Treasury released capital gains tax modelling that underpins billions in budget revenue projections, but the methodology skips over the assumptions that matter most: how investors actually respond when the rules change.
The modelling treats investor decisions as static. It assumes holding periods, portfolio turnover and transaction volume remain roughly constant after the CGT discount changes. Real-world investor behaviour since the announcement suggests otherwise.
The baseline problem
Treasury’s revenue estimate depends on a fixed denominator: the number of properties sold, the average gain realised, and the tax collected per transaction. Change any of those variables and the budget arithmetic breaks.
The first variable already moving: anecdotal reports from mortgage brokers and buyers’ agents point to investors holding assets longer to defer the higher tax bill. Longer holding periods mean fewer transactions. Fewer transactions mean less revenue collected in any given year, even if the rate per transaction rises.
The second variable: portfolio composition. If investors shift toward assets with lower embedded gains or longer time horizons, the taxable gains Treasury expects to capture simply don’t materialise on the same schedule.
What the model assumes and what it ignores
Treasury’s approach prices in two mechanisms: the change in the CGT discount rate and a rough estimate of how much unrealised gain sits in the investor property stock. It does not account for:
- Behavioural shifts in selling decisions. Investors can defer sales, restructure ownership, or accelerate transactions before the rule takes effect. Each decision changes the timing and size of revenue collected.
- Supply-side feedback loops. Fewer investment property sales mean less stock cycling through to owner-occupiers or new investors. Reduced turnover tightens supply in markets already constrained, which feeds back into price growth and affordability outcomes the policy was supposed to address.
- Cross-market substitution. If property becomes less tax-efficient relative to other asset classes, capital flows elsewhere. That reduces the pool of taxable gains in property without increasing revenue from the alternatives, because those gains may be deferred longer or structured differently.
- Holding-period distribution changes. The current CGT discount rewards holding beyond 12 months. A higher tax burden on long-held assets pushes some investors to sell sooner, others to hold indefinitely. The net effect on revenue depends on which group dominates, and Treasury hasn’t published that breakdown.
The discount vs reality gap
Official estimates rely on historical holding periods and transaction volumes. Those patterns formed under a different tax regime. Extrapolating them forward assumes investors don’t adjust, which contradicts every behaviour pattern observed when capital tax settings shift.
Data providers track days-on-market and transaction counts monthly. Since the CGT changes were flagged, transaction volumes across 93% of Australian suburbs have declined. Some of that reflects rate settings and affordability constraints, but the tax signal compounds it.
If holding periods extend by even six months on average, the revenue Treasury expects this financial year gets pushed into the next. If they extend by two years, the budget forward estimates no longer align with cash flow.
Second-order risks Treasury hasn’t modelled
The policy sits inside a system with feedback mechanisms Treasury’s static model doesn’t capture:
- Lender appetite for investment loans contracts. Banks price risk on projected cashflow and exit liquidity. A higher tax on exit reduces liquidity, which tightens credit or raises the price. Three lenders have already lifted pricing in recent weeks, citing funding costs and risk settings.
- Developers face compressed timelines and thinner margins. Higher CGT on project sales eats into feasibility. Projects already marginal drop out of the pipeline. Developer collapse timelines have compressed to days as lender confidence breaks on thinner buffers.
- Retirees adjust drawdown strategies. Investors using property to fund retirement may hold longer or sell other assets first, deferring the CGT event Treasury is counting on.
Each mechanism pulls revenue forward or pushes it out. The modelling assumes they cancel out or stay neutral. That’s an assumption, not a certainty.
What the data would need to show
To test whether Treasury’s revenue forecast holds, you’d need:
- Monthly transaction volumes by investor vs owner-occupier, segmented by holding period.
- Average capital gain per transaction, updated quarterly, compared to the historical baseline Treasury used.
- Portfolio turnover rates for investment property holdings, before and after the announcement.
- Substitution flows into other asset classes from investors who historically allocated to property.
- Regional breakdowns, because holding behaviour in Sydney’s inner ring differs from regional Queensland or Perth’s growth corridors.
Treasury hasn’t released that segmentation. Without it, the revenue estimate is a point forecast with no disclosed confidence interval.
The counterfactual Treasury won’t run
Every budget measure gets modelled against a counterfactual: what happens if we don’t do this. The CGT changes were modelled against a static baseline where nothing else moves.
The real counterfactual: what happens if investors respond by holding longer, shifting portfolios, or exiting the asset class entirely? What does that do to rental supply, to price growth in tightly supplied markets, to first-buyer access?
Treasury’s modelling treats those as externalities. They’re not. They’re the variables that determine whether the policy achieves its stated goal or produces the opposite result.
The catch
- Treasury’s revenue forecast assumes investor behaviour stays constant after the tax changes. Every precedent suggests it won’t.
- Longer holding periods defer revenue. Fewer transactions shrink the tax base. Portfolio shifts move gains outside the modelled timeframe.
- The policy sits inside a system with feedback loops: credit tightening, supply pipeline constraints, and affordability pressure. Treasury’s static model doesn’t capture those.
Scenarios over the next 12 months
Base case: Transaction volumes drop 8-12% as investors defer sales. Revenue comes in 15-20% below forecast in the first year, recovers partially in year two as some deferred sales complete.
Downside: Investors hold indefinitely or restructure ownership to avoid the higher rate. Revenue miss exceeds 25%, and the supply-side effect tightens rental markets in already-constrained cities.
Upside: Investors front-run the change, accelerating sales before the new rate takes effect. Revenue spikes in the transition year, then falls sharply afterward as the deferred pipeline empties.
Treasury’s forecast assumes the base case holds and the downside/upside risks cancel out. The modelling provides no sensitivity analysis to test that assumption.
Bottom line for investors and policymakers
If you’re holding investment property, the decision tree now includes a tax variable that wasn’t there 12 months ago. Run your own numbers: what’s the after-tax return if you hold another two years vs selling now? What’s the opportunity cost of that capital locked in property vs redeployed elsewhere?
For policymakers, the risk is straightforward: if the revenue doesn’t materialise because the behavioural assumptions were wrong, the budget hole gets filled some other way. Either through higher taxes elsewhere, spending cuts, or higher deficits.
The modelling assumes a static world. The market doesn’t work that way. Subscribe to Australian Property Review for weekly updates on how policy changes play out in lending, transaction data, and supply pipelines.
General info, not financial advice.
