In a world increasingly moving to defined contribution pension plans, an effective default fund is key since many workers won’t make active decisions about the fund that’s right for them. But default funds can take many forms and some may be more effective than others.
A new paper suggests that a default fund investment model that considers both time until retirement and wealth accumulated at various points during the accumulation phase has a higher probability of meeting a plan member’s retirement goals than a default that only considers age and gender.
The option can be implemented by any DC plan with enough data to estimate asset and return correlations, the paper said. “To build these policies, we formulate a stochastic dynamic programming model that uses different objective functions to take risk aversion, the performance of the default policy and a target terminal wealth into account.”
The researchers tested their model using data from Chile’s public pension system, which is one of the oldest pure DC systems in the world and has a lot of available data, including real returns, for testing.
Chile has five funds, each defined by their equity exposure and ranging along the risk spectrum. “Workers are free to choose in which funds they invest and can modify their investments freely,” the paper said. “A default policy exists to assign an investment strategy for workers who do not choose in which fund(s) to invest. The default policy in Chile . . . follows the lifecycle logic and depends only on the time to retirement.”
The paper described Chile’s default policy as open-looped, which means it doesn’t depend on the wealth accumulated at any point in time and is purely based on age and gender. It noted the fund performs well and about 70 per cent of the country’s population is defaulted into it.
The research used Chile’s policy as a benchmark to test its closed-loop policy, which considers both time to retirement and wealth accumulation. It found changes to investment strategies relative to the individual’s wealth level can reduce shortfalls relative to retirement targets.
Particularly, when testing the closed loop-policy, researchers looked to beat a 66 per cent replacement rate for workers. “All the models discussed obtained a significant reduction in downside risk and almost all resulted in expected returns that were superior to those of the default policy,” the paper said. “Additionally, the probabilities of reaching a fixed percentage of the final target was almost always significantly superior when closed-loop polices are employed.
“Lastly, we observed multiple cases in which the allocations used in the closed-loop policies dominated returns from the allocations in the default policy. In particular, by combining the riskier and less risky funds in the system, our models were able to obtain allocations with better expected return and lower volatility than the ones used by default policy.”
The paper was written by Daniel Duque and David Morton of Northwestern University and Bernardo Pagnoncelli of the Adolfo Ibanez University.
Updated: November 20, 2019