Australia’s retirement industry is the envy of the world, yet few Australians know if they’re on track for a comfortable retirement. Wade Matterson says data and algorithms are a key part of the solution.

Australians hold an incredible $2.3 trillion in retirement savings – more than the country’s annual GDP – and yet few people know if they’ll have enough to live the life they want in retirement.

It’s no surprise given the confusing and contradictory advice directed at couples saving for retirement. Media publications regularly quote retirement nest egg targets ranging anywhere from $640,000 to $2 million.

A key problem is the input data – it’s based on qualitative surveys of retirees that have only a tenuous relationship with their actual needs, wants and desires. We know this because big data has shone a light on investors’ actual lives and shown just how inaccurate even their own estimates are when compared to their real-world choices.

Knowing who people really are is just the first step – it’s equally important to know how to help them achieve their goals. Understanding people takes time, but when big data is combined with powerful algorithms, it can fast-track the process by creating direct and clear investment strategies linking investors and their goals.

Just as we use Google Maps to help us identify the best way of getting from point A to point B, big data combined with powerful algorithms can identify where retirees are now, where they want to go and the best way of making sure they get there.  

It’s time for data, analytics and finally, appropriate products, to form the basis for good financial advice, leading the way towards personalised retirement goals.

Clients: Know thyself with big data

Financial planning questionnaires tend to ask clients how they believe they will react in different circumstances (‘stated’ preferences) but good advisers know this rarely gives the full picture.

‘Revealed’ preferences – either real-world data or simulated events that trigger real-world responses – are far more accurate.

Without a deeper understanding of decision theory and behavioural economics, advice that appears sound can easily misdirect investors towards the wrong goals.

For example, more than two-thirds (69 per cent) of seniors in a recent ASIC survey (Building Seniors’ Financial Capability) said ‘having enough money to enjoy life and do what they want to do’ was a financial priority. Yet more than half of retirees spend less than the Age Pension, according to the Milliman Retirement Expectations and Spending Profiles (Retirement ESP) analysis of more than 300,000 retirees’ actual spending patterns.

This contradiction may be explained by a fear of outliving their retirement savings (not helped by out-of-reach retirement savings targets) or a desire to leave something for the next generation.

Similarly, many retirees state their strong desire for consistent income but few will buy annuity-type products: a latent desire to have more flexibility with their savings (and other drivers) often takes precedence.

A good adviser can better understand these discrepancies by using big data, which helps reveal not only what retirees are doing now, but what they are likely to do in the future.

While super savings targets vary widely, we know that retirees aged 65 to 69 actually spend a median of just $31,068 from all income sources each year. A savings target of around $640,000 equates to the top quartile of investors (although many of these will be the ones who seek personal financial advice).

More importantly, big data allows advice to be tailored across wealth bands, age, location and other filters. For example, a client may plan for their retirement expenditure to remain constant or even rise over retirement, whereas the Milliman Retirement ESP shows that, in practice, it tends to drop.

This type of real-world data can help recalibrate retirement expectations and behaviour, so that it better matches personal goals.

Algorithms power personal advice rather than replace it

Big data means little if it doesn’t carve a clear path that brings investors closer to their goals.

Financial planning is a complex business. At its heart is managing clients’ assets and income needs to meet their changing expenses over time. It’s an asset-liability problem, and one that has been successfully managed by large businesses and institutional investors for decades.

Algorithms can bridge this divide and, thanks to powerful cloud-based technology, households can now use the same institutional grade analytics as global firms. Risk tolerance, which was once the centrepiece of simple, product-based advice, is just one component of this holistic, goals-based approach.

This approach is inherently more complex, but a goals-modelling engine should be able to run what if, trade-off, tracking and optimisation analyses that incorporate client preferences and risk profiles. It should also allow for feedback loops and intelligent learning, ultimately offering real-time household asset-liability analysis across multiple advice channels.

Advice algorithms must be accurate, complete, consistent, fast (real-time), supported and backed by a strong company.

This confidence then allows advisers to focus on differentiation, their business strategy and the user experience. The result is more customised goals, which can radically change investment advice and recommended products.

Products: Beyond investment returns and closer to personal goals

The Milliman Retirement ESP has revealed many surprising aspects about retirees, including that more than half spend less than the Age Pension. While we don’t yet know the full qualitative reasons behind this behaviour, we know that risk is a key concern.

Longevity risk is a reasonable fear for many people, given that a 60-year-old man is now expected to live for a further 26.4 years and a 60-year-old woman for 29.1 years, according to the Government’s 2015 Intergenerational Report.

Many investors would prefer to have more savings before they retire, given these growing lifespans. It requires saving more, spending less or boosting returns.

Unfortunately, cash and other safe-haven investment returns, like interest rates, have been slashed in many parts of the world. Asset returns are expected to be more subdued, as central banks begin reining in the loose monetary policy that has powered markets in recent years.

Many investors also need to balance the desire for higher returns with a deep sense of loss aversion, given they no longer have time to recover from extended market downturns.

Sequencing risk describes the heightened risk that an investor with a large balance approaching retirement, or in retirement and drawing down a pension, faces. Younger investors, by way of contrast, are often making contributions (rather than withdrawals) and can benefit from an eventual market rebound.

The holistic advice journey that begins with big data and continues with algorithms should end with a personal goal and strategy, rather than sell relatively short-term one-year investment performance.

When investors can clearly see their five to 15 year personal goals, explicit risk management strategies make sense.

Avoiding large and sustained market downswings means investors are more likely to meet their actual goals. Firstly, it minimises sequencing risk and, secondly, it makes investors less susceptible to making poor decisions under stress.

For example, a post-GFC study by the Centre for Retirement Incomes and Financial Education Research found a surge of industry fund super switching beginning in October 2008 and ending in March 2009, just as the market reached its low point. The market then surged more than 44 per cent, but the report’s authors found investors still remained in their low-risk investment options by September 2009.

Milliman’s ‘Even Keel’ model portfolio strategy is one new offering that allows dealer groups and platforms to apply an institutional-grade, rules-based market risk management overlay to any model portfolio’s underlying equity investments. It allows investors to retain a strong exposure to growth assets, while the downside risk is managed across their own unique portfolio.

Innovative products, built on big data and algorithms, mark the final step in the holistic advice journey. It is only by bringing together these disparate elements to underpin quality personal advice that investors can achieve their desired retirement goals.

Wade Matterson is Practice Leader, Australia at Milliman.