Managed accounts are due for a transformation.
The existing model, offered in many defined contribution plans, is lopsided with providers benefiting more than plan participants as fees erode much of the value addition for participants. Consequently, it is essential for employers to understand the underlying methodologies of their managed account providers and assess not only the drivers of personalization, but also their utilization by employees. At NEPC, we believe managed account providers can and do construct efficient investment portfolios, but plan providers, as fiduciaries, should push for improved outcomes for their plan participants through negotiating lower fees and seeking to better align the interests of the providers of managed accounts with those of participants.
In this paper, we explore personalized investment management within the context of managed accounts and evaluate its overall impact on and ultimate benefit for participants’ retirement outcomes. To that end, we examined providers’ approaches, tested and modeled a range of personalized portfolios, and assessed fees. Using this analysis, we identify ways the industry can enhance the value of managed accounts for both plan sponsors and participants.
Key Findings
A. Lack of Consensus Among Managed Account Providers on Personalized Portfolio Assignment Approaches
Managed account providers vary significantly in how they determine a participant’s risk allocation, which in turn impacts their returns. Each provider’s unique methodology is influenced by factors such as risk tolerance or retirement readiness. Additional personalization factors—such as external assets or changes to the age of retirement—tend to have a modest impact. Therefore, it is essential for DC plan sponsors to understand the underlying methodologies of their managed account providers and assess what drives personalization, along with how many participants are providing the relevant information.
B. Range of Personalized Portfolio Returns is Narrow
Using a mix of global stocks and bonds, the maximum range of forecasted returns typically falls between 5% and 7%1. This limitation applies to managed account solutions, participants building their own portfolio from the core lineup, and to some extent, target date funds. In essence, the impact of dynamic changes in strategic asset allocation resulting from personalization are likely to be marginal. For example, a shift of less than 10% in the risk allocation may result in minimal changes in expected returns, often within a range of +/- 0.05% to +/- 0.20%.
1 Based on NEPC’s 06/30/2024 Capital Market Assumptions
C. Managed Accounts Show High-Quality Investment Management, but Fees Erode Value
Managed account providers can and do construct efficient investment portfolios; however, current fee levels make it challenging to justify the added value. A fee of 0.30% typically requires a participant to increase their equity exposure by 20-30%, or by two to three target date fund vintages, to achieve a similar net-of-fee return. On the other hand, participants paying a 0.15% managed account fee could anticipate returns comparable to those of a typical target date fund investor2.
2 Typical target date fund investor is defined as an investor in the most used TDF series in the U.S.
D. A Reimagined Managed Account Model
A subscription-based model could align business incentives more closely with improved participant outcomes. By implementing a lower base fee for less engaged participants, providers can offer an entry-level option. Meanwhile, more engaged participants could access additional services and expanded investment options through tiered subscription offerings. This approach allows for flexibility in costs and features, catering to diverse participant needs and engagement levels.
Key Terms
Providers of managed accounts are responsible for implementing a range of investment decisions for participants across multiple glidepaths based on the provider’s strategic asset allocation views.
At its core, managed accounts are a discretionary portfolio management solution selecting investments from the options offered within the DC plan.
TDFs are responsible for providing participants with an optimized investment decision across a glidepath based on the provider’s strategic asset allocation views.
At their heart, TDFs are a closed-source, multi-asset investment that may invest in diversifying asset classes that are not currently offered within the DC plan.
The term “target date” refers to a targeted retirement date and is often included in the fund’s name.
An agreed-upon investment policy that reduces risk once a certain objective is achieved. This is typically done by selling stocks and buying bonds.
When managed accounts, acting in an ERISA 3(21) advisory capacity, recommend that a participant consider changing their savings rate. The key word here is “recommendation,” as the onus to act resides with the participant who ultimately can accept or decline. This type of service is typically a single point-in-time recommendation on savings rate adjustments.
See The Real ROI: Analyzing Savings Advice in Managed Accounts for our research on this topic.
When managed accounts, acting in an ERISA 3(21) advisory capacity, recommend to a participant how much money they can sustainably withdraw from their DC plan at any given point-in-time. Like savings advice, the onus to act resides with the participant to follow through on the recommendation.
This is the action of buying and selling assets, under ERISA in a 3(38) capacity, on behalf of a participant in alignment to a stated investment policy and/or guideline. These investment guidelines are determined by the ERISA 3(38) investment manager operating the managed accounts.
Using a specific set of characteristics or data points to tailor a participant’s asset allocation. Ultimately, these become a set of rules and constraints defined by the managed accounts providers themselves.
What is Personalized Investment Management?
Managed accounts are a discretionary service in which participants delegate control of their retirement investments to a managed account provider. This provider determines how the participant’s assets are invested within the constraints of the plan’s investment design. Participants can also provide personal information to help the provider fine-tune or “personalize” their asset allocation. Traditionally, managed accounts have been touted as an administrative service or decision. We counter this notion while emphasizing the investment aspect of the offering.
Managed accounts are primarily a computer-driven investment and advice solution. The process follows a predetermined set of rules and constraints while reacting to certain data points that may or may not materially influence how the investment manager buys or sells investments. In their simplest form, managed accounts build portfolios from a combination of stocks and bonds, creating an output that ranges from 0% to 100% stocks or bonds. Through small 1% adjustments to stocks, the outputs could equate to about 101 different equity levels. The managed account provider then adjusts a participant’s portfolio over time. These changes may look similar to a target date fund glidepath by incrementally reducing risk over time, or they may follow a different path, depending on the managed account provider’s approach.
Managed accounts providers accept responsibility under ERISA 3(38) as the investment manager. For this reason, we will focus on this element of managed accounts and try to break down the concept of personalization.
How Managed Account Providers “Personalize” Participant Accounts
In an earlier post, we highlighted how managed account providers use personalization factors differently. The research revealed that the portfolio assignment methods used by managed account providers vary and ultimately result in a diverse set of outcomes (Exhibit 1). We polled five different managed account providers and asked them to propose a portfolio for the same sample participant.
All five providers received the same information but used their own unique methodologies to arrive at proposed investment portfolios. Providers A, C and E used retirement readiness as the primary basis for assigning risk and yet arrived at vastly different equity allocations (40% to 81%).
Providers B and D used selected risk tolerance to assign the risk level and arrived at equity allocations ranging from 68%-73%.
As we review the equity allocation outputs from these five providers, we are reminded of the 1991 Brinson, Hood and Brian D. Singer study3, which found that strategic asset allocation could account for 91.5% of return variance among the largest U.S. pension funds over 10 years. This suggests that outcomes for the proposed participant could vary significantly depending on the provider’s starting and “personalized” equity allocation.
3 Determinants of Portfolio Performance (1991) in the Financial Analyst Journal by Brinson, Hood, and Brian D. Singer
Impact of Personalization on Participant’s Asset Allocation
To understand how isolated changes to a participant’s circumstance impact the strategic asset allocation assigned by the managed account provider, we took our baseline participant (as noted above) and included several iterations where we changed one personalized data point (Exhibit 2).
Here are some key observations from two of the providers’ proposed portfolios –
- Provider 1 does not consider a participant’s retirement readiness, even if they are underfunded or overfunded compared to their retirement goal, whereas Provider 2 does not incorporate a participant’s risk tolerance.
- For Provider 1, changes in risk tolerance had the most significant impact, resulting in a 13% increase in risk assets. Altering outside assets and retirement date also affected the allocation to risk assets, but to a lesser extent given the “canceling out” effect. Provider 2’s primary basis for assigning risk is retirement readiness, which had the most material impact on portfolio construction while outside assets and an earlier retirement date had a modest impact.
- Focusing on outside assets further illustrates the difference in methodologies between Provider 1 and Provider 2, as using the same information results in Provider 1 increasing risk assets by 11% and Provider 2 decreasing risk assets by 10%.
- Under Provider 1’s approach, if the participant doesn’t further engage (e.g., updates their risk tolerance, outside assets or retirement date), any future changes to their portfolio will likely follow the normal time decay mechanics of a glidepath. Whereas under Provider 2’s approach, any future changes to their portfolio will likely be most influenced by their projected funded status over time.
While providers use different methods and personalization factors, we observe that a participant’s assigned equity risk allocation is predominantly driven by the provider’s primary basis for assigning risk (e.g., risk tolerance or retirement readiness). Personalization factors like outside assets and retirement age impact allocations to a lesser degree (e.g., 5-10%), when considered. Therefore, understanding each provider’s investment biases is crucial, as strategic asset allocation remains the key factor influencing participant returns.
Personalized Asset Allocation – The Risk and Return Impact4
As managed account providers personalize portfolios, it’s important to understand the impact on a participant’s risk and return profile. Exhibit 5 illustrates changes in risk and return across various portfolios using 10% equity increments, from 100% global equities to 0% global equities. This approach highlights that shifting 10% in equities typically results in an average return change of +/- 0.19% and a risk change of +/- 1.26%.
If you focus on the equity range between 40%-90%, a 10% shift in equity typically results in a +/- 0.13% change in return and a +/- 1.63% change in risk. We focused on this range because it aligns with most allocations observed across managed account providers, and it overlaps with the equity ranges commonly used by top target date fund providers.
We took the portfolios from Exhibit 5 based on their expected risk and return and compared them with vintages of the most used target date fund in Exhibit 6. For context, this target date fund starts with a 90% equity allocation, gradually decreasing to 30% over time. Essentially, a 10% change in equities corresponds to moving between different five-year vintages within a target date fund glidepath, such as transitioning from 2030 to 2025.
Expected portfolio returns, before fees, typically range from 5% to 7% for all portfolios (Exhibit 6). This highlights that managed account providers, despite having a broad equity range from 0% to 100%, are still constrained by forecasted returns. Similar to the Brinson, Hood and Brian D. Singer study mentioned earlier, this limitation is not exclusive to managed accounts– target date funds, do-it-yourself participants, and managed accounts alike face similar restrictions. The key difference impacting forecasted returns lies in access to more efficient asset classes and the fees associated with each option, as the analysis above is gross of managed account fees.
That said, we believe managed account providers can and do construct efficient portfolios. To illustrate this, we modeled a single participant who could elect between a sample managed account provider’s conservative, moderate, and aggressive glidepaths and a target date fund glidepath. Our model assumes a 30-year-old in a DC plan with an initial balance of $7,500, a $50,000 salary, a 3.7% salary growth rate, retirement at age 65, and an income replacement ratio of 45%. Exhibits 7 and 8 display the forecasted retirement balance and longevity for the median and 5th percentile outcomes, showing this managed account provider’s glidepaths outperform the most used target date fund on a gross-of-fee basis.
Balances are in thousands.
This suggests that it’s important to understand the managed account provider’s approach to assigning risk but also the available set of investable assets. In general, the more diversifying asset classes a managed account provider can leverage, the greater the likelihood of constructing more efficient portfolios. Conversely, if a managed account provider has access to a limited set of investment options, while the plan’s target date fund has a broader set of investable assets, it creates a disadvantage for the managed account provider, making it harder to build more efficient portfolios than the target date fund.
4 Expected return and risk based on NEPC’s 30-year capital market assumptions as of 06/30/2024
Impact of Managed Account Fees
To understand the impact of managed account fees, we conducted two different exercises:
- We reviewed fees in relation to expected returns and risk on an efficient frontier, and
- We modeled the impact of fees on participant outcomes
Fee Exercise 1 – A model designed to achieve the highest expected return would allocate 100% to global equity (Exhibit 6), but this approach could reach its limits. When factoring in a 0.30% managed accounts fee, the forecasted net-of-fee expected return for the 100% global equity portfolio decreases from 6.96% to 6.66%, while maintaining an expected return risk of 18.16%. This net-of-fee portfolio would fall well below the efficient frontier, meaning a rational investor would not choose this asset allocation on their own. Instead, a participant targeting a 6.66% expected return would likely opt for a lower-cost alternative with only 70% equity allocation. This still delivers a similar 6.7% return but with a significantly lower expected risk of 13.03%.
Fee Exercise 2 – We modeled different fee levels on a single participant’s forecasted outcomes (Exhibits 7 and 8). For the sake of simplicity, we focused on a participant with a tolerance for moderate risk. We found that the breakeven point relative to a target date fund5 is a managed account fee of roughly 15 basis points. In other words, based on the forecasted target date fund outcome, a managed accounts provider charging 0.15% can deliver comparable results by age 65; for every additional 0.15% in managed account fees, participants would lose about two years of retirement income, based on the median outcomes.
5 Target date fund is defined as the most used TDF series in the U.S.
A Case Study: Range of Experienced Outcomes
To bring this full circle, we use the case of a client where nearly 5,000 participants enrolled with a managed accounts provider. The provider reported that 80% of the participants had personalized portfolios, with an average of 2.5 personalization factors applied.
We then compared the portfolios created by the managed accounts provider to the portfolios plan participants would have been assigned to within a target date fund. We observed the following results, gross-of-fees (Exhibit 9):
- Of participants’ strategic asset allocations, 69% were within +/- 5% of the equity level in the target date fund
- When it came to relative expected returns, 90% differed by no more than 0.50%
- For the expected return variation risk, 61% fell within 1.50%
A close look reveals the observed differences align with our modeled scenarios, supporting the case for managed accounts being able to build efficient portfolios—in some cases, even more efficient than target date funds. In aggregate, the managed accounts increased equity levels for half the participants, improved expected returns of 94%, and increased expected return variation risk for about 74%.
What stands out is that the managed account provider was able to increase expected returns for more participants than it increased expected return variation risk, underscoring how a managed account provider can deliver more efficient portfolios than the most used target date fund.
While this gross-of-fee scenario may seem ideal, the true assessment lies in the net-of-fee results. In this plan, the average participant paid approximately 0.35% for managed accounts. After accounting for this fee, the scenario results change as follows:
- Of participants’ strategic asset allocations, 69% remained within +/- 5% of the target date fund’s equity level
- When it came to relative expected returns, 97% fell within 0.50% of each other
- For the expected return variation risk, 61% stayed within 1.50%
At first glance, these changes may seem minor – portfolio equity levels and expected return variance risk remain largely unaffected. However, the key shift occurred in expected returns. Before accounting for the 0.35% managed account fee, only 5% of participants were forecasted to achieve lower expected returns than their assigned target date fund’s allocation. After including fees, 79% of participants were forecasted to have lower expected returns compared to their age-appropriate target date fund (Exhibit 10). This is troublesome because the decrease in the expected return is not offset by a corresponding reduction in the expected return variation risk.
This case study highlights a scenario where managed accounts should have excelled, driven by high participant engagement and meaningful personalization. Our analysis demonstrates that, before accounting for fees, the managed accounts solution was a more efficient investment option compared to the target date fund for most of the approximately 5,000 participants. However, the fees set by the managed accounts provider, in partnership with the plan’s recordkeeper, eroded the advantage the managed account portfolios held over the target date funds.
One could make the case that a participant isn’t paying for just personalized investment management. There are other services that a managed account provides, including savings and distribution advice, and online tools and progress reports. In our paper The Real ROI: Analyzing Savings Advice in Managed Accounts, we found that savings advice through managed accounts adds value for participants – but that value declines over time. Ultimately, a better, cheaper alternative is implementing an auto escalation feature for all participants. This leaves distribution advice (only applicable to those in distribution status), and online tools and progress reports, which are often already available through the recordkeeper as part of the recordkeeping fee.
The Future for Managed Accounts
We find U.S. DC plans are becoming increasingly passive. This shift is driving more plan assets into publicly traded global stocks and bonds, which supports the simplistic scenarios described in this paper. This growing constraint affects both managed accounts and target date fund providers alike. Over time, as these investment solutions become more commodity like and less differentiated, competition is likely to focus increasingly on price.
In 2024, we began to see a shift in plan sponsor sentiment, with plans actively terminating managed accounts services due to stalled fee negotiations. This change is occurring at a much faster rate than providers had anticipated. At some point, providers will need to rethink their business models to better align the interests of the managed accounts provider, recordkeeper and the end participant. So, the key question is: who will lower their share of the managed account fee – the recordkeeper or the managed account provider? This reduction would need to reflect the value they each provide to the end participant.
To that end, we have already proposed to many providers a buy-up subscription model that better aligns business incentives with improved outcomes for participants. This could look like a single-digit base fee and a series of subscriptions for additional services and/or additional investment exposures for individuals that engage and want those additional features. This model could be used as a plan default investment option, providing an alternative to the most common solution today – the target date fund. This reimagined solution could be revenue neutral for the providers while revamping their offerings to improve participant outcomes.
As with any transformational moment, some market competitors will seize the opportunity to adapt and thrive. We are observing signs that providers recognize this pivotal moment. By adjusting behavior and evolving the relationship between managed account providers and recordkeepers, we anticipate a shift from what has mostly been an oligopoly to a more competitive market where prices are more dynamically set based on the value demanded by buyers of managed accounts.
Conclusion
As we have demonstrated, the current average fee levels erode much of the value participants could gain from personalized investment management. The existing managed accounts model is lopsided, with providers benefiting far more than most participants.
While our research places scrutiny on certain areas of managed accounts, we believe managed accounts can deliver high quality and efficient investment portfolios for participants. Our findings underscore the importance of DC plan sponsors reviewing how asset allocation is determined. The usage, application, and magnitude of personalization factors increasingly hinge on the approach of each provider. Furthermore, most providers seem to rely heavily on a single factor that shapes much of their portfolio construction methodology. Therefore, it is essential to understand the inherent biases that may affect managed account providers.
When considering participant outcomes, the potential range is narrower than one might expect. Forecasted risks and returns limit managed account solutions, do-it-yourself investors, and to some extent, target-date funds. This limitation implies that dynamic shifts in asset allocation resulting from personalization will likely yield only marginal benefits.
As fiduciaries, we bear a responsibility to critically evaluate the services provided, ensuring that participants receive tangible value that justifies the fees charged. We call for a reimagining of the managed accounts business model, advocating for a more balanced fee structure that benefits both participants and providers, while also ensuring effective portfolio management at a reasonable cost.
We need providers who share our commitment to these values and are willing to engage in a constructive discussion to benefit participants and build trust among plan sponsors. This entails a commitment to pursue meaningful change for the betterment of all stakeholders. To learn more and continue this conversation, please reach out to your NEPC consultant.
NEPC CAPITAL MARKET ASSUMPTIONS – ASSET ALLOCATION DISCLOSURES
NEPC’s capital market assumptions are forward-looking and fundamentally based forecasts developed with proprietary valuation models to generate 10-year and 30-year outlooks.
These assumptions are updated on a quarterly basis.
Asset class forecasts for expected returns, volatility, and correlations are based on a combination of forward-looking analysis and historical data.
Forecasts are produced for public market asset classes and alternative strategies with both pre tax and post-tax assumptions.
Historical information dating back to 1926, which includes monthly index returns, cash rates, inflation rates, bond yields, and valuation metrics are utilized to both frame the current economic environment and serve as the foundation for the volatility and correlation assumptions for all asset classes.
Volatility assumptions are based primarily on the long-term history of the asset class with some adjustments for the current environment, while correlation assumptions are based on a mix of both long-term history and current trend.
Expected return forecasts are based on current market prices and forward-looking estimates. The forward-looking estimates rely on a fundamental building blocks approach that broadly includes intermediate and long-term assumptions for economic growth, supply/demand dynamics, inflation, valuation changes, currency markets, forward-looking global yield curves, and credit spreads. The building blocks are specific to each major asset class and represent the primary drivers of future returns.
- For example, the equity forecast model is based upon assumptions for real earnings growth with adjustments incorporated for profit margin changes, inflation, dividend yield, and current valuations trending to long-term averages.
- Fixed income return forecasts are based upon changes in real interest rates and forward yield curves, with credit sectors including an assumption for changes in credit spreads and credit defaults.
- Alternative investment strategies are similarly built from the bottom up with a building blocks approach based upon public market beta exposures while also incorporating an appropriate risk premium for illiquidity.
