A retiree’s best strategy for creating a lifetime income stream may be to keep their savings in their defined contribution (DC) plan. Doing so may maintain
access to institutional investment strategies and, if the sponsor allows, a drawdown plan that can provide consistent monthly income. What’s more, for plan
sponsors, retaining retiree accounts can increase assets and bolster leverage to negotiate lower fees. This would benefit all participants – and help
employers attract and retain workers.
While few plan sponsors actively seek to retain retiree assets, most may desire to do so, according to PIMCO’s 2015 Defined Contribution Consulting Support and Trends Survey. To encourage plan sponsors to retain retiree assets, we offered suggestions in
the paper, “Defined Contribution Plan Sponsors Ask Retirees, ‘Why Don’t You Stay?’ Seven Questions for Plan Sponsors.” Most critical to success:
offering distribution flexibility and having retiree appropriate investment choices.
When it comes to retirement income, consultants said target-date strategies (at-retirement vintage) and diversified fixed income funds were the most
important. To that end, we suggest ways that plan sponsors can evaluate how likely these investment strategies are to deliver retirement income.
Evaluating the DC investment lineup for
To generate steady and sustainable retirement income – i.e., a lifetime paycheck – retirees seek opportunity for return as well as the ability to manage
specific risks. Chief among them: market, longevity and inflation risk. Each imposes unique demands on a retiree’s portfolio.
- Managing market risk requires structuring an investment portfolio to seek retirement income with stable payouts and minimal disruption from market
- Managing longevity risk requires assessing how long assets might last.
- Managing inflation risk requires populating a portfolio with assets that seek to keep pace with inflation.
We propose a set of measures to evaluate portfolios on their appropriateness to retirees’ needs:
1. Correlation to retirement liability: How well does the strategy correlate to the PIMCO retirement income cost estimate (PRICE)1?
2. Information ratio relative to retirement liability: What is the information ratio relative to PRICE?
3. Downside risk: What level of Value-at-Risk2 is appropriate given a retiree’s risk capacity?
4. Asset longevity: How long are assets likely to remain during the distribution phase?
5. Inflation beta: To what extent is a retiree’s account likely to keep pace with inflation and retain purchasing power?
In the analysis that follows, we examine target-date (at-retirement vintage) and fixed income strategies. For each, we look at two strategies: for
target-date funds, these include the Market Average3 and the Objective-Aligned Glide Paths (at-retirement vintage); for diversified fixed income
strategies, we model the Barclays Aggregate Bond Index and a multi-sector bespoke blend of major U.S. and non-U.S. bond indexes.
The concept of the PIMCO retirement income cost estimate (PRICE) is discussed in the September 2015 DC Design, “Benchmarking Target-Date Funds to the PRICE
of Retirement.” In short, the retirement cost can be defined as what an individual must pay to buy an annuity that provides lifetime income sufficient to
maintain her lifestyle during retirement. Interest rates will largely determine the cost.
Figure 2 shows the correlation of the four retirement income strategies to the objective of securing retirement income as defined by PRICE. Between the
target-date strategies, the Objective-Aligned approach correlates better than the Market Average Glide Path. Between the diversified fixed income
strategies, the Barclays U.S. Aggregate Index has a slightly higher correlation than the multi-sector bond portfolio.
Next, we modeled how well these four strategies perform relative to PRICE, including excess return, tracking error and information ratio. Between the
target-date strategies, the Objective-Aligned Glide Path (at-retirement vintage) had a higher information ratio than the Market Average Glide Path (at
retirement vintage). For the diversified fixed income strategies, the multi-sector bond portfolio had a higher information ratio.
To measure potential loss, we modeled risk exposure by assessing value-at-risk (VaR) at the 95% confidence level (VaR estimates the minimum expected loss
at a desired level of significance over 12 months). As Figure 4 shows, among the target-date strategies, the Objective-Aligned Glide path has lower
downside risk than the Market Average Glide Path. For the diversified fixed income strategies, the multi-sector bond strategy has lower downside risk.
How long might one’s money last? To answer this critical question, we modeled the distribution phase to evaluate the likely life of the assets. We assume
participants begin retirement at age 65 with an account balance of $680,000 and withdraw 50% or 30% of a final salary of $74,500. Figures 5 and 6
illustrate the projected longevity of assets at 95% and median confidence for final salary withdrawal levels of 50% and 30%.
We find that the Objective-Aligned (at-retirement vintage) glide path has higher asset longevity under the 95% confidence level and the same asset
longevity under the median confidence level than the Market-Average Glide Path (at-retirement vintage) for both 50% and 30% of final salary withdrawal
rates. The multi-sector bond portfolio has the same asset longevity as the Barclays Aggregate U.S. Index under the 95% confidence level and higher asset
longevity at the median confidence level, for both 50% and 30% of final salary withdrawal rates.
As to inflation responsiveness, we believe asset prices are much more sensitive to inflation surprises than actual levels of inflation. Asset classes with
a positive beta to inflation surprises have historically tended to perform well, thus preserving purchasing power.
As Figure 7 shows, for the target-date (at-retirement vintage) strategies, the Objective-Aligned Glide Path had a higher inflation beta than the Market
Average Glide Path. Among the diversified fixed income strategies, the multi-sector bond portfolio had a higher inflation beta.
Retirees may benefit from building a retirement income stream from their DC plan, including access to institutional investments, fiduciary oversight and
attractive pricing. Plans may also benefit by leveraging the purchasing power of retiree assets to bring down the costs of the overall plan. For retirees
to retain assets in plan, though, they need appropriate investment choices: strategies designed to keep pace with the real cost of retirement, reduce the
risk of significant loss and stay ahead of inflation.
According to PIMCO’s 2015 Defined Contribution Consulting Support and Trends Survey, consultants rank “at retirement” target-date vintages and diversified
fixed income strategies as most important. Plan fiduciaries should evaluate these strategies relative to retiree needs. Our analysis shows that an
Objective-Aligned Glide Path (at-retirement vintage) is superior to the Market Average Glide Path (at-retirement vintage) across all measures analyzed in
this article. Specifically, it outperforms in correlation to the cost of retirement (PRICE), risk-adjusted performance, downside risk, asset longevity in
retirement and inflation beta. In addition, our analysis demonstrates that compared to the Barclays U.S. Aggregate Index, a multi-sector bond strategy may
provide higher return, greater asset longevity and higher inflation protection, albeit with slightly higher downside risk.
In addition to the evaluation factors presented, we encourage plan fiduciaries to carefully consider the use of active versus passive investment
management. Consultants in our survey suggested that “active management should dominate” for all asset classes except U.S. large cap equity. Retirees may
benefit from increased return opportunity and reduced risk by selecting actively managed target-date and fixed income strategies. Sponsors also may
consider target-date strategies that embed downside hedging intended to guard retiree assets against a sudden market downturn.
Employers also may consider making an institutionally priced annuity purchase program available to retirees. This could enable retirees to access immediate
and deferred annuities. Retirees may gain greater comfort in managing longevity risk by purchasing a deferred annuity that begins paying income, for
instance, when the retiree reaches age 85. A deferred annuity may allow the retiree to shorten their investment horizon and possibly increase installment
payouts. Combining an objective-aligned target-date strategy or a multi-sector fixed income strategy with a deferred annuity may create a more secure
life-long retirement income stream.
1 PRICE (PIMCO retirement income cost estimate) is calculated as the discounted present value of a 20-year zero-coupon TIPS ladder. Details are discussed
in the September 2015 DC Design, “Benchmarking Target-Date Funds to the PRICE of Retirement.”
2 Value-at-Risk estimates the minimum expected loss at a desired level of significance over 12 months.
3 The Market Average Glide Path is constructed by NextCapital and is an average of the 45 largest target-date strategies in the market.
4 Multi-sector bond consists of: 20% Barclays U.S. Aggregate Index
20% Barclays Global Aggregate ex-USD (USD Hedged) Index
20% Barclays Capital Global Credit Hedged USD Index
20% BofA Merill Lynch Global High Yield BB-B Rated
20% JPMorgan EMBI Global Index
5 We employed a block bootstrap methodology to calculate asset longevity. We start by computing historical factor returns that underlie each asset class
proxy from January 1997 through the present. We then draw a set of 12 monthly returns within the dataset to derive an annual return number. This process is
repeated 25,000 times to have a return series with 25,000 annualized returns. The standard deviation of these annual returns is used to model the
volatility for each factor. We then use the same return series for each factor to compute covariances between factors. Next, the volatility and
distribution of each asset class proxy is calculated as a function of variances and covariances of risk factors that underlie that particular proxy. We
assume participants begin retirement at age 65 with an account balance of $680,000 and withdraw 50% or 30% of a final salary of $74,500. Finally, the asset
longevity for each strategy is estimated using the following return assumptions combined with the above described block bootstrap simulation methodology
and DC assumptions.