Thrown in Over Their Heads: Understanding 401(k) Participant Risk Tolerance vs. Risk Capacity

​We believe target-date strategies may expose participants to more risk than they can afford.

Defined contribution lore suggests that participants do not worry too much about risk in target-date strategies, because these participants tend not to move their money, regardless of market volatility. What’s more, we have been told that, in order to meet their goals, participants need to accept the level of risk that’s typically embedded in target-date strategies. However, a careful look at the net flow activity into or out of target-date funds suggests that common DC lore may well be mistaken.

According to the Plan Sponsor Council of America, over 76% of DC plans that automatically enroll participants into their plans use target-date strategies as the default. It’s not surprising then that, as the most prevalent default type, target-date mutual funds tracked by Morningstar as of 30 June 2012, have grown to over $440 billion, increasing from $75 billion since the Pension Protection Act of 2006 was enacted. [Note: Custom target-date strategies would add significantly to this asset value, but unfortunately cannot be tracked as readily.] Given the asset flow and the importance of the default strategy in meeting retirement income goals, we believe a close look at how participants are responding to the target-dates is critical.

Our analysis of this data shows that the closer a target-date vintage comes to the stated retirement date, the more mutual fund net asset flows become correlated to market movements. This may indicate that participant risk tolerance is lower the closer one moves to retirement. Further, when we take a look at the risk of loss imbedded within the market-average glide path, we suggest that the loss potential embedded within this type of glide path exceeds the risk capacity of most workers. Our conclusion is that the market-average glide path may throw participants in over their heads, in terms of both their tolerance and their capacity for accepting risk. At PIMCO, we believe there is a better way to manage target-date assets: If we focus first on the risk capacity relative to meeting an income goal, target-date strategies may be more likely to meet the income needs of workers.

Risk tolerance
To explore risk tolerance, we evaluated the net cash flow (disregarding investment returns) in target-date funds between January 2006 and June 2012, as provided by the Morningstar data universes. We studied five vintages: at-retirement vintage and vintage years 2020, 2030, 2040 and 2050 (see appendix for more information). We considered both the amount of net cash flow and the correlation to the equity market movement, represented by the S&P 500.
When we consider the total flow into all of the target-date vintages, market lore appears to be correct. Flows into all of the funds are positive throughout this period, and net flows appear not to closely correlate to market movement. As shown in Figure 1, net flows into all target-date vintages remained positive throughout the 2006 to 2012 time frame, despite market volatility. Please keep in mind that the net asset flow data may obscure our ability to see the outflow of assets, given the likely significant inflow of assets from automatic contributions.

However, when we look at target-date asset net flows by vintage, we observe meaningful activity that shows a very different picture from the aggregate data. While the aggregate target-date activity is positive, the net cash flow within certain vintages is not. As you can see in Figure 2, while net cash flow in the 2050 vintage remains positive, the “at-retirement” vintage for participants closest to or in retirement is negative at several points.

When we study the correlations between net flows by vintage and by movement in the stock market, using the S&P 500, we observe that the closer the vintage is to the retirement date, the more the net flows are correlated to the market; net flows into the at-retirement target-date funds show the highest correlation to the S&P 500. As you can see in Figure 2, net cash flows into the at-retirement target-date vintage do appear to respond to market movement. During the 2008 market downturn, investors appear to have responded by shifting money out of the at-retirement target-date funds. It may not be at all surprising that money flows into stable value during such downturns, according to the Aon Hewitt 401(k) Index data over time.
Evaluating the correlation of net flows by vintage, we note in Figure 3 that the correlation of flows to the S&P increases the closer participants get to a retirement date. Over the time frame analyzed, from January 2006 to June 2012, flow activity in the 2050 vintage has a correlation of only 0.18 to the S&P 500, whereas the at-retirement example shows a much stronger 0.47. Even more notable, the correlations appear to tighten in periods of market downturns. During the October 2007 to February 2009 time frame, the 2050 vintage showed a negative correlation of -0.20, while the at-retirement vintage tightly moved with the market at a 0.80 correlation. (See the appendix for the correlation table).

Also, we ran a regression analysis of target-date flows in different vintages on S&P 500 performance and found that coefficients are significant for near-term vintages (at-retirement and 2020). Please see detailed information in the appendix.

Our correlation analysis may imply that those closest to retirement have the lowest risk tolerance, and that fear may motivate them to transfer their assets out of the target-date funds. Yet this net cash flow activity may also be explained by the reality that those closest to retirement may be moving out of a DC plan altogether, perhaps simply because of retirement or because they are leaving their current employment for other reasons, e.g., downsizing or job change.

What’s important to take away from this analysis is that, regardless of the reason, this money is moving at an unfortunate time, when markets are down, thus potentially locking in losses. For those who may have moved because of fear, this analysis may show that they lack tolerance to market declines, given the asset allocation within the at-retirement target-date vintage. [Note: You may ask why those farthest from retirement in the 2050 vintage show a negative correlation to the market movement during the financial crisis (October 2007 to February 2009). This may be a result of the contribution inflows outweighing the outflows from these strategies. Further analysis using inflow and outflow data is needed to explain this data point.]

Stay the course
Many professionals in the market argue that participants who moved their assets during times of market stress did so irrationally; if they had just “stayed the course,” they would have recovered and been better off today as a result. This may well be the case, but would the retirement date of these individuals also have been delayed significantly as they waited for their accounts to regain value?

Let’s take a look at a hypothetical pre-retiree who was on track to retire in January 2011. We’ll assume that this 55-year-old participant’s DC account balance of $350,000 was invested in the market-average, at-retirement target-date fund in January 2006, and we’ll consider two scenarios: (1) the individual “stayed the course” and (2) the individual moved out of this strategy to stable value in February 2009.
As shown in Figure 4, when we account for both investment returns and ongoing contributions, it may have taken participants who moved out of the target-date strategy to a stable value strategy 17 months to return to their 2006 starting balance of $350,000 (see appendix for further details). By contrast, if they had stayed the course, it would have taken only a few months to return to the $350,000 balance. At PIMCO, we believe 401(k)s cannot succeed if participants jump out of markets at the bottom and possibly miss a rebound. Plans need to have tolerable downside risk, so participants can ride the market waves. 

To consider what level of risk may be tolerable for participants, we suggest first considering the participant’s risk capacity as measured by the amount of loss workers can absorb at different time horizons without derailing their retirement income goals.

Risk capacity
Unlike risk tolerance, which may be driven by human behavior, risk capacity can be estimated using basic math. Starting with the amount of final pay a DC participant may need to replace in retirement, we then consider what percentage of their salary is expected to be saved during their career and what investment returns may be realized. By evaluating how a participant’s account may grow over time relative to the money needed at the retirement date, we can determine their capacity for absorbing an investment loss at different points during their career (based on the number of years they have to work and save prior to retirement).

As addressed in the PIMCO Viewpoint article titled “Loss Capacity Drives 401(k) Investment Default Evaluation” (May 2012), we believe that, based on DC participant savings patterns and a 50% final income replacement goal, the value at risk (VaR) a participant can accept in a year without derailing their retirement plans (e.g., date of retirement or lifestyle in retirement) is about 10% at 10 years prior to retirement, and less than 7% at retirement. (Note: This analysis assumes a VaR confidence level set at 95%.)  We may define these loss percentages as the capacity the worker has to accept a loss at different time horizons.
The concern is that the market-average glide path builds in a risk of loss at a much higher percentage. As referenced in “Loss Capacity,” we calculate the value at risk 10 years from retirement at less than 17%, and the at-retirement vintage at a potential loss of more than 11%. During the 2008 market downturn, those invested in the market-average at-retirement vintage lost 24% of their market value. During this time frame, we also witnessed the greatest outflows from the at-retirement strategies. While we may attribute these transfers to fear or other factors, what’s critical to recognize is that many participants in those strategies were subjected to a risk of loss that exceeded their capacity for loss. 

Managing target-date glide paths within a set risk capacity
When we consider the appropriate glide path for a defined contribution participant, we believe it’s critical to start with the investment objective: how much of one’s final income will need to be replaced in retirement? Then we consider how much participants are likely to contribute to their DC plans as a percentage of pay during their working years. From this we can determine the amount of loss a participant may be able to accept, given the number of years to retirement. This outcome-based approach sets the risk budget for each target-date vintage. Remaining within this risk budget, we then work to maximize both risk diversification and the potential for return, always with a keen eye toward what the future economic environment may deliver.

Our risk management within the glide path seeks to diversify the drivers of risk within each vintage in such a manner that the participant will experience lower volatility, regardless of the market environment. Further, we add hedging strategies to guard participant assets against market shocks such as the one experienced in 2008. These risk management strategies are important because we recognize the limitation of asset diversification alone in protecting participant assets.

PIMCO’s outcome-focused approach to target-date management acknowledges that participants have but one chance to get it right. We are committed to helping workers reach their retirement income goals. To create successful target-date strategies, we know that we must help workers by building and preserving their purchasing power, by managing volatility throughout their working and retirement years, and by guarding their assets against market shocks that could derail their retirement or their desired lifestyle in retirement.

Thanks to our PIMCO colleagues Fiona Cole and Matt Bartch for their input to the content and their contributions to the analytic work within this article. In addition, we thank Chase Haymon, MBA, 2012 summer intern, for the foundational analysis that he provided for this article.



Morningstar categories
The table shows the labels we use for each target-date vintage and the corresponding Morningstar universes.




It is not possible to invest directly in an unmanaged index.

Starting time: January 2006 

Starting age: 55 years old

Starting balance: $350,000

Employee contribution rate: 12%

Employer match rate: 3.5%

Market-average glide path/stable value: Participants invested in market-average glide path at retirement starting at January 2006, then moved out in February 2009 and invested in stable value.

Market-average glide path: Participants invested in the market-average glide path at retirement starting from January 2006 through June 2012.

Returns for market-average glide path are back-tested and adjusted by inflation rate. Returns for stable value are from the Hueler Stable Value Index and are inflation-adjusted.

End balance
Market-average glide path: $490,864

Market-average glide path/stable value: $371,262

Bounce back to $350,000
Market-average glide path/stable value: from February 2009 takes 17 months to bounce back in July 2010.

Market-average glide path: takes three months to bounce back in May 2009.


Past performance is not a guarantee or a reliable indicator of future results. All investments contain risk and may lose value. No representation is being made that any account, product, or strategy will or is likely to achieve profits, losses, or results similar to those shown. Hypothetical or simulated performance results have several inherent limitations. Unlike an actual performance record, simulated results do not represent actual performance and are generally prepared with the benefit of hindsight. There are frequently sharp differences between simulated performance results and the actual results subsequently achieved by any particular account, product, or strategy. In addition, since trades have not actually been executed, simulated results cannot account for the impact of certain market risks such as lack of liquidity. There are numerous other factors related to the markets in general or the implementation of any specific investment strategy, which cannot be fully accounted for in the preparation of simulated results and all of which can adversely affect actual results.

Stress testing is a simulation technique used on a portfolio to determine its reactions to different hypothetical situations. PIMCO employs methodologies which may include market or other assumptions, subjective judgments and valuation models. Such assumptions, judgments and models may reflect PIMCO’s current thinking and may be changed or modified in response to PIMCO’s perception of market conditions, or otherwise.

The “Market Average” glide path is constructed by MarketGlide. No representation is being made that the structure of the average portfolio or any account will be the same or that similar results will be achieved. Results shown may not be attained and should not be construed as the only possibilities that exist. Different weightings in the asset allocation illustration will produce different results. Actual results will vary and are subject to change with market conditions. There is no guarantee that results will be achieved. No fees or expenses were included in the estimated results and distribution. The scenarios assume a set of assumptions that may, individually or collectively, not develop over time. The analysis reflected in this information is based upon data at time of analysis. Forecasts, estimates, and certain information contained herein are based upon proprietary research and should not be considered as investment advice or a recommendation of any particular security, strategy or investment product. PIMCO routinely reviews, modifies, and adds risk factors to its proprietary models. Due to the dynamic nature of factors affecting markets, there is no guarantee that simulations will capture all relevant risk factors or that the implementation of any resulting solutions will protect against loss. Simulated risk analysis contains inherent limitations and is generally prepared with the benefit of hindsight. Realized losses may be larger than predicted by a given model due to additional factors that cannot be accurately forecasted or incorporated into a model based on historical or assumed data.

Value at Risk (VAR) estimates the risk of loss of an investment or portfolio over a given time period under normal market conditions in terms of a specific percentile threshold of loss (i.e., for a given threshold of X%, under the specific modeling assumptions used, the portfolio will incur a loss in excess of the VAR X percent of the time. Different VAR calculation methodologies may be used. VAR models can help understand what future return or loss profiles might be. However, the effectiveness of a VAR calculation is in fact constrained by its limited assumptions (for example, assumptions may involve, among other things, probability distributions, historical return modeling, factor selection, risk factor correlation, simulation methodologies). It is important that investors understand the nature of these limitations when relying upon VAR analyses.

There is no guarantee that these investment strategies will work under all market conditions or are suitable for all investors and each investor should evaluate their ability to invest long-term, especially during periods of downturn in the market. No representation is being made that any account, product, or strategy will or is likely to achieve profits, losses, or results similar to those shown.

This material contains the opinions of the authors but not necessarily those of PIMCO and such opinions are subject to change without notice. This material has been distributed for informational purposes only and should not be considered as investment advice or a recommendation of any particular security, strategy or investment product. Information contained herein has been obtained from sources believed to be reliable, but not guaranteed. No part of this material may be reproduced in any form, or referred to in any other publication, without express written permission. ©2012, PIMCO.