How Much Risk Should A 65-Year-Old Take?

Stacy Schaus, Executive Vice President at PIMCO and her colleague, Ying Gao, recently wrote a white paper titled, “Loss Capacity Drives 401(k) Investment Default Evaluation” that tackles some of the most important issues in retirement planning. The white paper discusses the development of asset allocation strategy over an investor’s working years and into retirement. (Schaus has worked on this problem for years and has written a book called Designing Successful Target Date Strategies for Defined Contribution Plans: Putting Participants on the Optimal Glide Path on this topic).

While the white paper covers a number of elements of the life cycle asset allocation problem, I’d like to focus on one in particular: How much can a 65-year-old investor approaching retirement really afford to lose?

In a recent survey by PIMCO, representatives of a number of leading consulting firms that help in designing retirement plans were asked to estimate how large a 12-month loss investors could sustain and still be able to meet their long-term income needs in retirement. A total of 39 firms were surveyed and 26 responded:

18 firms estimated that the acceptable level of loss was 0.0%-5.0%
6 firms answered “don’t know”

Not a single firm indicated that a potential 12-month loss worse than 5% (e.g. return worse than -5%) of the portfolio value was acceptable.

Surprising Results

When I first saw these survey results, I was very surprised. After all, we know that the average 2010 Target Date Fund lost 23% in 2008 and the standard assumption is that an investor in a 2010 fund would be about age 65 in 2010. Furthermore, we have spent a lot of time analyzing risk levels in Target Date Funds, and while some have altered their asset allocations since 2008, our results suggest that 2010 funds—even today—have a far-worse 12-month loss potential than just a 5% loss. Our risk analysis suggests that an investor in the average 2010 Target Date Fund could lose as much as 20% in a 12-month period (e.g. worst case estimated 12-month return of -20%).

When I emailed Schaus about this result, she was kind enough to send me a copy of PIMCO’s own analysis (not yet published at that time) which came to a remarkably similar conclusion. Schaus’ group estimated that the average maximum 12-month loss potential in 2010 Target Date Funds is a return of -22%. Schaus’ group also noted this substantial discrepancy between the consultants’ estimates of how much risk a 65-year-old can bear and the risk levels in 2010 Target Date Funds. This is a major finding from PIMCO that really resonates with our own risk projections here at Folio Investing.

Estimating the “Worst Case Scenario”

Using their proprietary risk model, Schaus’ group created an asset allocation that evolves over an investor’s life to optimally balance investment risk with the need to provide sufficient return to be able to fund retirement.

According to their own risk analysis, the PIMCO asset allocation at retirement age has a worst 1-in-20 (5th percentile) projected loss of -6%, quite close to the consensus view of the consultants of the appropriate loss potential. PIMCO’s projections for the ‘worst case’ loss were the modeled 99.5th percentile loss level (this is the worst estimated 1-in-200 12 month period) and was -12.3%, far less than the average loss levels among 2010 Target Date Funds today. It is also worthwhile to note that the PIMCO analysis assumes that people exit the Target Date strategy at retirement and purchase an annuity—in other words this analysis is a ‘To’ rather than ‘Through’ analysis.

Personalize Your Risk Level with Target Date Folios

In our own Target Date Folios (an alternative to Target Date Funds) we have three risk levels for the 2010 Target Date Folios. The Conservative risk level is the one that is designed for investors who anticipate that they may buy an annuity at retirement rather than staying invested. When we analyzed the loss potential of the 2010 Conservative Folio using the same approach that we applied to the Target Date Funds, we calculated a 12-month loss potential of -7.4%. When we applied the same approach to an asset allocation matching PIMCO’s recommended asset allocation at retirement age, using ETFs to represent each asset class, we came up with a loss potential of -10%.

As Schaus’ analysis notes, potential loss estimates are highly approximate. In real life, portfolios often lose considerably more than models have estimated. My analysis uses a standard statistical measure of two standard deviations below average, which occurs with a probability of 1-in-44. It makes sense, then, that my loss projections lie between PIMCO’s 1-in-20 and 1-in-400 loss estimates. Our numbers are, given the uncertainties associated with risk models, remarkably consistent. This two-standard-deviation loss is the same measure of potential loss that I used when I analyzed the 2010 mutual funds and came up with a loss potential of about -20% cited above.

Managing Risk is Critical in Retirement Planning

So, where does all of this leave us? First, we agree with PIMCO’s conclusions that the current crop of 2010 Target Date Funds could lose as much as about 20% in really bad market conditions. For an investor in a 2010 Fund, (someone in their mid 60s), this could be a catastrophic loss. Second, we agree with PIMCO that the ‘real life’ worst case scenario is likely to be worse than our estimates. When markets go haywire, the range of possible outcomes is considerably more extreme than any possible models estimate.

The consistency of our own risk projections and those in PIMCO’s white paper support the idea that there are reasonable standards for calculating portfolio risk levels and this is very encouraging. Also, in recent comments on the SEC’s proposed standards for Target Date Fund disclosures, Folio Investing proposed that Target Date Funds should provide documentation of their target risk levels using a standardized measure. The broad consistency of our risk estimates with PIMCO’s supports the idea that a risk reporting standard is possible.

Finally, I am left wondering whether investors in 2010 Target Date Funds understand that risk projections suggest that they have the potential to lose as much as 20% in a 12-month period. (Unfortunately, my belief is that very few investors or 401(k) plan sponsors realize that this is the case). While risk management is crucial for all investments, it is especially critical for investors who are approaching retirement and have less opportunities to adjust their long-term plans (such as savings rates) in response to portfolio losses (see “Risk Budgeting: A Critical Tool for Portfolio Management”). Investors approaching retirement need to know how much risk they are willing to stomach and set their asset allocation accordingly.