Cherry-picking cost base after share sale is a tax gift

Indeed, Harry, thanks for helping to support my point. The only reason to list them separately is to take advantage of the ATOs tax gift. The shares are otherwise identical. If the ATO did not offer you a choice of cost base, you would treat them as the same.

A point overlooked by this discussion is that the ATO’s current “generous gift” to shareholders contributes to liquidity in what is a small market by global standards.
If investors were to be constrained by FIFO, LIFO or any other limiting cost base, it would reduce the reason and incentive to the very conduct of trading, which is an integral facet of “right pricing” in our free market system.

Harry, you're doing a Mike Willesee to land a John Hewson birthday cake moment. The GST on the cake and the decorations are treated differently but somehow we raise billions in GST. I am raising a general principle for a homogeneous parcel of shares or units, where the ATO is unnecessarily giving investors a tax-minimising choice. There is no need to know "how to identify the particular shares or units they have disposed of". They are all the same and investors know exactly which shares they are selling - whichever minimise their tax. Every investment property is different but even if they were identical, any tax foregone by a cherry-pick is tiny compared with shares and funds.

They aren’t a homogeneous parcel of shares, they are shares bought in different parcels at different times. You are giving them an equivalence that doesn’t exist. They are judged separately when determining whether they should receive a capital gains tax discount based on the date of purchase because they are seperate capital assets. Because they are separate capital assets then you can dispose of them separately according any ordering you desire.

There is a portfolio of 1,000 Macquarie Bank shares. All the same. They all have the same place in the company's capital structure, they all receive the same dividends, all the same franking credits, they rank equally in a liquidation, they are all measured the same way by APRA for prudential requirements, they are treated the same in a buyback. Etc, etc, etc. They are homogeneous. Does anyone list their shares as 'Macquarie Bank 1999, Macquarie Bank 2005, Macquarie Bank 2015, Macquarie Bank 2021'?

Agree Harry, I number Lots and journalise decision to purchase/why, expected outcome/when, which capital source and actual result. It follows if one adheres to investment plan which after all is what 'experts' advise, that one be able to choose which disposal Lot to meet a plan. Different Lots yield different results according to outcomes chosen at purchase so how otherwise calculate accurate result for particular capital assigned? I don't cherrypick due to plan so sometimes 100% CGT but journal reminds allowed for so no use crying as net result which matters. Add to mix rebalancing portfolios & choosing which disposal so not to distort results of other Lots/investment decisions still awaiting their outcomes. No ATO choice is perfect so I advocate accuracy of result to know plan successful (or not).
Assigning capital in Lots is no different from those with variously future - dated F.T. Deposits who in essence make a gain on each assigned cash lot with no CGT so why is it wrong to choose a particular Lot's disposal when CGT IS paid at subjective rate? There's clues in word Portfolio----folio meaning numbered page as identifier reference, the very descriptive noun tells us its' contents are identifiable and moveable re use of 'port'---to carry/move.

Actually I list them separately and with dates attached because they were purchased separately and are individual assets in my portfolio.
The same would apply to purchases of gold. Same intrinsic metal, but different actual asset and purchase date.
I need to track them individually for their capital gains status but somehow you want to impose an ordering on my selling of them.
The ATO recognises them a separate individual assets purchased at separate dates.

It’s traditional and rather helpful for a blog to publish the post that is being responded to. Your post makes much less sense without my original question being published.

Hi Harry, I agree, but the website was set up with a limit on the number of comments on comments on comments in the same stream. We can look into changing this. Which is why I put your name at the start of my comment.

The level of reader disagreement with Graham’s thesis here, is surely reason enough for the ATO, the government and any industry advisory body to stay well clear of recommending similar. Remember, someone thought denial of dividend franking credits were a quick revenue raiser!

This may be oversimplistic to some, but the ATO classes every share as an individual asset since you can sell any one share on its own. Why is it therefore assumed that first in, first out should apply? I can choose whichever of my other assets to sell and when. Do I have to sell my oldest car first, just because I bought it first? If I make a profit on sale, it's also subject to CGT!

Hi Kim S, when you sell one of your cars, there is no tax implication, so the ATO does not care which you sell. In any case, the shares are all the same, the cars are all different.

If instead of a car it was one of a set of investment properties then Kim’s question is valid.

Well, if it's costing the treasury tax revenue, watch it dry up pretty soon then. In a scene like that out of Oliver Twist - people getting a lower tax bill or MORE of a tax refund ?! Can't be having that, can we ?! How dare we !

Over time the same amount of tax is paid as all shares eventually end up sold. Therefore, this would only be a one off benefit to tax revenue and not a lasting benefit. It is incorrect to say that it will raise billions of dollars a year. It will only do this for a few years as the change is introduced and then the amount of tax collected will be the same as it is now.

Hi Greg, your statement that "all shares eventually end up sold" is not correct. My main examples are managed funds, and most of them will last forever, and will always hold the majority of their stake in large banks, supermarkets and major miners. Think of the big LICs like AFIC and Argo, the big index funds, even the big active funds which hug the index. They will be compelled to hold as they grow because they are Australian equity funds and they must hold the largest companies because the ASX is not big enough for them to divest (total managed funds $4.5 trillion, total value of ASX $2.4 trillion). But when they sell at the margin as they trim or add to their positions, the tax system allows them to select the most expensive purchase cost to minimise tax.

Also, in Australia, when someone inherit shares, they are only liable for CGT if they sell them. They also might hold for the rest of their lives, and then pass them on.

Graham, I think that what you say here actually supports the argument against what you're saying. If shares are held for a long time and it's only at the margin that they are bought and sold (or 'traded') then surely the tax implications of those transactions should be confined to the shares that are in that margin. Shares bought 10 years ago should not be compelled to be exited if they're part of a core long term holding that's topped up from time to time by shares that are then traded out relatively quickly.

I agree with Warren's comment on this. Suppose you are offered a rights issue for a large company whose shares you have owned for many years, you take the offer up as it is at a discount to current price, then you sell those shares. Why should you then be deemed to have sold some of your original holding and not the ones you took up in the rights issue? If the ATO want to be prescriptive on this, then last in first out (LIFO) would make more sense to me.

Thank you very little for raising this issue Graham. It is one of the very few legal tax levers I have to pull in my very modest non- government pension retirement.

Hi Warren, the ATO allows a cherry-pick of cost bases to minimise tax based on a spurious notion about identifying the particular shares disposed of. No fund managers says, "I'm only selling $50 million of my Macquarie Bank shares because that's all that's in my trading account, and I can't sell those others because they are in my core account." But let's say for a minute that they did have two distinct accounts. Do you believe that when shares are sold, the fund manager and CFO would then select a different cost base depending on which pool they were in? "Oh, I sold from my core account so we must use FIFO." No, they would still select whichever cost base allows them to minimise tax. In reality, nobody records the fund manager's intentions to hold or trade at the time of purchase.

Hi James, on your point "Just curious Graham, why you so strongly advocate this? As one who admits to have made the most of past generous superannuation legislation and opportunity (your comments in a previous topic), now reduced for most, surely this stance is a little contradictory/perplexing?"

You're right that I have made the most of super concessions, but that does not mean I support every piece of tax legislation or government policy as it currently stands. I won't trawl through my hundreds of articles but here are a couple:

I don't believe someone with a $5 million home should be eligible for a full age pension, but should use a Home Equity Access Scheme to fund their retirement. There should be an eligibility cap.
https://www.firstlinks.com.au/10-reasons-wealthy-homeowners-not-receive-welfare

I don't believe superannuation is only for retirement.
https://www.firstlinks.com.au/that-horse-bolted-super-not-only-retirement

Thanks for replying. Looks like you're a bit on the "Pat Malone" side with your primacy of share sale suggestion though!

If the ATO's happy, just like sleeping tiger's or ex wives: leave them be!

Graham, I disagree with the assumption underlying your paper, that taxpayers choose sale parcels to mininimise capital gains and short-term tax: sometimes, they may do the opposite. I do agree that they recognise and use the rules to benefit their position.
In my SMSF, which is mostly in pension mode, I maximise current capital gains by selling low priced parcels because of the 0% tax rate on pensions, and so maximise the remaining cost base - ie minimise the CGT payable in the future by beneficiaries.
For investments outside super, I try to massage my taxable income to reflect my expected marginal tax rates. In 2022-2023, my income was lower than the previous year because of fewer takeovers that weren't under my control and because of lower "recovery" dividends. I expect my 2023-2024 income will be higher, resulting in a higher marginal tax rate. I chose to sell some CSL (at virtually 100% capital gain) in 2023 which brought that capital gain forward into a year with expected lower marginal tax rate for me.
Packer was right.

Hi John W, you are right that I use the example of minimising capital gains (maximising losses) to reduce tax, but the main point of my article is that taxpayers are given a range of cost bases to choose the tax position that best suits them to minimise tax. In your case, you have an opportunity to increase the remaining cost base, as you say, to minimise CGT in future. Fair enough, a la Packer, you are using the current rules. My point is that the ATO/legislation should not allow taxpayers to choose, but I realise any loss of flexibility for an 'investor' is unwelcome.

"My point is that the ATO/legislation should not allow taxpayers to choose,...."

Just curious Graham, why you so strongly advocate this?

As one who admits to have made the most of past generous superannuation legislation and opportunity (your comments in a previous topic), now reduced for most, surely this stance is a little contradictory/perplexing?

Surely it is not our job to advocate and suggest to government, when and where to tax us more. I doubt they need any more suggestions in that area! Most often they do not do what is right or fair but what they can electorally (self interest) get away with, rather than perhaps the wholesale, bottom up and top down reform that is really required!

Moreover, the first questions should be what areas to appropriately spend money on, how much and then ensure that the money spent is results based, not just thrown at areas with little or no improvement or accountability e.g billions and billions on education and indigenous support!

Thanks John W. My approach in my SMSF is the same. Seems odd going for the highest gain for the current year - but it makes sense.
I can't see how anyone could complain about choice - the ATO certainly differentiates between parcels of shares in its assessment - so why shouldn't the taxpayer?

I don't completely agree with the logic of arbitrarily enforcing FIFO. If you are minimizing your realized taxable gains then you are also maximizing your unrealized gains. Eventually, in time, that gain will also be realized. What's the average holding period for a stock in Australia these days? I'm guessing it's not 'infinity' and most holdings are eventually fully sold down. If markets trend upwards over time you could also argue that allowing people to minimize their tax in the short term could actually increase it in the long term. Maybe I'm wrong but It would take a lot of modelling to prove that parcel selection costs the Govt tax revenue.

I just let my software auto select which model generates the lowest tax and use that. My understanding is you can use any method you like for each transaction. Hope I am right, nobody take this as advice. Having said that I agree with the theme of the article that tax should be as simple as possible. Each complication has a cost.

Hopefully this is not a "thought bubble" policy which the Government is currently working on behind the scenes as part of possible changes to capital gains tax regime.

We already have the intended taxing of unrealised gains on assets in SMSF's which has a number of obvious flaws but will bring forward tax paid, and opens up the potential for similar application to assets outside super. In an earlier article you discussed changes to the 50% discount of gains and replacing this with indexation which may have some merits but will again bring forward revenue collection. And FIFO is also now being floated as a means of bringing forward tax paid. The other common thread, unfortunately, is loss of simplicity and additional administration cost.

The article primarily uses listed shares as an example although units are mentioned? Is the intent of the proposal to primarily create a new capital gains tax regime for listed share assets only? If so is this equitable?
If not, there will be different implications for other asset classes which have different characteristics. (a) A holding in a bond held in a custody account and accumulated through multiple purchases. Is it being proosed that FIFO should apply noting the initial purchase price will be a function of the remaining duration and interest rate expectations at that time, and unlike shares the redemption value and capital gain/loss if held is known at the time of purchase? (b) Holdings in unlisted entities and family companies which may also have different share classes? (c) Physical assets such as gold bullion accumulated over time (d) Multiple units in the same building purchased at different times

There will be predictable administrative difficulties within SMSF's with segregated assets (which is appropriate choice when considering the different risk profiles of older and younger members). How will FIFO be administered ? Will it apply to the assets of each member or the fund as a whole? What happens when different members "own" part of segregated asset purchased in one transaction, and only one member wishes to divest their holding or decides to exit the fund?

If an entity/person has incurred a capital gain and they wish to balance this out in a year with a loss, under FIFO they would be required to sell the initially purchased part of another holding where they may have a small gain or loss in addition to selling the holding purchased at a later date at a higher price which has incurred a larger loss. What is the benefit to either to tax revenue or the person?

If a person has accounts with multiple brokers in addition to direct holdings. Will they need to sell down the last package purchased? If not, will this be considered avoidance?

Your contact is not well informed about the super and managed fund industry. Any RSE or RE allowing this practice would subject themselves to serious fiduciary breaches under SIS and Corps Law. They must treat all members/investors equally and this includes the tax treatment embedded into the unit prices/crediting rates. As scuh any public offer trust not using FIFO would be the exception IMO. As such this potentially adds merit to your argument, however, I personally prefer direct investing and the Kerry Packer approach to the ATO.

Completely disagree with Graham.
Just because shares purchased at different times are the same as each other, does not mean they merge to become one amorphous block of shares.
Buy shares over several days and see how the registry records each purchase separately on the Holding Statement.
Why are shares different from (say) bars of gold, loads of coal and the like where one can demonstrate that the one has been sold and the other kept?

Shares are, properly, bought on a voluntary basis. Those shares have a purchase price and a purchase date. It is also proper for the shareholder to choose the sell price and the date of disposal on a voluntary basis.

If you can’t identify individual parcels, how do you identify the first?

To Jim. Parcels SHOULD be easily identified by both the cost price and the date.
If you have bought four parcels of the same share, it’s highly unlikely they would have all been bought on the same day and at an identical price. In this case, which parcel is sold first is immaterial. One could easily identify the date and cost prices of each parcel by referring to the confirmation note records issued by your broker.

Exactly. My point is that it is inconsistent to claim (a) that all the shares are the same and (b) that the first-bought shares are somehow special and should be sold first.

I’m curious as to how this behaviour affects investors in industry/retail funds. It seems to me that the current generation of retirees benefit from selling the shares which have just been purchased and thus paying minimal tax, but the coming generations are stuck with the much lower cost base and thus higher CGT of shares purchased decades ago.

Is this correct? And if so how does it meet best interest duty for the super funds?

Graham, I respectfully disagree. First, your logic that all shares are the same leads to the conclusion that some sort of average of the purchase price of all shares held is used as the cost base, not that its first in first out as you say in the article. An averaging approach treats all shares acquired over time as going into a pool and all sales as being from that pool, with a proportion of all individual transactions over the years being sold. BUT I still don't think that's right. Investors may perfectly legitimately hold parcels of shares for different investment reasons over different time periods. One example. I hold CBA shares allocated to me over the years I worked for Colonial. I intend never to sell those shares. However, if I took a view today that CBA shares were at a good price with a bit of upside over the next couple of years and bought some more, my intention may well be to sell them in 2025. What you are advocating is that I'd have to be deemed as selling a portion of my employee allocation from years ago which would trigger a much larger capital gain than my actual investment strategy warrants. What you propose - or even an averaging approach - simply discourages me from ever buying more CBA shares because eventual sale of those shares would wrongly and unfairly trigger a capital gain and a tax payment that isn't the result of my investment strategy. I think the ATO is doing exactly the right thing in allowing investors to determine the parcel of shares they have sold. Sure it mathematically looks like it costs the taxpayer, but good tax policy requires an alignment with financial reality, including not creating an unintended* distortion to financial decisions; and fairness to all tax payers. * footnote tax policy can be used to create incentives, or distortions, with superannuation being the prime example. But capital gains tax policy is not intended to distort decisions about what to buy and sell so should be structured in a way that avoids doing so.

Hi Warren, appreciate your position, but when you say "Investors may perfectly legitimately hold parcels of shares for different investment reasons over different time periods", perhaps there's a case for individual investors with some sort of bucket approach, but I'm guessing that's a small minority. But large funds (super and non-super) hold shares in companies purchased at many different times, maybe decades apart, and they are not in buckets for lifestyle or spending or strategy reasons. They are all part of one portfolio. But the standard industry software (HiPortfolio) is set to calculate the least tax, or a loss. I don't see why such a generous treatment should apply to every taxpayer simply because a few think of their shares in different buckets.

Graham, that's a reasonable response, though I don't know how the ATO can possibly distinguish the intentions of different groups of investors.

I'm not sure that a lot of your readers would see it the same way (ie that holding shares in long term versus medium term buckets is a small minority). Lots of individual investors have a long term buy and hold strategy overlaid with something that has a shorter timeframe. Perhaps this is a survey question you could ask!

Also, the reason that Hi Port is set up that way is that most holdings in managed funds are actually kept for a long time. Active management is at the margin. When someone goes overweight a stock, it's not necessarily a long term hold but will be sold when a short term profit is realised. Those shares should be able to be identified as the source of realised gains rather than shares bought 10 years ago to obtain the core holding.

In my early years at CFS I found the opposite issue was also true. Bonds that had been a few years before that were out of money were going to realise capital losses if sold, which would impair the distributable income we could pay. So we wanted to make sure that it was the more recently purchased securities that were not out of the money that were, in an accounting sense, sold. We did the ATO a favour because we maximised the income payments made to our investors rather than diluting those payments by realising losses.

I'm probably unique in that way too. (When I started at CFS we only had $28 million in bond funds and we eventually grew that to around $30 billion. Not a lot of investors have had to deal with that sort of growth in funds!!! I was blessed to be part of the greatest corporate ride in funds management, I know.)

But my point is that this is a more complex situation than meets they eye. I wouldn't be as critical of the ATO and the industry as your article has been.

Australia must have the perfect tax system or at least no taxes which should be adjusted to be less favourable to investors. Any suggestions to reduce the benefits are always criticised in these pages.

Hmmmm. I wonder if the author has recently discovered he does not hold good records of his own shareholdings, and has perhaps missed an opportunity to legitimately manage his tax upon their sale in prior years? If you are willing to use a decent investment advisory firm, they will use a platform that holds tax records on a parcel-by-parcel basis, enabling sensible tax management. If you own 5 investment properties, you may elect to sell the one that realises a different tax profile to the others. What is the difference? Oh yes, that's right - most journalists hate equities; they think we should only invest in 'property and cash'...

I suspect the only thing James and I agree on is that a) governments waste money, and b) taxing unrealised gains in super is not a good idea.

Many areas of choice in the system are to allow for options to fit the circumstances or available record keeping, with a consequence that some are more favourable to taxpayers. This one is a clear opportunity to simplify the accounting and improve revenue collection.

Interesting to note that funds that have to value holdings to the market or businesses that change cost base of assets have to show the impact in the P&L, which at least closes the loop on the cost base adjustment.

I find Graham’s view that the first purchase should be the cost base ie CGT should be based on FIFO hard to understand. He offers no logic for that view. Unlike perishable inventory, shares don’t go off so there is no reason to mandate FIFO. The fact that, in rising markets, it will maximize tax revenue is a very poor argument. My view is that the share owner should reasonably be able to make the choice but, if the law must mandate a treatment, LIFO has more logic. Alternatively, average cost avoids arguments and minimizing record keeping demands.

Agree with James's sentiments. If an investor has choice as to when they purchase an asset, they should have choice as to the timing of the sale. Because they are the same asset, in the same class, doesn't mean the parcels are merged into one. Each parcel is distinctly identifiable. The averaging option is probably allowed as record keeping has been poor. With the digitisation of share trading and the registries etc, this method should be eventually fazed out. In the meantime, anyone that works with older client's will attest to the incidence of portfolio cost bases being not identified until the person passes away and the assets have to be dealt with. The cost to reconstruct is not in proportion to the tax leakage if the executor is able to use the averaging method.

Capital losses are applied before CGT discount. For example 3m capital loss is applied against 3m of gross capital gains before discount so saving 23.5% cash tax if on top marginal tax rate, or $700k in your example. Also, whilst FIFO may be more appropriate from a policy perspective in this example, it is not appropriate in many circumstances. For example, what if the same taxpayer has a long term investment portfolio and also a short term trading portfolio - here LIFO or discretionary is likely more justified to match the gains/losses in the trading portfolio. Similarly this same principle applies to other CGT assets such as crypto where nearly every dealing is a gain/loss and likely has similar mixed purposes to the previous example but with many more taxing points/trades such that LIFO is likely the most appropriate. For example in the space of a few minutes, AUD transferred to exchange to buy ETH (taxable), transferred to a cold wallet that holds other ETH (not taxable) and then swapped to another token say stETH/wstETH (likely taxable) - it is the cost base of the ETH just transferred in that should be relevant to the gain/loss on the last swap (small as only held briefly) not the cost base of the earliest unspent ETH in that wallet.

It is no scoop that harvesting of the highest cost price shares to calculate CG is widespread - the word epidemic is an emotive word thrown in for rhetorical effect. Perhaps it's is not understood by the author that shares belong to the owner - not the ATO, Treasury or anyone else. It is quite legitimate - and proper - for an owner to choose which shares one wishes to sell. And end of FY tax management is also a legitimate exercise if one accepts the principle that it is legitimate to minimise tax liability. If this principle is not accepted, then the corollary is that one must maximise tax liability - and that would be a 100% tax rate. And the author also forgets that, sooner or later, shares will have to be sold, so hanging on to lower cost price shares delays the inevitable payment of, in all likelihood, more CG tax - CG tax is not avoided.

Your last point is spot-on Nick - CGT is not avoided, it is merely deferred. In the end, when the whole parcel has been sold in multiple transactions (with the final sale possibly being upon death), every share in the holding has had CGT calculated on it.

100% agree with all of James' comments above

"The tax lost by not adopting FIFO is billions of dollars a year."

So? Given that the amount of tax lost by misplaced, often stupid government policies, boondoggles and excesses with no accountability or link to outcomes, this pales into insignificance. How much of GDP revenue supports a bloated public service at council, state and federal level for a small population of 26 million?

Why take this discretion away from investors? Australia is becoming more socialist by the day, why promote further changes removing choice, to further it?

As for the notion of taxing unrealised gains on superannuation accounts, albeit over a largish threshold, who's to say this precedent won't spread to other areas. Is this policy fair?

There are many unfair rules in the tax system, but choice of shares sold for efficiency of capital gains treatment, is not one of them!

James, I am reminded of Kerry Packers famous response to the committee who were essentially accusing him of tax “avoidance”;
“ I don't know anybody that doesn't minimise their tax," Mr Packer growled as he stirred his delicate parliamentary china cup of tea with a teaspoon. "I'm not evading tax in any way shape or form. Of course I'm minimising my tax. If anybody in this country doesn't minimise their tax they want their head read. As a government I can tell you you're not spending it that well that we should be paying extra."
Here Here !