Sometimes we can be too clever for our own good, and owning it is the most useful thing I can tell you. Early on, as a planner, I gave in to it — proud of how tightly I could model. Plans tuned to the last dollar, contributions timed to the birthday, every lever pulled. Technically beautiful. Then the real world turned up, and it was a nightmare! But let me come back to that.
Let me start with the answer, because your clients want it and the noise has buried it: for most clients, super is still king. Even with Division 296 on top, it remains one of the most tax-effective structures they can access. The political theatre has been loud enough that people who will never go near a $3 million balance are asking whether to pull money out. None of it is one-size-fits-all — the right answer always turns on the individual client — but the shape is clear enough to lead with, and it comes down to three clear answers, not a bigger spreadsheet.
Here’s what twenty years of watching advisers respond to tax and super change has taught me (other than “here we go again”): the instinct is to retreat into the engine room and build a more complicated model. I’m not immune — there’s a reason some of us end up as techies and in advocacy. So I know the pull first-hand, and Division 296 is made for it: realised earnings, cost-base elections, proportioning, carried-forward losses — enough machinery to justify a very impressive fee and a very confused client. Resist it. Your value here is the clarity of your answer to the three questions every affected — and, ironically, unaffected — client is asking.
First, what actually changed, because a lot of what you read is out of date
- It taxes realised earnings, not paper gains. The unrealised-gains model is gone. Earnings are built on the fund’s taxable income attributable to the member — interest, dividends, rent, realised capital gains, with the usual CGT discount still applying (wherever those finally land).
- Two tiers. An extra 15% on the balance between $3 million and $10 million, and an extra 25% above $10 million. Stack that on the existing 15%, and effective rates land at roughly 30% and 40%.
- Both thresholds are now indexed — the lower in $150,000 steps, the higher in $500,000 steps — tracking the transfer balance cap. The “bracket creep will drag everyone in” argument has lost most of its force.
- Nobody pays for a while. The first assessment is based on your balance at 30 June 2027, and the bill lands after that.
Question 1: “Is super still tax-effective for me?”
For most clients, yes — and this is the number to lead with. Not the marginal rate on the top slice, which is designed to frighten, but the average rate across the whole structure. If a client panics about a 30% headline, show them the effective rate on their entire superannuation position — the grossed-up earnings across the fund, not just the portion Division 296 catches. The story usually holds: super is generally still taxed well below what the same earnings would attract in the client’s personal name, and — depending on marginal rate and structure — often below a trust once you account for distributions. They need a rough rate they can hold in their head, not three decimal places — the precise figure is your housekeeping; the average is their peace of mind.
Question 2: “How much Division 296 am I actually paying?”
Less than they fear, and later than they think. Only the slice above the threshold is caught: if someone has over $3 million, only the proportion tied to the excess, and only on realised earnings. A fund holding long-term assets it isn’t selling has a much smaller footprint than the headline implies. And the final law is kinder than the draft: negative earnings in a year attract no Division 296, and that amount carries forward to offset future positive earnings — as do carried-forward capital losses. The early drafts stranded losses; the final law does not. The “losses just disappear” take describes a bill that no longer exists.
An admission: you can be too clever for your own good on super tax
Back to that nightmare I alluded to, and here’s where the real world turned up. The strategy was one of those beauties — three years, maximise concessional contributions, trigger the non-concessional cap next year to give you an extra year of contributions, move into pension phase. Then the employer paid a concessional contribution early, in the current financial year. That single slip pushed the client over a cap I’d planned to run right up to, triggering excess-contribution issues across three years that flowed through both concessional and non-concessional limits. I’d been right — the employer stuffed up, not me. But I’d tuned the plan so tightly there was no room to absorb one ordinary mistake, so we compensated the client anyway. A lot!
The lesson has stayed with me: precision is fragile, and a plan with no margin of safety isn’t sophisticated — it’s brittle. Banks make errors, employers make errors, the ATO makes errors. Division 296 will tempt every one of us to over-engineer to shave a few dollars off the top slice, and some of those strategies break the first time reality doesn’t cooperate, some break the next time the law changes – and it will. Leave the buffer. Being technically right and strategically wrong is still wrong.
A word on defined benefits
Defined-benefit pensions (PSS, CSS, MilitarySuper, a legacy SMSF pension, etc.) work differently. They are still only in draft — a notional value, a proposed 17.5% haircut, and a hit generally lower than an equivalent account balance. But the margin-of-safety point bites hardest here: a notional valuation or ordinary indexation can nudge a client over $3 million without a dollar of real money moving, so “we’ll know once the regulations are final” is a perfectly professional answer.
Question 3: “Should I restructure?”
Usually: manage it, don’t flee it — but one move is worth making deliberately. The flight instinct sends clients toward trusts, investment bonds, a partner’s name — each with its own tax, its own cost, and its own loss of the concessional environment they’re leaving. For most clients, unwinding super to dodge Division 296 means paying a certain cost to avoid a smaller, later, indexed one — the “too clever” trap all over again.
The one piece of housekeeping worth assessing is the one-off cost-base reset. Because Division 296 only bites gains realised from 1 July 2026, funds can elect to reset an asset’s cost base to market value at that date, so pre-existing gains aren’t dragged in later. But it isn’t automatically right: a fund sitting on unrealised losses, or not planning to sell, can be worse off electing. Model it against the actual asset mix — assess for every affected fund, default for none.
And for SMSFs holding illiquid assets — the classic being direct property — the restructure conversation is really a liquidity one if the fund can’t fund the bill without a forced sale, surface that now, calmly, well before a 2027 assessment turns it into a fire drill.
The real job
Division 296 doesn’t end super — it tests whether we can do the thing that actually earns our fees: take a noisy, frightening change and hand a client back a clear picture of where they stand and what, if anything, they need to do. The adviser who wins the next twelve months isn’t the one with the most decimal places — it’s the one whose clients walk out able to explain, in a sentence, why their super is still working for them.
So here’s the question — the same one I should have asked myself across the desk from my client all those years ago: when your next affected client sits down, are you going to reach for the dollar-perfect model, or the answer that helps them sleep at night?
This article is general information for advisers, not personal financial advice. Some of the detail above — particularly the defined-benefit valuation rules — still sits in draft regulations and may change before the rules are final.
Ben Marshan, Marshan Consulting
