As an industry we generally tell people what their retirement will look like on average; this is known as deterministic modelling. Typically, we provide little information around the range of retirement outcomes which may be experienced, known as stochastic modelling.

Deterministic modelling is used widely across the superannuation and wealth management industry, through the software that underpins the advice process to the tools attached to platforms and superannuation fund web portals and everything in between. Stochastic modelling is not.

Effectively we tell people what to expect but we haven’t coached them for what may occur. We leave consumers unaware and uneducated. This is all fine when markets are performing well, but what happens during the tough times? Most likely, people will experience a worse retirement than the one communicated to them. But it goes further than that: people don’t know the size of the impairment to their retirement plans, and they will likely have reduced confidence in any updated guidance provided to them. Hardly the platform to spend with confidence in retirement.

It would be understandable if people felt that they have lost control of their retirement. Can anyone blame people for feeling anxious in such a situation? No wonder panic-led decisions are made in areas such as switching.

As an industry we are part of the problem.

Where’s the guidance?

Advice and guidance are both failing when it comes to communicating the range of possible retirement outcomes.

Guidance is the collection of general information and interactive tools provided to consumers. Examples include retirement income estimates provided on annual super statements, and interactive projection tools.

This is a collective industry problem. Even ASIC’s moneysmart retirement planner, which is better than most retirement calculators provided by super funds, provides deterministic guidance.

A brief explanation of deterministic & stochastic modelling

Consider a single male homeowner who is about to retire aged 67 with a superannuation balance of $500,000. This example already accounts for the age pension.

The flaws of deterministic guidance have been known for many years. Some firms have been looking to implement better advice and guidance solutions, but it is hard to name with certainty any group which has implemented better practices.

The opportunities are great.

Consider the advice message: “As a household you face an uncertain retirement. We work to assess those sources of variability and account for them in the plan we have designed for you. These are the range of retirement outcomes that we see as possible. Are you comfortable? Let’s consider some of the bad outcomes and work through the impact on you. If necessary we can modify the financial plan.”

So why does the status quo, deterministic advice, persist?

When it comes to retirement we can do so much more for consumers. Why don’t we?

There are many possible reasons but two carry the most weight. One is cost.

Complex models are expensive to develop, particularly in the Australian market where the means-tested age pension adds to modelling complexity. The financial services industry has experienced much structural change over the last decade; perhaps this upgrade to advice modelling has continually been ‘parked’.

The other issue is more concerning. Sometimes I wonder if we, the wealth management and superannuation industry, are complexity-averse. I don’t mean complexity for the sake of it. I mean solving complex problems and then tackling the associated challenge of being able to communicate effectively with consumers.

Does the medical industry not develop new treatments? Are new energy technologies not explored? Does the motor industry not introduce new safety features? Each of those industries embraces the challenge of solving a problem and explaining the benefits in consumer-friendly language.

I hope our industry is one which embraces the complex challenges.

I’m sure some industry participants will raise the regulatory environment as an additional obstacle. I’m sure our regulators are looking to be enablers rather than roadblocks, and that there is an opportunity to consult.

Advice and guidance provided to consumers is one area where I’ll be looking to make a positive impact with The Conexus Institute, leading to better outcomes and less anxiety for Australians.

8 comments on “Flawed retirement modelling compounds client anxiety”
    David Phillips

    David dynamic cash flow modelling software can demonstrate any outcome with any change..

    Hi Adam – thanks for sharing your thoughts. Your points are well made. The point on modelling to life expectancy is a particularly important one. I know of some planners who, to be conservative, add 5yrs to life expectancy in their modelling. Unfortunately this still doesn’t capture the possibility of some extreme outcomes, or the interaction of investment returns with age. Cheers, David

    Hi Ron, respectfully I compare your practices to those highlighted by Chris Craggs (comment above yours). Chris’s clients have been educated that there are a whole range of possible outcomes. The financial plan produced accounts for those and the steps you would take during a crisis. So if / when a crisis comes along, I’m confident that Chris’s clients would be better prepared, feel like they haven’t lost control of their retirement, and while stressed, I’m sure their levels of anxiety are lower. Regards, David

    Chris – I’d been looking to call out some great examples in practice, so well done to you and your firm. I’m sure that you had been able to show clients a possible path not too dissimilar to the present one led to your clients being better informed and felt like they hadn’t entirely lost control of their retirement. In terms of the diagrams themselves… you are absolutely right. I suspect we could have an interesting question about this which could last a day! To try and make the point to the readership I chose to show the “fall-of-the-cliff” diagram. However, our submission to the Retirement Income Review goes into a lot more detail. Check it out if you like: https://conexusfinancial.com.au/wp-content/uploads/2020/03/Retirement-Income-Review-Submission-The-Conexus-Institute.pdf Best, David

    This is certainly an area that needs to be explored further. The other flaw in retirement modelling is the failure to account for potential increase in expenditure in the frailty years at the end of retirement. On average this could represent 15-25% of retirement years. Planning ahead and setting expectations can mean a higher quality of life during this period.

    Chris Craggs

    Hi Ron,
    I have been using Monte Carlo simulations for almost 20 years that map out best and worst outcomes for clients. Instead of creating anxiety it prepares clients. We are in bad market conditions right now, and my clients have a Lifepath that has modelled this exact situation. I can point to the chart and show the clients where they are at, and how this current situation was planned for.

    In my opinion, deterministic models are akin to deceit (we absolutely know that the outcome portrayed will not eventuate, but we present it to the client anyway.)

    David’s stochastic model isn’t much better (as displayed). If this was shown to a client in February, the client would not be wrong to ask why their capital is lower now, because the graph doesn’t display an immediate drop in capital, only the long term consequence of volatility.

    A monte carlo simulation shows an immediate drop in capital, thereby showing clearly to the client what could happen if they happen to pick the very worst day to start investing.

    And these models need not be complex, they can be created quickly within Excel.

    If anything is flawed, it is David’s suggestion to build stochastic modelling into the advisory process. For a start, the introduction of uncertainty will only increase the anxiety of both the client and the adviser. Further, there will be no offsetting benefits from this greater anxiety in a world where the typical advice at a time of crisis is to just do nothing and everything will work out OK at some distant time in the future.

    Adam Miliszewski

    Very valid point. I am often surprised at the emphasis I have seen placed on modeling long-term outcomes to show that clients are in a better position for following advice when the assumptions are so variable. Over-emphasis on a financial model surely creates some anxiety for clients when there is a negative deviation – i.e market downturn or unavoidable and unexpected expenses. In my opinion, the value of ongoing advice is the ability to support clients through the unexpected events and rationalize their financial position so that their anxiety is reduced at critical moments. Stochastic modeling would certainly help shape these conversations better. The biggest variable in retirement modeling is how long someone will live. I don’t want my clients to run out of money but I would hate to see people who have worked hard for 30-40 years to build a strong financial position further defer the life goals – travel, family assistance etc because of some simple model that says they will live to the average life expectancy. Focusing on the variables and adapting as necessary is surely more valuable.

Join the discussion