Following user feedback, our Div 296 Modelling Tool has been improved. It now shows the significance of the more important assumptions so you can see why the result you are receiving is not intuitively obvious. Let me illustrate this by way of an example. John is 65, in an SMSF (note that the results differ for APRA Pooled and APRA Non-Pooled funds) has $2m in pension and $13m in accumulation. He is considering withdrawing some of his balance to avoid paying Div 296 tax. Intuitively, based on much of the sector comment, you would expect that withdrawing about $5m and placing it into an insurance bond would be a worthwhile strategy. This is what our optimiser calculates, based on our default settings.
Why does it produce this result and why does it take 14 years for the alternative investments to restore John’s wealth position to what it would be if he hadn’t done anything. The reason is obvious when you consider the significance of various of the default assumptions – all of which can be modelled separately.
The default individual tax rate is at marginal rates. This makes it attractive to place some investments in John’s name. If his external income is such that he would be on the top marginal rate, the sensitivity analysis shows that the whole balance would go into the insurance bond but the strategy would not be better than the do-nothing alternative until the final year of the 20 year projection and that is only because the model liquidates all structures at the end of the term and insurance bonds don’t take a final year dip in value as they are not subject to the end of term capital gains tax cost that superannuation is subject to.
Available fund cash has a significant effect. Perhaps the fund has sufficient cash due to its normal investment processes. That’s fine. The issue is when the fund needs to bring forward liquidation of assets to enable the Div 296 withdrawal strategy. This brings forward the realisation of capital gains, from when they would normally be realised under the turnover frequency of the fund’s investment strategy, to now. The resultant costs and capital gains tax lowers John’s current wealth so the alternative strategy has a lower starting base than the do-nothing strategy. The default assumes no available cash for the Div 296 withdrawal strategy. The sensitivity analysis shows how the results for the alternative strategy improve if either 50% or 100% of the cash required is available without forced sales. You will note that the time for the portfolio value to recover is reduced from year 15 to year 7 and year 2. There is also an increase in the wealth advantage overall.
Another important driver, linked to the need to liquidate assets, is the assumed CGT cost base. The default cost base is 75% of the asset value. The sensitivity analysis considers a drop in the cost base to 50% and 25% You can see that this reduces the value of the result, though, as it changes the optimised asset strategy, it does reduce the recovery time. Our Alliance Partners can view each fund’s unrealised cost base percentage, from the Unrealised Gains label on the left-hand portal menu, to assist with providing this input.
Insurance bonds have an administration fee structure of 0.3% to 0.7% over the underlying asset management charges. We have assumed a default of 0.4%. The sensitivity analysis shows what happens if this value drifts by 0.1% either way. If it drops to 0.3%, the value is improved and the recovery time is reduced by a year but if it rises to 0.5%, the optimiser abandons the insurance bond option and favours an individual portfolio instead.
The investment strategy influences the result by playing off the different tax rates and CGT treatment of the various alternatives considering income, growth and turnover. The different portfolio types have a dramatic effect on both the strategy and results. Not all options are available in all structures but, for comparative purposes, the model disregards this. Turnover and franking credits are particularly relevant. Our default strategy is a balanced portfolio.
So, the decision to withdraw from superannuation to invest in alternative product types, if made simply to avoid Div 296 tax, can cause serious harm to John’s future wealth if the nuances are overlooked. If, like many of the comments I have seen, we disregard initial costs, administration costs and the cost of early liquidation, assuming John’s marginal tax rate is 47%, we get this result.
But to borrow a term, commonly used to describe the answers provided by AI, this is an hallucination.
Final note. The model considers individual, company and insurance bond ownership. It does not consider trusts, as they distribute to these structures as the ultimate beneficiaries. If Budget announcements require trusts to be included as a separate alternative or alter the capital gains tax treatment of any structure, the model will be adjusted.
