5 years that changed Australian superannuation reporting, and what comes next.

When we started Nuj, regulatory reporting wasn't the plan. We were a collection of data and technology humans looking for a problem to solve.

I'd recently departed from a business trying to build and brand a new super administration platform, and the experience convinced me that replacing legacy systems that create data wasn't the answer. Those robust heritage systems weren't going anywhere. What was missing was something that could sit across them and make sense of the information they held.

At the same time, there was a regulatory wave of change happening, underpinned by a drive for greater transparency and better financial outcomes for members. APRA's Superannuation Data Transformation (SDT) program started to take shape. It needed information from every corner of a fund, with new taxonomies and validations, far more than any one administrator could pull together. We jumped on it.

Five years after the first submissions, the industry has changed more than most people expected, and it still hasn't finished changing. Looking back, it's clear none of this was ever going to be set-and-forget.

It's easy to forget how contained reporting used to be.

Before 2021, APRA's regulatory reporting mostly sat with administrators, internal or external. The data was relatively high-level; it tied back to the financials, and it could be run pretty much rinse-and-repeat. The weakness was in what came back to the industry. The collections didn’t accurately capture the diversity in the super fund offerings, and the annual publications gave little sense of how funds were tracking against each other. If the goal was a transparent view across the industry, the old approach couldn't deliver it.

SDT looked to change that view entirely. It required a far broader set of information, at a much deeper grain: insurance, expenses, asset allocation, investment performance, fees, liquidity, valuations, and product-level detail. No single team or system holds all of that, so reporting that one function used to manage now draws on both internal (investment operations, finance, product and administration) and external (administrators, custodians, insurers and investment managers/data providers) teams. Super funds either adopt a centralised or decentralised approach to solving the reporting based on the collection of skill sets that sit within their business. The team pulls data together every quarter into a schema that must be validated and correlated before anyone can submit with confidence.

It also asked funds to look at their own data in a new way. APRA was designing one structure to fit all fund types - retail funds, corporate funds, industry funds, platforms and insurance-only, and in places it wanted data that funds had never collected or analysed internally. Interpreting what APRA wanted, then finding and shaping the data to match, has been the hardest part for most funds. More contributors raised a question many hadn't needed to answer before: who is accountable? When investment operations supplies one part and administration another, who signs off? More parties means more governance, more review, and more process, and funds will tell you the cost of compliance has grown sharply in headcount, systems, and change. That's the price of an industry where the outcomes of the underlying members are the most important thing.

We got plenty wrong along the way.

Our first mistake was believing the program would arrive as a finished set piece. We followed APRA closely and built to the plan, then the plan changed, and changed again. That is understandable for a program of this scale and nuance. But it meant building for continual change in the data, the validations and the way they're interpreted. Changes to Phase 1 still come through today.

We also underestimated the crunch at the end of every reporting period. Period-end figures aren't ready until things are closed out, so everything bottlenecks into the final stretch, and whatever you build has to withstand that load. But not everything has to wait. Work can be sequenced and not all left to when everything has dropped. Product and fee structures, pathways and combinations, insurance arrangements and asset allocations can all be prepared and validated earlier, so the close is about the final numbers rather than everything at once.

And we learned how hard the handover culture is to shift. Every party wants to do its piece, sign off and step away, leaving the fund to work out what fits and what doesn't. The better model gets those parties working directly with each other, so issues are resolved in the flow instead of bouncing back and forth.

None of those lessons were about the reporting standards themselves. They were about how data moves through a fund, and that's what separates the funds in a good position today. Funds in a strong position control their data, know exactly who is responsible at each step, and have a structure that takes information from the source, transforms it, and compiles it into something validated before it goes out. Funds in a weaker position still rely on individuals, spreadsheets, email and fragmented workflows. It can be done, but the risk is higher, confidence is lower, and there is an increased likelihood of further interrogation from the regulator. Those programs are always in project mode and always dependent on key people, and that's where we've seen things break down. That reliance is about to matter more.

Having data doesn't mean having insights.

For much of the past five years, APRA has been in a data cleansing phase. Phase 1 data has come through it, with the resubmission process mostly behind us. Phase 2 data is still settling, and Phase 3 is largely a migration of the remaining legacy forms so it won't change much. With all reported data now flowing through APRA Connect, the next stage is analysis.

That shifts the pressure. Now the regulator has a full picture of every fund, any fund that doesn’t have the same view of itself will be exposed, spending time and resources investigating questions it could have seen coming. Regulation will keep evolving with the data, and so will the cost for any fund that hasn’t invested in the right structures. The cost of regulatory reporting is unavoidable; however, the value you get from the information can be significant. It shows the regulator what's happening and where to focus. Use that to build your own picture, get ahead of the questions and decide where to act. The informed decisions you make as a trustee could be what sets you apart.

Retirement is the obvious next frontier. It was part of the early SDT conversation and has since moved into its own framework, with data expected from late 2027. When it lands, expect what we saw with Phase 1: different interpretations, resubmissions and a settling period before transparency flows back to the industry. As the population ages and the balance shifts from accumulation towards drawdown, more attention will go to this phase. Performance testing exists in accumulation, so why treat retirement in isolation? I can't tell you the timing, but I can't see it not ending up as a heat map, a performance test or something similar.

The annual performance test itself is evolving too. This year, 12 products failed, and Treasury is looking at changes, including how private markets are treated. When the test came in, there were plenty of industry concerns, resulting in many conversations, debates and arguments. But its first job was to raise the low-water mark so no product sat below it. With that largely done, it makes sense to tilt it towards bigger challenges, like private markets and housing, without dropping the standard. This is our money that fuels our retirement, however, our retirement will be impacted much more than just the balance that we have in our super.

As for AI, regulatory reporting won't be the first place funds use it, and it shouldn't be. Getting it wrong means fines and incorrect information published back to the industry, and the knock-on effects are severe. Funds will start where the spend and the engagement challenge are biggest: member administration, member engagement, advice gaps, call centres. APRA's message is to adopt it but govern it carefully, and super is a long game. I don't expect meaningful AI in regulatory reporting in the short term, but we have embraced it and intend to be well ahead of it when it comes.

Other markets are on a path that looks familiar.

None of this is unique to super. Wherever regulation grows faster than the systems feeding it, the same pattern appears: more parties, more data and more pressure on a process built for something smaller. In conversations outside super, Excel is still often the main reporting tool, with all the manual effort and risk that comes with it.

Looking north, the UK's Pensions Dashboards, Value for Money Framework and small pots consolidation looks to follow a similar sequence to Australia's: first help people see where their money is, then show how schemes compare, push for scale and then enable consolidation. Data seems to be the hard part there too, because it's fragmented across so many different schemes and structures. Every market is different, and I'm wary of assuming what worked here simply transfers. But Super taught us that structure, data quality, and visibility mattered most as a starting point. Those needs don't belong to one regulator or one country.

That's also why we work the way we do. We don't come in and upend how a team works, or spend years building before anyone sees value. We add something small and meaningful that works alongside the people and systems already in place, shows value quickly and can grow in value from there. We think that's the most effective way to have a deep and meaningful impact on an industry. Prove value all the way through, and people know they're on the right track.

Regulations will keep changing. The need to know your data, trust it and act on it won't.