The Missing Link in Managed Accounts Research

Do managed accounts actually move the needle for retirement savers? Or do they just look good on paper? It's a question the industry has debated for years, and one the Center for Retirement & Policy Studies decided to take on directly with our latest research, Analyzing the Value of Managed Accounts. 

What makes this research different is the model behind it. DCOM—the Defined Contribution Outcomes Model—simulates retirement outcomes using data from millions of participants across thousands of plans, while accounting for how plan design and participant behavior actually interact over time. That last part is harder than it sounds, and it's something earlier research hasn't been able to do well enough to placate naysayers. 

Since the paper came out, we've heard a lot of the same questions. This post is our attempt to answer the most important ones. If you have not read the research yet, we highly recommend you do so. You can find the original publication here.

We believe DCOM is a revolutionary model in the retirement space. It’s a statistical model that simulates how changes in policy, plan design, and investor behavior could affect overall retirement outlooks. But what makes it special is that DCOM focuses on detailed participant behavior within defined-contribution plans—a granularity other simulation models have not been able to achieve before. It incorporates detailed DC plan logic and allows for close analysis of how contribution behavior, asset allocation, plan features, and managed account usage interact over time. 

DCOM was created to address major gaps in existing DC plan research, which often relies on descriptive comparisons that can’t isolate how factors such as participant behavior or plan design affect retirement outcomes. Unlike traditional approaches that treat participant behavior as fixed across a career-span, DCOM models behavior as responsive to plan design, leading to what we consider a more realistic approach. 

To measure the value of managed accounts in a DC plan, for example, DCOM runs two parallel simulations for each participant: 

• A baseline scenario in which the participant remains in a target-date fund or a self-directed portfolio.
• An alternate scenario in which the participant adopts a managed account at a 40-basis-point fee.

For each scenario, DCOM projects the participant’s wealth-to-salary ratio (i.e., how much money a participant has saved by retirement relative to what they earn) at age 65 and calculates the percentage change between the managed account scenario and baseline. This approach isolates the effect of managed accounts, while controlling for participant and plan design factors that have previously made other analyses on managed accounts’ effectiveness harder to interpret. 

Our analysis found that managed accounts boost retirement outcomes across investor types, with self-directed (DIY) investors seeing the largest gains. Managed accounts increase the median wealth-to-salary ratio at age 65 by 5.9% for TDF investors and by 11.4% for DIY investors. Across all plan participants, adopting managed accounts led to an overall increase of 7.7%. 

We also found that the benefits of managed accounts were more pronounced among younger investors and those new to their employer because of their time spent in the service. They have a longer time horizon for the effects of higher savings rates to compound, leading to larger projected differences in wealth at age 65. For the youngest age group, adopting managed accounts led to an increase in median wealth-to-salary ratio at age 65 by 9.9% for TDF investors and more than 22% for DIY investors. Compare that to an increase of 3% and 7.5% respectively in the oldest age group. 

A common critique of managed accounts research has to do with the effect of selection bias. Or, in other words, the idea that more motivated, educated, and engaged investors are more likely to adopt managed accounts in the first place. 

Our model controls for age, wage, tenure, and plan design (i.e., auto-enrollment, auto-escalation, company match, etc.). This helps to isolate behavioral differences rather than differences in who selects a managed account. Keeping this in mind, the largest gains appear among younger, lower-income, and newer participants who are least likely to already be strong savers. If self-selection were driving the results, we’d expect the opposite. 

We won't claim this approach is perfect, but it does meaningfully reduce the bias that has held back earlier studies and gives us a clearer picture of how managed accounts actually influence behavior over time. Plus, this is a topic we plan to explore more in-depth in the near future.

Managed accounts do add value in well-designed plans. And by “well-designed,” we mean plans that include features like auto-enrollment and auto-escalation. To put it simply, automating plan features and adding personalization through a service like managed accounts actually complement each other nicely. 

The benefits of managed accounts were seen across all plan types we studied. There were increases in wealth-to-salary ratios for the following: 

• Plans that use voluntary enrollment: 6.7% for TDF investors, 11.7% for DIY investors
• Plans that use auto-enrollment without escalation: 11.6% for TDF, 18.5% for DIY
• Plans that use auto-enrollment with auto-escalation: 2.7% for TDF, 7.8% for DIY

But the results still indicate that there’s considerable variability across plan types, depending on plan features, as well as participant demographics. Plans with auto-escalation show smaller gains, which is consistent with higher baseline savings rates. Yet even in those plans, managed accounts continue to provide additional lift by personalizing savings recommendations and portfolios. In fact, even among auto-escalated plans, 92% still showed positive projected improvements for participants who switched to managed accounts from TDFs. 

It’s also worth noting that access to managed accounts has increased significantly over the last decade. Today, 45% of plans offer this feature, giving 79% of participants access. Adoption, however, has remained relatively constant for the last five years, with each year indicating that around 7% of eligible participants have enrolled. The gap between managed accounts access and use remains substantial—presenting a meaningful opportunity for more thoughtful plan designs that encourage and incorporate managed accounts usage and participant engagement.

The primary driver of improved outcomes in our analysis is savings behavior—specifically that managed accounts users tend to contribute more and sustain higher contribution rates over time.  

Asset allocation does play a role too, particularly for self-directed investors. But plans have far more ability to influence how much people save than they do to influence what markets return. That's what makes the savings finding significant; it's something plans can actually act on. 

One related question we received was whether a one-time savings increase at enrollment could explain our results. It doesn’t. Savings behavior in DCOM is modeled dynamically over time based on participant characteristics and plan features, not a single moment of change.

A common critique of industry research is that it lacks practical applicability. Broad averages rarely tell sponsors whether a specific plan or population is likely to benefit. 

The 41 plan design prototypes serve as a starting point, not a constraint. When we conduct plan-specific analysis, we incorporate actual plan features rather than applying a generic grouping. In practice, we first assess whether an existing prototype closely matches the sponsor’s plan design. If it does not, we can adapt the framework or create a new prototype that reflects the plan’s specific design. The analysis can include building customized age-wage profiles, savings-rate models, and asset-allocation models that reflect the plan’s participant population and observed behavior. The analysis also uses the plan’s actual investment lineup and fund-level fees. 

Our research also shows that wider adoption of managed accounts could help address the retirement savings gap—though managed accounts alone cannot be a full solution. The shift from generalized conclusions to plan-specific analysis is where the framework becomes most useful.

Personalization is becoming central to how defined contribution plans evolve. As a result, the industry needs better tools to evaluate it and see when and where personalization can drive real value. This research doesn’t promise uniform results across all plans or populations. But it does provide a framework grounded in how participants actually behave, which can bring us closer to understanding where managed accounts can improve retirement outcomes the most. Future research will extend the framework into deeper analysis of how managed accounts complements plan design, the effect of QDIA, and much more.  

If you’re interested in a plan-specific analysis using the framework from our research, please feel free to reach out to our sales team at retirement@morningstar.com.