked in, a fact that affects member behavior; people may be reluctant to buy an annuity because there is a chance that they may regret that decision if their circumstances change. From a purely financial management point of view, if annuities are to be purchased, it is better to spread the purchase over time, but it may not be feasible to do this, and individuals may not have the financial understanding to handle such a spread, even if it were feasible.

Country differences

Policies and practices are under review in many countries, and change is likely. In countries that use lump sums as the primary payout method, there is concern that pension assets are being spent too rapidly. In countries that mandate annuitization, there is concern about the lack of individuals’ control over pension assets, and the possibility of easing that requirement is being discussed. The Organisation for Economic Co-operation and Development is focused on that issue, and a 2008 paper ⁽²⁾  provides an overview of the policy issues and recommendations.

Source: OECD, “Forms of Benefit Payment at Retirement,” Pablo Antolin, Colin Pugh and Fiona Stewart, September 2008.

Decumulation: What we have learned from behavioral finance

In order to consider how member behavior affects decumulation, it is worth reviewing how people make decisions. The timeframe becomes very important when we think about decumulation. The time horizon needs to be the rest of one’s life, but research shows that many retirees suffer from myopia-thinking only about the next five years or even shorter term and not focusing on the long term. At the other end of the retirement spectrum, people have concerns about how much they will need to allow for long-term care costs a large unknown that can puzzle even the best actuaries.

As we have mentioned before, research has shown that members’ decisions are based on a number of factors, including past experience, logical processing and emotion. Emotion is likely to be involved when the individual decides how much control to maintain over funds, and whether a programmed withdrawal will be managed inside the plan or through an independently chosen service. Individuals tend to be overly optimistic about investment outcomes, underestimate their income requirements in retirement and/or underestimate how long they will live.

As people age they become more experienced (or wiser) and have the potential to be better educated through the benefit of hindsight, learning from their previous mistakes and other financial decisions they have made. However, their prior experience may not help as they focus on the drawdown of their funds, as these decisions may be very different from those taken in the past. They may also be influenced at that time by peers who are making or have made similar decisions.

Spend-down and payout decisions are extremely important and will affect the rest of one’s life. Yet the average individual is very poorly prepared to manage a lump sum that must last for years. There is also an increasing chance that as people age, their ability to make appropriate decisions declines.

Given the expected growth of DC funds globally, it is becoming increasingly important for individuals to seek expert help, and quality advice must become more available, so people can  make wise decisions as the value of their own fund becomes increasingly significant for their retirement. Unfortunately, many do not make wise decisions, possibly because they are overconfident of their own financial understanding, are reluctant to pay the necessary charge for this advice or are simply distrustful of the financial service sector. Ultimately, members may make inappropriate decisions that could reduce their income throughout their retirement.

Decisions about investments – The second article in the series linked behavioral finance to investment decisions during the accumulation period. The same issues apply to investments during the payout period, and will be applicable unless an annuity payout is selected. However, these may be further complicated as a result of the need for income along with exposure to alternative products that may compete with the investment options previously available. The total control that members now have over their own funds, and their access to the open market, could exaggerate the impact of their behavior.

Inertia – Inertia is most likely to affect members who draw down on their funds rather than buy annuities. Even if annuity prices fall to a level that would be acceptable, members are unlikely to actively buy an annuity they might need as they progress through retirement. Financial research suggests that there may be a right time (optimal age based on changing mortality risk) for many people to buy an annuity. Therefore, it may be worthwhile for trustees and plan sponsors - where  permissible - to offer free financial advice on retirement to help members assess when this “right time” might be and to adopt a planned process to buy the annuity.

Prospect theory – People view gains and losses and will base their decisions on perceived gains rather than on perceived losses, taking more risk to make up a loss than if they had made an immediate gain. With the recent falls in equity prices worldwide, some members may be prepared to take higher risks now than they would have otherwise, by retaining their investments in equities on the expectation that the market will recover, rather than reducing their exposure to market risks as they approach retirement. The theory of mean reversion of equity markets has yet to be proven, which suggests that this is a risky approach. In the UK, the “lifecycle investment strategy,” which typically switches funds from equities to bonds as a member approaches retirement, has partly reduced the impact, as these funds were hit less by the fall in equity market prices the closer the member was to retirement. This automated process of derisking helps remove the impact of prospect theory.

Framing – The way information is presented can affect both how it is interpreted by members and their subsequent actions. How decumulation decisions are framed can give a positive or negative slant to the same information, and can produce very different results. Pictures of a retiree on a yacht could encourage excessive spending while pictures of an elderly person in poorer circumstances could encourage more cautious spending.

Myopic loss aversion – Members tend to focus more on the short term than the long term. When faced with buying an annuity, members tend to focus on the annuity that provides the largest immediate income rather than the one that may provide the largest income in the long term. One way to counter this behavior is for communications to focus on the longer-term implications in terms of everyday spending, such as what a member’s “basket of goods,” or living accommodations would look like. This could encourage members to take a longer-term view, while linking this to their current cost of living.

Home bias – People naturally favor information or choices with which they are more familiar. This creates challenges during the decumulation phase, because decisions people would make while working could be very different from those needed in retirement. For example, they may retain investments with a favored manager, despite the asset class required, which may be of a very different nature and not within the manager’s expertise. Similarly, members may buy an annuity from a provider with a familiar name or an existing fund manager, rather than shop around for the best price for the annuity. 

Choice overload – Research shows that more choices may not increase satisfaction or the quality of the choice. This applies to decumulation as well as to the accumulation period, because the number of choices or decisions members need to make once they are near, at or in retirement can be considerable and daunting to the financially uneducated. To counter this problem, it would be worth staggering the various decumulation decisions into more manageable chunks to allow members adequate time to think and plan without the confusion of too much choice at one point in time.

Encouraging engagement: What can trustees/sponsors do?

Plan trustees and sponsors need to support payout decisions and also address governance and fiduciary requirements, all of which vary by country. But given the length of time people are expected to live during retirement, greater flexibility and protection are needed to help members plan for and anticipate their changing needs during retirement. Reassurance that they can revise earlier decisions could encourage more members to make their own decisions. Plan trustees and sponsors also need to decide on a basic philosophy with regard to decumulation decisions. As in the accumulation period, the basics require that for good governance during the decumulation period, trustees and sponsors need to pay attention to the best knowledge available and structure plans to meet participants’ needs in light of realities that have emerged. 

Since practices and markets vary so much by geography, feasible strategies and options also vary. What is offered to members also depends on the role that sponsors and trustees intend to take in assisting members. In Mercer’s Global DC Survey 2009, these roles were broadly broken down into Facilitative, Paternal or Compliant, to describe the levels of employer engagement. The exhibit below provides examples of the conceptual options linked to different situations.

Directions for sponsor/trustee action Linked to alternative philosophy (without regard to legal constraints)

Trustee and plan sponsor actions related to decumulation options

Reminder communications and framing before the payout period starts – An important role for the plan trustees and sponsors before the payout starts is to create expectations about decumulation. Is the basic purpose of the plan to enable the retiree to replace his or her salary with an alternative income in retirement?  If so, then providing messages as part of regular statements and communications, showing that the plan is to provide income after retirement, helps lay the groundwork for the payout period. Another important set of messages to lay groundwork for the payout period is that life spans are uncertain; some indication of the range of life spans should be provided as well. Many employees underestimate their life spans. Even if the plan does not provide income directly, it can help employees understand the amount of monthly income or annual spending that can be produced by a given account balance. Of course, having a default distribution method is a powerful motivator.

Timing of retirement and planning for retirement – Understanding how much regular income can be generated by the DC plan is critical. Behavioral finance is linked to success in this effort, because employees are likely to overestimate what they can get from an account balance, how well they can manage it and how long it will last. Employers can play a key role in helping employees plan by helping them estimate whether they have adequate resources. Planning is particularly important where there will be no DB income and where DC funds are vital to retirement security.

Investment options that include built-in income – Some investment options automatically provide for a minimum income during the decumulation period. These options are relatively new, and whether they are available or desirable varies by geography and cost. Their popularity may grow in the future.

Structuring the default distribution system – Unlike the accumulation period, when many plan members will use a default option, during the decumulation period lump sums are usually chosen, if available, in preference to an annuity. However, little has been done to structure DC default distributions that focus on encouraging regular life income. Trustees should communicate how they envisage the default distribution working, in order to enable members to understand the principles and encourage them to apply these principles to their own circumstances. 

Encouraging members to engage by linking their pension savings to their other resources in retirement (for example, social benefits, other pension income, personal savings and investments, house and car). Members can make financial decisions in a similar way. This would also help them recognize how their total financial picture is affected by their decisions on how to use funds in retirement.

Encouraging a long-term view of pension resources – Focus on the rest of life rather than on the short term to reduce the potential of a big drop in resources later in life. A view of what the future could look like, and how retirees would fit in, could be a powerful tool for more conservative spending.

Limiting the number and nature of choices – Where choices are offered, the choices should be sufficiently different so that plan members can easily differentiate between them, and they should be presented in a systematic and timely way. Staggering decisions over an extended period should help members make more efficient and effective decisions.

Beyond structuring of choices: Providing the right messages

Behavioral finance gives us clues about how to structure retirement plans and it also gives us clues about how to structure information that will be effective. As we indicated in the last two articles, communications need to set up the context well and focus on the purpose of the program, linking it to the interests of plan members at different stages of their lives not just their working lifetimes. This would ensure that members relate the information to their own personal circumstances and envisage what these would be in retirement. Relevant information at the appropriate time can reduce the impact of poor choices.

Communication issues to be addressed include the following:

What to include on regular statements – What will be useful to whom? Customized information to the specific membership is very important and can encourage progressive understanding. Are amounts communicated as income rather than as lump sums? If income, will it be an annuity or a programmed withdrawal? Income is probably easier for a member to relate to. The lump sum required for a long and healthy retirement is probably beyond most members’ comprehension. In some countries, DC benefits are paid primarily as income; in others they are paid as lump sums, and in some countries, participants have a choice. What is communicated about how assets are building up and how likely they are to provide adequate retirement benefits can help people focus on their future decisions.

How to explain options at retirement – At the time of retirement, there are often major choices to be made. The default in many members’ eyes will act as a natural anchor upon which they may frame their own decisions. These are often the biggest financial choices of a lifetime.

Understanding that there are tradeoffs – In virtually any plan that offers choices (or a cash lump sum) at the time of termination or retirement, there are important tradeoffs to be made.  There is often a fundamental choice between “guaranteed lifetime payments versus flexibility” and “liquidity versus the potential for a bequest.”  There is a tradeoff between the opportunity for gain or loss through self-investment versus turning the task over to someone else. There are decisions that are locked in and some that are changeable later. There are decisions that provide a potential for fixed returns versus higher or lower returns through investment choices. Some choices may have better tax treatment than others. The specific tradeoffs depend on the particular plan, the marketplace for financial products in the country, whether the plan sponsor offers special products in the workplace and the tax rules in the country. The plan sponsor can help the individual understand the applicable tradeoffs and prepare for them, so that he or she can make decisions appropriate for his or her own circumstances.

Education and evaluation are key

Arguably the most important stage of a DC plan’s lifecycle is retirement, where the efforts and sacrifices of plan members will be rewarded. However, with it comes some of the most important decisions members need to make. The nature of modern DC plans is that they offer choice and transfer responsibility to the individual. Plan members need to remember that decisions they make throughout their lives will define their retirement security. Choices at payout are vital for their financial security in retirement.

For plan sponsors, there is a tremendous opportunity to assist employees rather than leaving them to their own devices: through plan design, support for retirement planning and the creation of the appropriate context for employee decision making.

This article has built on the earlier articles focusing on joining the plan and the accumulation period. Collectively, the three articles describe the key stages of the development of a DC plan. If members can be educated progressively throughout their working lifetimes about the growth and financial importance of their fund,  there is a realistic expectation that they will be suitably prepared for the critical decisions required to make along the way, and ultimately that they will be financially secure in retirement.

With thanks to the following academics, whose research in the field of behavioral finance and whose articles and papers have been utilized in the views set out in this article: Shlomo Benartzi, Gur Huberman, Sheena Iyengar, Wei Jiang, Daniel Kahneman, Olivia Mitchell, Hersh Shefrin, Robert Shiller, Richard Thaler, Amos Tversky and Stephen Utkus

Notes:

1. Programmed withdrawals are also called phased withdrawals or allocated pensions (for example,. in Australia).

2. Antolin, P. “Policy Options for the Payout Phase,” OECD Working Papers on Pensions and Private Insurance, Number 25, OECD.