A revolutionary program that offers robo-advice on optimal investment plans for individuals could change the face of retirement offerings for superannuation funds.
The model, which has been created by three actuaries, factors in detailed statistics on investment, health and mortality, alongside personal data on wealth, income requirements and risk preferences.
The multiple inputs have been made possible by advances in IT which can factor in billions of permutations in under a minute for individuals – around five years ago such calculations might have taken several days to process.
Dubbed the complex adaptive retirement strategy or CARS, by its creators, the Melbourne based Jeffrey Chee and Paul Newfield of Towers Watson and David Schneider, head of research at UniSuper, it has been built primarily out of intellectual curiosity around the possibilities of an all-encompassing post-retirement decision framework.
The model factors in 50 years of investment data on 44 asset classes with 2000 simulations of what could happen to them each year. It adds to this 1000 demographic simulations on health and mortality over 50 years, annuity rates, popular retirement products.
It can quickly produce suggested investment and consumption plans for those in defined contribution plans and it is expected to be most relevant for those with retirement balances of $200,000 upwards, as those under this figure will have a greater reliance on the age pension.
It is believed there are no immediate plans to commercialise the model, until it has been peer-reviewed to work out where its weaknesses lie.
Newfield said: “We absolutely welcome anyone to tell us which areas of the model or framework need improving and we will debate and discuss that and see if we can make improvements.”
Initial testing has already served to expose the weakness of many retirement beliefs. The language of retirement planning is to target a sustainable or a desirable income, but the model illustrates how as a person’s health and spending needs changes, then so should their investments and their income levels.
While the wide variety of scenarios created by the program challenges widely held beliefs that individuals should target $1 million in savings for a comfortable retirement. Testing shows that across millions of potential scenarios an individual might need anywhere from $300,000 to $2 million to fund their retirement.
The suggested optimal investment plans created by the model are also expected to challenge the maxim that one should derisk their investments as they get older.
Schneider recommends that retirees should use the model on a quarterly basis to update their spending and investment plans. “As your spending needs change, as your health status changes, as you learn more about the investment scenarios, you adapt your investing and your spending,” he said. “The question we have sought to answer is how to build a system that can guide people in an environment which is inherently chaotic.”
While the model could be used in conjunction with a financial adviser, Jeffrey Chee says in many ways the model goes further than existing best practice. He said that not only does advice often lead individuals down a path of risk of their own choosing, rather than their optimal strategy, but financial advisers often lack advanced predictions of what could happen to asset classes.
“Where advisers tend to be less strong is around the understanding of stochastic modelling and the distribution of outcomes and some of these more complex statistical ideas,” Chee said. “This could augment the toolkit that a financial planner already uses to help them have metrics they can communicate to their clients. It could also improve the consistency of advice provided to individuals in the same situation.”