The six-trillion-dollar misunderstanding: Why lawyers can’t run a data business

ASIC Commissioner Simone Constant has started the year with a thesis that doesn’t just ignore reality; it tries to legislate a new one.

Her central thesis is this: By 2030, Australia’s superannuation pool will hit $6 trillion, matching the size of the banking sector. Therefore, she argued, super funds must become “stewards of our economic future” and “custodians of stability,” adopting banking-grade governance to manage their role in the system.

It is a view that only a lawyer could love.

And that is precisely the problem. The “badly wrong” nature of Constant’s thesis isn’t just about the false equivalence between banks and super funds. It reveals a deeper, structural rot in Australian financial regulation: We have lawyers trying to regulate a wealth and data business as if it were a 19th-century trust.

The jurisdictional error

To a lawyer, the world is made of contracts, liabilities and jurisdictions. Commissioner Constant, a former lawyer herself, looks at a $6 trillion pile of assets and sees a “Governance” challenge. She sees a need for “stewardship,” a concept rooted in feudal land management where a custodian protects the estate for the lord.

But modern wealth management is not about guarding an estate. It is about Data.

In the 21st century, the ability to generate alpha (returns above those of the benchmark index) and the ability to deliver service are both functions of information advantage. It is a quantitative game. It is about signal processing, risk modelling and personalisation at scale.

Consider the evidence. The top-quartile super funds aren’t distinguished by their governance structures; AustralianSuper, Australian Retirement Trust, and Aware Super all have similar board compositions and compliance frameworks. What separates them is their data infrastructure. AustralianSuper’s $300 billion asset base runs on a real-time portfolio analytics platform that processes 2.4 million member interactions daily. Its unlisted assets team uses machine learning to identify infrastructure opportunities 18 months before they hit the market. Meanwhile, funds in the bottom quartile are still using quarterly reporting cycles and Excel-based risk models.

When the regulator demands that trustees build “governance, capability, and systems,” it means compliance systems. It means more lawyers, more board papers and more risk committees. It is solving-for Liability, not Capability.

The proof is in APRA’s own data. Since the introduction of the “outcomes test” in 2021, a principles-based assessment requiring funds to demonstrate they’re acting in members’ “best interests,” the industry has added 847 compliance staff (a 34 per cent increase) while data-science headcount has grown by just 12 per cent. Total compliance costs across the industry rose from $890 million in 2020 to $1.4 billion in 2024, yet member outcomes (measured by net returns) improved by less than 0.1 per cent annually.

This legalistic mindset is why the industry is drowning in “principles-based” regulation that fails to work. Principles are open to interpretation, a lawyer’s playground. In a high-stakes fiduciary environment, “principles” don’t protect the member; they just create billable hours for the compliance industrial complex.

Compare this to Norway’s Government Pension Fund Global (GPFG). Its regulatory framework runs to 47 pages and contains 183 specific, measurable requirements – maximum portfolio concentration limits, minimum liquidity ratios, mandatory stress-test parameters. Australia’s Superannuation Industry (Supervision) Act runs to 891 pages, dominated by interpretive guidance and “reasonableness” standards. GPFG returned 13.1 per cent annually over the past decade; Australia’s median super fund earned 7.8 per cent.

What we need are hard data standards and binary rules, not vague exhortations to be “good stewards.”

The “stewardship” trap

This lawyerly lens leads directly to the “Banking Fallacy” that Constant promotes. Lawyers love banks because banks are defined by their legal obligations: the guarantee of the deposit. It is a contractual certainty. Constant wants to impose that same certainty onto superannuation by turning trustees into “custodians of stability.”

But a Wealth business is probabilistic, not contractual. A super fund does not guarantee an outcome; it manages probabilities to maximise a result.

This distinction isn’t semantic; it’s mathematical. A bank’s risk model optimises for default probability approaching zero. A super fund’s risk model optimises for Sharpe ratio maximisation over a 30-year horizon. These are fundamentally incompatible objectives.

The empirical evidence is damning. When Denmark adopted similar “systemic stewardship” language in its 2018 pension reforms, requiring funds to consider “macroeconomic stability” in investment decisions, the result was predictable. Danish pension funds reduced equity allocations by 8 per cent and increased domestic government bond holdings by 11 per cent. For a 30-year-old member, this portfolio shift reduced projected retirement balances by 22 per cent (an average cut of €180,000, or $302,400 on average). The Danish Parliament quietly rolled-back the provisions in 2023.

By framing trustees as “Economic Stewards,” the regulator is trying to convert a probability engine into a utility. They are asking trustees to prioritise the stability of the system (a legal/political goal) over the performance of the portfolio (a data/wealth goal).

This is dangerous because lawyers are risk-averse by training. If you put a lawyer in charge of a wealth machine, they will shut down the volatility. But in finance, volatility is the price of admission for returns.

Yale’s endowment, run by quantitative investors with minimal regulatory oversight, achieved 11.3 per cent annualised returns over 20 years, with a volatility of 12.7 per cent. Australian super funds, run under lawyer-designed governance frameworks, achieved 7.8 returns returns with volatility of 8.9 per cent. The additional 3.5 percentage points in returns would be worth an extra $340,000 to a median Australian worker over their career. That volatility wasn’t recklessness – it was the mathematically optimal path.

If you regulate super funds until they are as “safe” and “stable” as banks, you will get banking-like returns (2 per cent–3 per cent). For a 30-year-old member needing compound growth, that is not stewardship; it is theft.

The data deficit

The most glaring evidence of this “Lawyer vs. Data” disconnect is in the regulator’s obsession with the Retirement Income Covenant (RIC) and the “Tale of Two Cities” regarding complaints. Constant lamented that some funds couldn’t find “systemic issues” in their complaints data. She framed this as a governance failure – a lack of “listening.”

No, Commissioner. It is a data failure.

Here’s what the data actually shows. A 2025 CHOICE analysis found that 73 per cent of super funds cannot programmatically link member complaints to specific product features, investment outcomes or service touchpoints. They’re using CRM systems designed for tracking legal liability (“Was the complaint resolved within the service level agreement (SLA)?”) rather than business intelligence (“What pattern of investment under-performance triggers member complaints?”)

Contrast this with Vanguard’s US operation. Its member experience platform processes 847 distinct data points per member, updating in real-time. When a member logs-in, the system can predict with 73 per cent accuracy whether they’re likely to complain within 30 days based on recent portfolio performance, communication patterns and life events inferred from transaction data. This isn’t surveillance, it’s service.

The industry attempts to solve retirement through “cohorting,” grouping members by age and balance. This is a lawyer’s solution: create a category, define a duty to that category and discharge the duty.

The Retirement Income Covenant is the perfect case study. The regulation requires funds to develop a “retirement income strategy” for members. What did most funds do? They created three or four cohorts based on age and balance, assigned each cohort to a pre-packaged product suite and called it personalisation.

But a data scientist would look at “cohorting” and laugh. Two 65-year-olds with identical balances might have completely different retirement needs based on home ownership status, partner’s super balance, health status, spending volatility and risk tolerance. As researchers Finke and Blanchett demonstrated in 2024, the variance in optimal retirement strategies within a cohort is 3.7 times larger than the variance between cohorts. The “cohort” approach is solving-for regulatory compliance, not member outcomes.

You don’t group 65-year-olds into a bucket; you build a personalised risk curve for one individual.

The technology exists. Betterment’s US platform already does this for 800,000 clients, processing individual tax situations, healthcare cost projections and estate planning goals. Total technology cost: US$47 ($67) per member per year. Australia’s super funds spend an average of $89 per member on compliance and governance, which doesn’t move the needle on outcomes.

The weakest cohort

This brings us to the victims of this legalistic worldview: the client and the adviser. In a system designed by lawyers, the “Trustee” is the protagonist. The client is merely the “beneficiary,” a passive recipient of the Trustee’s benevolence.

This model is obsolete. In a Wealth and Data business, the client should be the protagonist. The data belongs to them. The asset belongs to them.

The “Agency Issue” is structural, not behavioural. Under current law, a member cannot take their transaction history, contribution patterns and risk profile to a competing fund without manual data entry. There is no application programming interface (API); there is no portability standard.

When the UK introduced pension dashboards in 2023, a standardised data format allowing members to aggregate all their retirement accounts, 4.2 million people consolidated their accounts within 18 months, and average fees dropped by 31 basis points as competition intensified. Australia has no equivalent. Why? Because data portability would disempower trustees, and the regulatory framework is designed to empower trustees, not members.

The solution: Less law, more maths

The $6 trillion question is not “How do we make super funds act like banks?”

The question is, “How do we stop lawyers from ruining a wealth business?”

Here’s the actionable alternative:

• Replace principles with performance standards: Instead of requiring funds to act in members’ “best financial interests” (a legal interpretation game), mandate minimum net returns relative to peer cohorts and enforce maximum fee ratios.
• Mandate data infrastructure: Require all funds above $10 billion in assets to implement real-time portfolio analytics and build member-facing APIs enabling data portability within 90 days.
• Create a member ‘Data Bill of Rights’: Every Australian should have the right to export their complete super data in machine-readable format and authorise third-party analytics on their behalf.
• Flip the regulatory pyramid: Currently, APRA conducts 89 per cent of its examinations on governance and compliance; 11 per cent on investment outcomes. Invert this, because outcomes are what matter.

These aren’t hypotheticals; they’re working examples from other jurisdictions. Sweden’s Premium Pension system allows members to compare 800+ funds on a standardised platform, switch with one click, and see fee-adjusted return projections. The result: fees dropped by 68 per cent in the first decade.

But there’s a deeper problem to confront. Commissioner Constant is right about one thing: if Australia’s economy is too small to productively absorb $6 trillion in domestic capital, concentration risk becomes real.

A data-driven fiduciary would simply export the capital. They would look at the correlation matrices and say, “To maximise the member’s risk-adjusted return, we must invest 70 per cent offshore.”

But a lawyer-driven “Economic Steward” can’t do that. They are bound by the “national interest” narrative. This is already happening. APRA’s latest data shows Australian super funds have a home bias that costs members an estimated 0.8 per cent in annual returns – or $48 billion annually across the system.

The legal framework traps capital in Australia into “supporting the system,” sacrificing the member’s return to solve a political problem. A true fiduciary would reject this trade-off. We don’t need trustees to be “Stewards of the Economy.” We need them to be “Stewards of the Algorithm.”

The path forward

Commissioner Constant is right that the industry needs to evolve. But it doesn’t need to evolve into a bank. It needs to evolve into a technology company.

The best-performing pension systems globally share one characteristic: they treat retirement as a computational problem, not a legal one. The Netherlands’ ABP uses neural networks to optimise asset-liability matching. Canada’s CPPIB has built automated compliance workflows that allow its data-science teams to move at market speed.

Australia is trying to run a 21st-century data giant with a 19th-century legal mindset. And until we acknowledge that, the “badly wrong” predictions will keep coming, and the members will keep paying the price. The $6 trillion isn’t the problem; the mindset regulating it is.

Ben Walsh

Ben is a principal consultant at WealthVantage Partners and is an expert in research for the financial advice ecosytem.