The Government has now released draft legislation to implement the capital gains tax changes announced in the May 2026 Federal Budget. So how might this impact people considering withdrawals from super in light of Division 296 tax?
In a nutshell – the changes make it less attractive to withdraw money from super. That doesn’t mean no-one should do so, but the changes weaken the case for taking action.
Previously, the comparison between super vs non super has been quite simple for Division 296 tax. Because the amount of a capital gain has been calculated in the same way for all types of taxpayer (sale price less original cost), the only differences have been:
I’ve found this table quite a useful way of explaining things to clients:
| Tax on capital gains (allowing for discounting) before any changes |
|||
|
Super
|
Own name
|
Company
|
|
| $0 - $2m (pension phase) | 0% | Up to 23.5% | 30% |
| $2m - $3m | 10% | ||
| $3m - $10m | 20% | ||
| $10m + | 26.67% | ||
The super rates all allow for the fact that capital gains are eligible for a 1/3rd discount. They also reflect the fact that different rates of tax apply for different “slices” of an individual’s super when we take Division 296 tax into account. The 26.67% figure in the table above, for example, is 2/3rds of 40% (the total tax applied to earnings relating to the balance above $10m – being 15% fund tax plus an extra 25% Division 296 tax).
The tax rate shown for an individual assumes that the people we’re talking about (people with Division 296 tax problems), usually already have income in their own name. While that income alone might not be able to push them into the top marginal tax rate every single year, a large capital gain certainly would. Hence I’ve assumed the capital gain, after discounting, would be taxed at 47% (and 23.5% is 47% x 50%).
In future, the tax rate (let’s assume it’s 47%) will be applied to a capital gain calculated in a totally different way (sale price less an inflation adjusted cost base). That means we can’t just compare tax rates.
The solution in my view is to come up with a specific metric for the comparison. To compare personal ownership with super on a like-for-like basis, I have expressed the personal tax outcome as a percentage of the ordinary capital gain (sale proceeds less original cost). Remember, this isn’t a statutory tax rate. It is simply a comparison tool.
So what are the variables at play?
Obviously inflation is one – particularly when measured relative to the growth in the asset. If inflation is high and asset growth is low, the new method will actually demand very low amounts of tax.
To use an extreme example – if inflation is 2.5% pa and the asset only grows by 3% pa, a $100,000 asset will have grown to around $134,400 in 10 years. The indexed cost base will have grown to $128,000. The new approach will see only $6,400 ($134,400 - $128,000) subject to tax (a tax bill of $3,000 at the highest marginal rate of 47%). This is equivalent to a tax bill of 8.7% of the whole capital gain (ie, $3,000 is 8.7% of $34,400). I’ve referred to this as the “comparison rate” – ie the rate we can compare to the super fund rates above.
In contrast, the same asset held in a super fund would have a $34,400 capital gain ($22,900 after discounting) and a tax bill of $9,200 if we assumed the highest possible tax rate of 40%. (This isn’t actually how super fund taxes or Division 296 tax work but hold that thought for a moment.)
In contrast, if the asset grew at 8%, the tax bill if it’s sold after being held personally would be more like $41,300 (which equates to a tax rate of around 35.6% on the capital gain of $115,900). Remember this isn’t a real tax rate (this is still assumed to be 47% on the capital gain relative to an indexed cost base), it’s just a way of expressing the tax paid in a way that makes it directly comparable to the rates in the table above.
The table below shows a range of “comparison rates” – ie, the effective tax burden for a given mix of inflation and growth, expressed as a percentage of the overall growth.
| Timeframe (yrs) | Inflation 2.5% / Growth 3% | Inflation 2.5% / Growth 5% | Inflation 2.5% / Growth 8% | Inflation 2.5% / Growth 12% |
| 5 | 8.2% | 24.6% | 33.8% | 38.9% |
| 10 | 8.7% | 26.1% | 35.6% | 40.7% |
| 15 | 9.2% | 27.5% | 37.3% | 42.3% |
| 20 | 9.8% | 28.8% | 38.8% | 43.5% |
This isn’t relevant for my analysis here – I’ve assumed a tax rate of 47% in all scenarios (thanks to the individual’s other income). While it’s a highly relevant change for other people, I’m guessing those with very large amounts in super who might make big withdrawals to avoid Division 296 tax aren’t paying tax at less than 30% on any of their income – let alone capital gains.
When personal assets are sold and subject to capital gains tax, it’s easy to focus on marginal rates because we can assume “other income” will push the recipient into a higher rate of tax. (This is why the above focussed entirely on 47%)
It’s not quite the same for super funds.
Because the tax rates depend on wealth not income, even when a member has more than $10m in super, every dollar of income and discounted capital gains will be split into various parts – with some taxed at 0 - 15%, some at 30% and some at 40%.
Technically, then, the super fund tax rates applicable to a particular amount of capital gain will be lower than the ones shown in the table above. But in order to consider whether to leave money in super vs withdraw it, we usually feel the “marginal rate” approach is a more helpful proxy for the real impact of the choice to leave money in super or withdraw it.
As you can imagine, I will be doing much more comprehensive modelling for our forthcoming Super Intensive Day!
The most comprehensive SMSF CPD event available for financial professionals, delivering up to 15 hours of specialist education across the full superannuation landscape, in a single focused day. Secure your spot now: