<!-- Swiftype Variables -->

Encouraging change

Retirement income: New strategies should be welcomed

Mark J. Warshawsky is director of retirement research for Towers Watson & Co.

After an extensive request for information by the departments of Treasury and Labor and a two-day hearing last year, the Obama administration is now beginning to propose regulatory changes affecting retirement plans and accounts to encourage the use of lifetime income solutions for retirees.

The Treasury is proposing changes in minimum distribution and other rules; Labor is discussing annual disclosures of the retirement account value converted to income flow form.

This initiative is being taken as members of the baby-boom generation heading into retirement face a somewhat poorer and less structured benefit environment than past generations. Today, it is left to retirees themselves to maximize retirement income and minimize risk from accounts. Several products and strategies have been put forward to help retirees. I am suggesting a new strategy.

At this time of development and experimentation, it is critical that the federal government not pick winners, but stay focused on the broader goals of removing obstacles and encouraging lifelong income flows from retirement accounts.

Given the competing desires for income and wealth, for security and flexibility, and for current and future needs, it is appropriate to compare carefully various retirement distribution strategies and products to evaluate their properties and likely outcomes.

Let's examine six different strategies applied to the same initial account balance at a retirement age of 65 for an individual:

  1. constant percentage annual withdrawals from the retirement account balance;
  2. purchase with the entire balance a straight immediate fixed-life annuity;
  3. purchase with the entire balance a straight immediate variable-life annuity;
  4. purchase with the account balance a deferred variable annuity with a guaranteed minimum withdrawal benefit rider;
  5. a mix of constant withdrawals from the balance and a one-time purchase of a fixed annuity, split at the time of retirement at 70% balance and 30% annuity; and
  6. a mix of constant withdrawals from balances and laddered purchases of immediate fixed annuities, gradually moving to complete annuitization by age 75.

The relatively new innovation in deferred annuities of the guaranteed minimum withdrawal benefit, or GMWB, rider protects the retiree against income declines and offers the possibility of market improvements in fixed-percentage (generally 5% annually) withdrawal amounts. While the minimum income flow is guaranteed by the insurer, the account balance is not and indeed might go down, even to zero, after distributions and charges for the GMWB rider on top of the investment and variable annuity charges. As long as there is an account balance, this product provides liquidity to the investor, and there is also no timing risk to the income flow from price fluctuations for life annuities. (I did not consider longevity insurance because its relative pricing efficiency was unknown, even as it has a long tail of fiduciary/solvency risk.)

The comparison is done through a stochastic analysis. Various fee charges are included; annuity pricing is modeled as well. Investments are made to a mix of equity and bonds, either in the account balance or the variable annuity. The model has stochastic equity and bond investment returns, interest rates, inflation and mortality.

The results are stated in terms of probabilistic outcomes. For real balances, the first strategy produces the highest levels in almost all scenarios. The fourth and fifth strategies produce the next highest levels of real balances, although the fifth strategy does better in tougher investment and inflation environments. The second and third strategies have zero balances, by design. In terms of real income, however, the first strategy produces a relatively low level while the second and third strategies produce the highest levels, especially noticeable at younger ages in retirement. The main advantages of the third strategy are that it eliminates timing risk in the annuity purchase, while giving some upside potential with risk that income will fall short. The sixth strategy produces the highest levels of real income at later ages and in tough investment and inflation environments, and averages out the annuity price fluctuations over time. The fourth strategy, despite its guarantees, is not noticeably superior to the first strategy in terms of income across the scenarios, owing to the force of extra charges and inflation erosion, while the fifth strategy is slightly stronger overall. Hence, the evidence points to the fifth and sixth strategies serving well as the basis for income and wealth management for the average retired household. The fourth strategy does have, however, the advantage of being a complete package to the retiree.

Consistent with the goals of the minimum distribution requirements and retirement income disclosures, one would posit preferences that are dominated by concerns about income security risks (and only secondarily about real balances). Then the model shows that a good strategy for an individual retiree at age 65 is to start with moving 10% or 15% of the retirement portfolio to an immediate life annuity initially and thereafter to gradually convert the rest of the remaining portfolio to immediate life annuities over the next 20 years. Withdrawals from the un-annuitized portfolio should be 5% per year. The equity share of the remaining asset portfolio, while mainly determined by the risk aversion of the retiree, should increase with the degree of annuitization.

Life annuities are clearly valuable parts of income strategies from a retirement account, but should not be mandated because retirees in poor health or with other sources of income might be disadvantaged. Similarly, whether the strategies are offered inside or out of a retirement plan, or just on a retail basis, should be the sponsor's choice, depending on relative costs, any pricing inefficiencies (with or without reflection of gender differences), administrative burdens and fiduciary risk, and the methods available to manage and minimize them.

Mark J. Warshawsky is director of retirement research for Towers Watson & Co. in Arlington, Va. His commentary is adopted from “Retirement Income: Risks and Strategies,” MIT Press, Cambridge, Mass.