Financial Services

Creating better retirement outcomes using data, technology and transparency

August 01, 2018


August 01, 2018

Jake Safane


Jake is an editor for The Economist Intelligence Unit’s thought leadership division in the Americas. Based in New York, Jake’s areas of focus is financial services. Previously, he ran his own content marketing firm, primarily helping startup software firms develop their blogs, social media channels and website copy. Prior to that, Jake worked as a B2B financial journalist covering asset management and asset servicing. He has also covered a variety of other topics as a journalist, ranging from local small business issues to sports. Jake received his bachelor’s degree in Broadcast Journalism from Boston University.

AI can help model future earnings, create budgets, track spending habits, suggest withdrawal amounts and rebalance accounts based on the participant’s income needs.

In the US, 95% of salaried new hires have defined contribution (DC) plans as their only employer-sponsored retirement plan option, according to Willis Towers Watson (1). Other markets around the world are following a similar path. But the societal shift from defined benefit to DC plans has not always been smooth for employees. Thus, there’s a clear impetus for retirement stakeholders to forge plans that work for DC participants. Fortunately, improved data, transparency and the use of new technologies such as artificial intelligence (AI) are paving the way for DC plans that help participants reach their goals.

Traditionally, a participant’s default DC investment plan has been based on simple data points such as age, earnings and expected retirement date. This has led to some participants—particularly those who are not engaged with their plans— having assets in funds that are not optimal for their risk profiles or goals.

Now, enhanced data and technology capabilities and improved transparency enable participants to access greater expertise and have more control and personalization of their investments. For less-engaged participants, technologies like AI can provide more and better information and help take some of the guesswork out of the process. As a result, they can make more informed investment choices.

Plan sponsors can also take advantage of data, transparency and technology to understand trends in investing and participant activity, as well as participants’ goals in their retirement plans. With better information, sponsors can offer more personalized options such as separately managed accounts (SMAs), self-directed accounts (SDAs), customized target-date funds or more rebalancing options. By having more information and clarity on how personalized investment options support retirement goals, the result will be increased DC participation and savings rates and optimized investing, benefiting all stakeholders.


Perhaps the greatest upside of improved data and AI is the potential for greater engagement among participants, particularly those who would take a more active role if only they had the tools to help them. Currently, however, US plan sponsors cite issues such as a lack of employee understanding and interest as some of the top barriers to more effective plans, according to Deloitte (2). 

“From an engagement perspective, members don’t engage,” says Lydia Fearn, head of DC and financial wellbeing at consultancy Redington in the UK. About 92% of the UK’s DC participants are enrolled in the default investment strategy. “The majority feel that it is all dealt with by their employer,” she says.

Jillian Kennedy, a partner in Mercer’s wealth business who is responsible for leading the company’s DC and financial wellness strategy in Canada, echoes this sentiment. She notes that there is a “culture of inertia where many feel they don’t have to do anything with their investments. They believe that the default investment is in their best interest.”

So how can better data, transparency and technology help with engagement? All three can make it easier for participants to better understand their investment options and take more interest in their savings goals by connecting with them on a personal rather than one-size-fits-all level. To improve customer experience, for example, AI tools can engage participants through digital questionnaires about their life and retirement goals—a format that might be preferred by some digitally savvy participants. And this data can be used to create custom investment plans.

Several vendors are exploring AI-enabled voice response technology to scale support on issues like financial literacy and savings. This type of technology can increase a plan sponsor’s ability to interact more with participants. Taking it a step further, AI can guide participants on decisions about their retirement choices similar to how predictive commerce makes suggestions based on previous buying behavior and other customer analytics. With advancements in AI technology, future participants may not even realize they are interacting with a computer.

“Engagement is where we see the most leverage of AI and machine learning trends,” says Ms. Kennedy. “We really think AI will help participants understand how to invest and make their own decisions on investing,” she adds.


As DC participants take more of an interest in their plans and start taking more action through digital tools, they generate more useful data. For example, when members log in to their recordkeeper’s site or app, adjust their enrollment, rebalance their investments, opt out or whatever else they may do while accessing their account, they often currently interact with an advanced data capture system.

“Plan members’ actions are put into the recordkeeping system, and the plan sponsor is receiving data analytics to see how they’re interacting,” Ms. Kennedy says. “This is being introduced to plan sponsors as an enhanced way of reporting and collecting data.”

“Plan sponsors are hungry for more data and more analytics,” she adds. They want to slice and dice that information to develop strategies to manage the plans, set success measures, monitor engagement trends and participation rates, build more predictive models, and better customize their offerings for participants. “That’s also true from an administration perspective, a governance perspective and a communications perspective. Many things are jumping off of this trend,” explains Ms. Kennedy.

Plan sponsors that have strong partnerships with asset managers may be able to then share their findings and develop new offerings for participants—ranging from SMAs to more index fund options, and more environmental, social and governance investment strategies. The data can also be used to personalize investment strategies and marketing, e.g. by providing educational content that aligns with an investor’s financial literacy level.


Lack of transparency is a legacy of many DC plan sponsors and asset managers. For example, 77% of those with investment responsibilities for pension schemes worry about transparency regarding the funds they invest in, according to The Asset Management Exchange (3). Fortunately, the industry is moving towards a more open environment.

Around the world, plan sponsors, asset managers and other retirement stakeholders are providing more transparency around fees, for example. These costs are increasingly broken down to help participants understand how fees affect performance, as well as what various costs are for, such as sales loads vs. management fees.

Plan sponsors and participants may also gain more transparency and control over investment strategies through the use of SMAs and SDAs. In an SMA, the plan sponsor is the direct owner of the assets (as opposed to a mutual fund) and therefore has more discretion and control. In an SDA, participants can choose their own investments rather than having to select from the limited offerings within a traditional DC plan. Data analysis tools might help participants realize their investment goals are not attainable through the current options provided by their plan sponsors, and thus an SDA would provide them with the control they need. Enhanced data and transparency also help participants weigh the potential risks of new account structures and fund types, such as added fees, against the potential benefits of control or ease-of-use.

Within SMAs and other types of accounts, AI can also help advisors and managers make investment decisions that fit the needs of their investors. For example, a 2050 target-date fund may be a suitable default investment for a 35-year-old, but this one-size-fits-all approach may not be the best choice for someone who has a substantial balance already in an individual retirement account or who is planning to take a few years off to start a family. AI tools can help create a customized plan that adjusts investments accordingly, rather than shifting allocations based on a more generic target-date timeline.

Even in instances where investments are not customized, big data can provide more insights that significantly help participants in larger ways. For example, the traditional retirement age of 65 may not be practical for Millennials. In addition, many plan sponsors are allowing participants to keep assets in their plan after retirement, and AI can help model future earnings, create budgets, track spending habits, suggest withdrawal amounts and rebalance accounts based on the participant’s income needs. Big data in areas such as demographic trends and spending rates can help asset managers offer strategies based on more accurate predictions of expected retirement dates and lifespans.

“We’ve gone through this continuum of where plan sponsors focused more on investments using two-dimensional, performance-related criteria that the employees would look at. Now we have more data about behavioral finance and the importance of financial wellness and the importance of engaging employees and smart defaults,” says Ms. Kennedy. “All of those things have driven our market towards the question: ‘What more can we do?’ And that’s why AI is so important.”


As the retirement industry shifts to DC plans worldwide, many plan sponsors will try to help their participants navigate this market and find the best value for their retirement. “If we can be clear about what they are paying for and what they are getting, then we can create a much better and much more trusted DC pension structure than we’ve ever had before,” says Redington’s Ms. Fearn. “Transparency and understanding data are critical for that.”

As AI and machine learning advance, asset managers and plan sponsors will be able to provide improved services to a more engaged market. “Technology is helping to create more efficiencies and hopefully better transparencies, which then in turn should drive better investments,” adds Ms. Fearn. “That’s what we’re trying to get to.”





Enjoy in-depth insights and expert analysis - subscribe to our Perspectives newsletter, delivered every week