Capital losses die with you: the "use it or lose it" tax asset retirees overlook

One of the quieter features of Australia's capital gains tax (CGT) system is that a deceased person's carried-forward net capital losses are extinguished when they die. An inherited asset's cost base passes to the beneficiary, which defers the tax; but capital losses do not work that way. The Australian Taxation Office (ATO) is explicit that if someone had unapplied net capital losses when they died, those losses don't transfer to a beneficiary or to the legal personal representative, who cannot use them to offset any capital gains. For a retiree who has built up substantial carried-forward losses — from the GFC, a failed business or investment, a margin-loan wipe-out, or speculative shares that went to zero — those losses carry a "use it or lose it" character that sharpens with age. If they are never applied against realised gains during the person's lifetime, they vanish at death. That sets up a specific and often-missed planning point: a retiree holding both large losses and large unrealised gains may benefit from realising some of those gains while alive, so the losses are put to work rather than wasted. It is not automatic, though — it has to be weighed against the value of letting assets pass to beneficiaries at the deceased's cost base.

How do capital losses actually work?

The mechanics are familiar but worth stating precisely. A capital loss arises when you sell a CGT asset for less than its cost base, and a capital loss can only be applied against capital gains — never against ordinary income such as salary, pension payments, interest, or dividends. In the year a loss is realised it offsets any capital gains that year, and any excess becomes a net capital loss carried forward indefinitely, with no expiry, to use against future gains. The ATO's own method statement makes the sequencing clear: you subtract your capital losses from your capital gains (including any net capital loss carried forward from earlier years, which you subtract first) before you apply the 50% CGT discount. That order matters, and it shapes how much a loss is really worth, as we'll see.

Why don't capital losses move to other people or entities?

The feature most retirees don't know is that a capital loss belongs solely to the taxpayer who incurred it, and it can only ever offset that same person's capital gains. A loss cannot be transferred to a spouse, even where the spouse has gains it could offset and even where assets are jointly held; each spouse's losses offset only their own gains. An individual's losses cannot be applied against gains made by their family trust, their company, or their self-managed super fund, because the loss is personal to the individual, not to the broader family or any entity. And on death, as the ATO confirms, unapplied net capital losses are simply extinguished — they do not pass to the estate, the executor, or the beneficiaries. The loss is a personal tax attribute that ends when the person does.

What's the asymmetry that creates this planning issue?

Here is the contrast that drives the whole planning question. When a beneficiary inherits a post-CGT asset, they generally take it at the deceased's cost base — the unrealised gain travels with the asset, and the CGT is deferred until the beneficiary eventually sells, perhaps years later, perhaps in a lower bracket, perhaps with the 50% discount. The CGT event on the transfer to the beneficiary is disregarded for the estate. So gains defer gracefully across death — good for beneficiaries — while losses are wasted at death. The deceased's losses can still be used against any capital gains in their final, date-of-death tax return, but anything left over after that return is gone for good. The estate is a fresh taxpayer during administration and does not inherit the deceased's pre-death losses.

What is the "use it or lose it" strategy?

This is where the planning point bites. A retiree who dies holding large carried-forward losses they never used has effectively thrown away a tax asset. Where a retiree has both carried-forward losses and unrealised gains — a long-held appreciated share parcel, an investment property sitting on a big latent gain, a managed fund grown substantially — realising some of those gains lets the losses absorb them. Realise a $100,000 gain fully offset by $100,000 of carried-forward losses and the gain is effectively tax-free, shielded by the loss. The retiree then holds the cash, or can buy back the same or a similar asset at today's higher price, giving the new holding a higher cost base. Either way the loss has been monetised — as tax saved on a gain the retiree wanted to realise anyway, or as a cost-base uplift that shrinks the future taxable gain for the beneficiaries. A loss that would have died with the retiree has been put to use.

Why isn't this strategy automatic for every retiree?

Realising gains early sacrifices the deferral you'd otherwise hand to beneficiaries, so the decision is a genuine trade-off: the value of using the loss now, against the deferral benefit given up. Realising to use the losses tends to win where the losses are large and would otherwise be wholly wasted, where the beneficiaries are high-income or likely to sell soon anyway, and where the retiree is in declining health with a closing window. Deferral tends to win where the losses are modest, the beneficiaries are low-income long-term holders, or the asset will attract further concessions later. There is also a sequencing refinement worth knowing: because losses are applied before the 50% discount, using a loss against a discountable (12-month-plus) gain "spends" it against a gain that would have been halved anyway. The ATO suggests subtracting your losses from any gains not eligible for the discount first, for the lowest CGT. That said, a wasted loss is worth exactly zero, so using one even against a discountable gain still beats losing it entirely.

Two practical cautions sit alongside this. First, the family home is generally exempt under the main residence rules, so it produces no gain to absorb losses — only investment assets do. Second, executing the strategy requires the retiree, or an attorney under an enduring power of attorney, to actively sell, and whether an attorney can undertake tax-planning sales depends on the power's terms and state law. A retiree who has lost capacity may simply be unable to act, which is why the window can close quietly. The sensible approach is to use the losses opportunistically — alongside a rebalancing, a de-risking, or a sale that is happening anyway — rather than forcing transactions purely for tax. And because carried-forward losses are often very old (GFC-era losses can still sit on a return 15-plus years later), keeping documentation of the original loss events is worth doing, in case the ATO ever queries the balance.

Worked examples

These two cases show the loss-extinguishment issue in practice. They are illustrative only and not personal advice.

Geoffrey, 78, widower, in declining health. He has about $240,000 of carried-forward capital losses dating back to the GFC and some failed mining stocks, and a long-held blue-chip parcel with an unrealised gain of roughly $300,000 (cost base $150,000, value $450,000). He intends to leave the portfolio to his two adult children, both high-income professionals. If Geoffrey does nothing and dies, the $240,000 of losses are extinguished — wasted — and his children inherit at his $150,000 cost base, facing CGT on the full gain when they eventually sell at their high marginal rates. On these facts the loss-usage strategy is compelling: realising around $240,000 of the gross gain now, fully offset by the carried-forward losses, makes that realised gain tax-free, and buying back gives the new parcel a cost base roughly $240,000 higher — reducing the eventual taxable gain for his children. On these facts that is generally rational, with the loss converted into a cost-base uplift; given his health, the window is closing, so capacity and whether his attorney can act both matter.

Patricia, 70, good health, modest portfolio. She has just $15,000 of carried-forward losses from one bad investment, and her main assets are her exempt home plus a $200,000 share portfolio she means to hold long-term for her low-income daughter. Here the strategy is much weaker: the loss is small, there is no closing health window, and her daughter would likely face minimal CGT on eventual sale anyway, with the discount and a low marginal rate. Forcing a sale purely to use $15,000 of loss would disrupt a portfolio she wants to keep, for little benefit. On these facts it is generally rational to simply keep the $15,000 loss available to offset any gains that arise naturally if she rebalances, note it on file, and revisit only if her circumstances change — letting the deferral benefit of passing the portfolio at cost base win.

For retirees with carried-forward capital losses, the extinguishment of those losses at death is a genuinely overlooked point. The work is to find the carried-forward loss balance on the tax returns, flag its "use it or lose it" character in the estate-planning review, identify unrealised gains that could absorb it, and model the use-now-versus-defer trade-off rather than assuming either answer — coordinating any loss usage with portfolio changes already in train, and watching the health-and-capacity window that makes it time-sensitive. The asymmetry at the centre of it — gains defer across death, losses don't — means a retiree's carried-forward losses are a real tax asset that simply evaporates if left unused. The first step is knowing they are there, and that they won't survive the client.

Sources

• ATO — How to calculate your CGT (losses applied before the discount)
• ATO — How CGT applies to inherited assets (unapplied losses don't transfer on death)
• ATO — Cost base of inherited assets