I have watched investors buy beautifully and end up nowhere.

The property checked out. A good street, a sound building, a tenant who paid on time, a price that looked sharp on the day. Two years on, the same investor was stuck. The bank would not lend them another dollar. The land tax bill had crept up in a way nobody mentioned at the time. The one asset they owned sat in the wrong name, in the wrong corner of the city, to do anything useful next. The purchase was good. The position was not.

That gap is the reason this practice exists. A single property, judged on its own terms, can be excellent and still be the wrong thing to own. It depends entirely on what sits around it, what came before it, and what you intend to do after it. A property is a thing. A portfolio is a plan. Most people are sold the first and told they have bought the second.

Here is the distinction in plain terms. When you assess a property in isolation, you ask a short list of questions. Is it well built. Is the area sound. Will it rent. Is the price fair. All sensible questions, and all incomplete. When you assess the same property as one move inside a portfolio, the list changes. What does this asset do that my existing holdings do not. Does it concentrate my risk or spread it. After this purchase, can I still borrow, and at what point does the next acquisition become possible. How does it sit for land tax once it is added to everything else I hold. In whose name, and in what structure, should it be bought so that the eventual sale does not hand back years of gains. When, and to whom, will I sell it, and is that buyer likely to exist in ten years. The property has not changed. The questions have. The second set is the one that builds wealth, and it is the one almost nobody asks before they sign.

The same property, two sets of questions

Judged on its own

• Is it well built?
• Is the area sound?
• Will it rent?
• Is the price fair?

Judged as one move in a portfolio

• What does it do that my holdings do not?
• Does it concentrate risk, or spread it?
• Can I still borrow after this?
• What does it do to land tax across everything I hold?
• Whose name, and in what structure?
• When do I sell, and to whom?

Two forces have made this more urgent than it was even a year ago, and both arrived in the space of a fortnight.

The first is the May budget. From the first of July 2027, negative gearing on established residential property bought after budget night is being withdrawn. Rental losses will be quarantined against rental income and carried forward, rather than written off against your salary. In the same package, the 50% capital gains tax discount is being replaced for individuals, trusts and partnerships with cost base indexation and a 30% minimum tax rate. New builds keep their concessions. Commercial property is untouched by the gearing change, though the new capital gains treatment applies across asset classes. I am not here to argue the politics of it. I am here to point out what it does to a certain kind of investor. For two decades, a particular logic underpinned a great deal of Australian property buying. Buy one good house, let it lose a little money each year, claim the loss against your income, and wait for the capital growth and the tax discount to do the heavy lifting at the end. That logic rested on a single asset carrying a single bet. Take away the annual deduction and soften the prize at the finish line, and the standalone purchase loses the very scaffolding that made it stand up.

The second force is the cost of money. The Reserve Bank has raised the cash rate three times this year, to 4.35%, and serviceability is now being assessed above 9%. Sydney's gross residential yield sits at about 3.1%, the lowest of any capital city, while values across the upper end of the market have slipped for months. An investor carrying one negatively geared house, bought for the tax and the hope of growth, is now holding an asset that costs more to fund, returns little in income, and, if it was bought after budget night, no longer carries the concessions that made the shortfall bearable.

I want to be careful here, because this is the moment people panic, and panic is expensive. None of this means property is finished. Sydney's vacancy rate is 1.1%. Rents on houses have risen close to 6% over the year. The city is still short of housing, still growing through migration, and the long-run fundamentals that drove decades of growth have not changed. What has changed is that the easy version of the strategy, the one-property-and-a-tax-return version, has stopped working. The considered version has not.

The considered version is a portfolio. Not more properties for their own sake. I have no interest in counting doors. A portfolio means a small number of assets chosen so that each one does a job the others cannot. One might be there for income that you can bank today. Another for growth over a ten-year horizon. A third might sit in a structure that protects the others and keeps the next purchase within reach. The income from one funds the holding cost of another. The equity in one becomes the deposit for the next, at a time of your choosing rather than the market's. The whole thing is assembled so that it survives a year like this one, where rates rise and the rules shift, without forcing a sale at the worst possible moment.

This is the part that gets lost. People imagine portfolio investing is about ambition, about scale, about owning as much as the bank will allow. It is closer to the opposite. It is about restraint, and sequence, and knowing what each piece is for. The best single purchase I could find you this month might still be the wrong one, not because the property is poor, but because it does not fit. It uses up your borrowing capacity on the wrong asset. It concentrates you where you are already exposed. It looks like progress and quietly closes a door.

The question I ask a new client is never “what do you want to buy”. It is “what do you want to own in ten years, and what has to be true along the way for you to get there”. Once that is on the table, a great many properties that looked attractive fall away, and the right ones become obvious. The purchase stops being the event. It becomes a move inside something larger, judged on the role it plays.

A property is a thing you can fall in love with at an inspection. A portfolio is something you build on purpose, over years, with a clear idea of where it ends. I am far more interested in the second. So, I suspect, are you.

There is a structural problem at the heart of how most buyer's agents are paid, and it is worth saying out loud.

Most of them only earn when you buy. The fee is tied to the purchase. No purchase, no fee. Sit with that for a moment and the consequence is obvious. The person advising you on whether to spend a million dollars has a direct financial interest in your spending it. Not a hidden one. Not a sinister one. Simply a built-in pull in one direction, present in every conversation, whether anyone names it or not. It is very hard to be told to wait by someone who does not get paid until you go.

I built this practice to sit on the other side of that problem. Clients engage me for a relationship that runs across years and across a portfolio, not for a single transaction. That changes what I am free to tell you. I am paid for my judgement over time, which means the most valuable sentence I can offer is sometimes the one that costs me a deal this month. Do not buy this. Or, more often, not yet.

I keep track of how often I say it. Across a year, I recommend against buying in close to a third of the situations I am brought into. Sometimes the property is wrong. More often the property is fine and the timing is wrong, or the structure is wrong, or the client's position cannot carry it without putting everything else at risk. I publish that number on purpose. It is the cleanest evidence I can give that the advice is honest, because a number that could embarrass me in a quiet year is not a number you fake.

Recommendations against buying, across a year

A third

of the situations I am brought into end in advice not to buy, or not yet. I publish the number because a figure that could embarrass me in a quiet year is not one you fake.

This year has made the discipline matter more than usual. The temptation right now is to read the headlines and rush. Sydney values have softened. The upper end of the market has slipped for months. There is a story going around that this is the dip, that the budget changes and the rate rises have scared everyone off, and that the clever move is to buy while others are nervous. Some of that is even true. Buying when sentiment is poor has made a great many investors wealthy. But the headline hides the trap.

Borrowing power has fallen faster than prices. With the cash rate at 4.35% and lenders assessing you above 9% under the unchanged serviceability buffer, the amount you can responsibly borrow has dropped by more than the discount you might win on the purchase price. A property that looks twenty or thirty thousand dollars cheaper than it was last spring is not a bargain if funding it leaves you unable to move for three years, or unable to absorb another rate rise without selling. The price tag fell. The real cost of carrying it rose. Saying no, in that situation, is not caution for its own sake. It is arithmetic.

The budget sharpened the same point. From the first of July 2027, an established residential property bought after budget night no longer lets you offset rental losses against your wages, and the capital gains discount that softened the eventual sale is being replaced. For an investor who was relying on the tax position to make a thin deal work, the deal may simply no longer work. The honest answer is to say so before contracts are exchanged, not after. That is uncomfortable to deliver when a client has fallen for a place and wants to be told it is fine. It is also the entire value of having someone in the room who is not paid to agree with you.

I should be clear about what “no” actually means, because it is rarely a flat refusal. Usually it is a redirection. Not this property, but here is the kind that would fit. Not this quarter, but here is what needs to be true before you move, and here is how we get there. Not residential at all this time, perhaps, but a neighbourhood commercial asset that the budget's gearing changes left alone and that yields well above anything residential can offer right now. A good “no” almost always comes with a better “instead”. The aim is never to stop you investing. The aim is to stop you investing badly, which is a different and far more expensive thing.

There is a quieter reason this matters, and it has to do with how trust compounds. The first time I talk a client out of a purchase, they are usually a little deflated. The second time, they relax. By the third, something has shifted. They stop bracing for a sales pitch and start treating the conversation as what it is, which is counsel. From that point on, every yes I give carries more weight, precisely because they have watched me say no. A recommendation to buy means nothing from someone who has never recommended against it. It means a great deal from someone who walks away from a third of the deals they see.

People sometimes ask whether it is awkward to charge for advice that ends in “do nothing”. It is the reverse. The work that goes into a confident “no” is often greater than the work behind a yes. It takes a full read of your position, the numbers run properly, the comparable sales checked, the structure and the tax and the serviceability tested against everything else you hold. The output is a single word, but the word is earned. You are not paying me to find you a property. You are paying me to tell you the truth about one, including the truth you were hoping not to hear.

The best thing I can do for a portfolio is sometimes to keep something out of it. I would rather lose a fee this month than lose your confidence over a decade. That is not a noble pose. It is simply the only model that holds up when the relationship is meant to last.

Almost everything written for property investors is about buying. The next suburb. The next deal. The next door.

Very little is written about the moment that matters just as much and arrives for every successful investor eventually, which is the moment you stop.

You have built the thing. The properties are bought, the structure is in place, the plan you set out with has largely been carried out. The instinct, drummed in by an entire industry, is to keep going. Buy another. Always be acquiring. I want to make the case for the opposite, because the years after you finish building are where portfolios are quietly won or lost, and where the work is least understood.

Call it playing defence. In the accumulation years, you are on offence. You are extending, borrowing, adding, taking calculated risk to grow the asset base. Defence is the discipline of protecting what you have built so that it actually delivers what you built it for, which was never the properties themselves. It was the income, the security, and the eventual freedom they were meant to produce. A portfolio that keeps expanding past the point of purpose is not ambitious. It is unfinished for the sake of being unfinished.

Two phases, two disciplines

The building years

• Extend, borrow, add
• Debt as leverage
• Growth in the asset base
• Measured by what you acquire

The years after

• Protect, tune, occasionally prune
• Debt as the thing to structure down
• Income, buffers and structure
• Measured by what it delivers

The defensive years have their own work, and it is substantial. Debt is the first piece. In the building phase, debt is your friend, the lever that lets a deposit control a much larger asset. In the holding phase, with the cash rate at 4.35% and the cost of money higher than it has been in years, debt becomes the thing to manage down and structure carefully. Which loans to pay off first. Which to keep interest-only and which to switch. How much of a buffer to hold against the next rate move, given the Reserve Bank has lifted three times this year and is not obviously finished. None of this is glamorous. All of it decides whether a portfolio that looks impressive on paper actually leaves you better off.

Then there is the question the May budget forced onto every established investor in the country, which is what not to sell. From the first of July 2027, the 50% capital gains discount is being replaced for individuals, trusts and partnerships with indexation and a 30% minimum tax rate. The design matters. Gains you have already accrued by that date keep the discount treatment when you eventually sell; it is the growth that comes afterwards that falls under the new rules. Read that carefully and two things follow. There is no cliff, and no reason to rush an asset to market this year just to beat a date, because what you have already earned is protected whenever you sell. And the tax concession on future growth is smaller than it was, which means the case for holding any asset now rests more nakedly on the asset itself: its income, its prospects, and the role it plays. For a great many investors the core of the portfolio will still deserve holding. But it should be held because you have decided to, not because you never looked.

Defence is also about the assets themselves earning their keep. Leases reviewed and rents brought to market, particularly with Sydney's vacancy rate at 1.1% and house rents up close to 6% over the year, so that you are not quietly subsidising a tenant out of habit. Commercial holdings, if you have them, managed against their lease expiries and their tenants' health, because a vacant shopfront costs far more than an empty flat. Insurance and structure checked against a life that has moved on since you set them up. Land tax modelled across the whole holding rather than discovered, unpleasantly, when the assessment arrives. This is unglamorous, ongoing, attentive work. It is also where the returns of the building years are either banked or frittered away.

I will say the harder thing too, because defence sometimes means subtraction. Not every property earns its place forever. An asset that has done its job, that no longer fits the plan, that ties up equity or attention better used elsewhere, can be the right thing to release even with the new tax position factored in. Defence is not the same as never selling. It is selling on purpose, for a reason that serves the whole, rather than holding everything out of inertia or selling something good out of fear. The point is that the decision is made inside a plan, not in reaction to a headline.

This is, frankly, where an ongoing relationship earns its keep in a way a one-off purchase never can. The buyer's agent who found you a property at a transaction five years ago is long gone. The questions that matter now are not about finding the next deal. They are about whether the portfolio you have is structured to ride out a higher-rate decade, whether the income is doing what you need it to, whether the holding costs are under control, and whether the plan you wrote at the start still describes the life you actually want. Those questions do not have an answer at settlement. They have an answer that changes every year, which is exactly why I stay in the room for years.

There is a quiet satisfaction in this phase that the accumulation years rarely offer. The striving is largely done. The job becomes stewardship, which is a more demanding discipline than acquisition and a far less celebrated one. A portfolio you have finished building is not a finished thing to be left alone. It is a living position to be defended, tuned, and occasionally pruned, so that the years of careful buying actually become the income and the security they were always meant to be.

Knowing when to stop building is its own kind of skill. Knowing what to do next is the one almost nobody teaches. It is the part I find most worth doing.

I studied psychology before I ever worked in property. At the time it felt like a detour.

Twenty years on, across sales, management, and buying, I think it was the most useful training I could have had, because I have used it every working day since and I have barely opened a valuation textbook.

Property is presented as a numbers exercise. Yields, comparable sales, growth rates, serviceability. The industry produces spreadsheets because spreadsheets look like the way decisions get made. They are not. I have sat on every side of the table, for vendors, for landlords, and now for buyers, and I could count on one hand the transactions I have seen decided purely on the numbers. Property decisions are made by people, under pressure, with incomplete information and a great deal of feeling. Buying well is not the removal of all that. It is the craft of knowing exactly where the feeling sits, whose it is, and what it is doing to the price.

There are three people to read in any purchase: the vendor, the agent, and yourself. Most buyers study none of them.

Three reads in any purchase

1. 01The vendor. The number already living in their head, where it came from, and what the money means to them beyond its size.
2. 02The agent. A designed environment: the guide as anchor, the open home as social proof, the auction as choreography.
3. 03Yourself. The gap between the brief you wrote and what you reach for under fatigue, competition and attachment.

Start with the vendor, because the vendor is the least visible and the most important. Every property arrives on the market with a number already living in its owner's head. Sometimes it is what they paid plus what they spent. Sometimes it is what the house two doors up sold for in the autumn of 2021. Sometimes it is the figure their retirement plan quietly requires. Psychologists call the pattern underneath this loss aversion, and it is one of the most reliable findings in the discipline: a loss hurts roughly twice as much as the equivalent gain feels good. A vendor who has mentally banked a price will experience any offer below it as a loss, and will hold out against arithmetic for months to avoid it. They are not being greedy. They are protecting a number they have already spent in their imagination. Once you work out which number is in the vendor's head, and where it came from, you know whether a campaign will move, stall, or quietly collapse.

You also know what an offer is worth beyond its dollar figure. A vendor bridging between two properties values certainty and a short settlement. An executor values simplicity and an end to the file. A separating couple values a date. The strongest offer I negotiate is often not the largest one my client could have made. It is the one shaped to what the money means to the person receiving it. That information is never in the listing. It is in the pauses on the phone, the timing of the campaign, the state of the garden, and a hundred small signals you learn to read by having stood in a lot of hallways.

The agent is the second read. The selling agent is not your enemy, but they are not working for you either, and they are running a designed environment. The price guide is an anchor, set with care, and every offer that follows is measured against it rather than against the evidence. The Saturday open home is social proof made physical; the crowd is part of the marketing, and it works on you whether you believe in it or not. The auction is the most carefully engineered decision environment in Australian commerce: a public arena, a clock, an audience, and bidding increments sized so that another twenty-five thousand dollars feels like one more small step rather than a material sum. None of this is underhanded. It is the vendor's agent doing their job well, and I admire good campaign craft when I see it.

The defence is not cynicism. It is preparation. You walk in carrying your own anchor, built from comparable sales you have actually inspected rather than a guide someone chose for effect. You know what the property is worth to your strategy before anyone tells you what it is worth to the market. And you set the walk-away number in a quiet room, days before the heat, because the worst place to make a decision is inside an environment engineered to stop you making one.

The hardest read is the third one. Yourself, or in my work, the client. People are honest about their goals and unreliable about their behaviour. Investors tell me they want yield and then fall for the pretty house with the terrible lease profile. They tell me they are patient and then ring me after a fortnight of quiet feeling that they are missing out. The gap between what people say they want and what they reach for under pressure is where most buying mistakes live, and no spreadsheet closes it.

Two moments deserve particular respect. The first is the inspection where a property becomes theirs in their head before it is theirs on paper. From that point, every fault becomes forgivable and every comparable sale that disagrees becomes irrelevant. The second is fatigue. A couple who has searched for a year stops trying to buy well and starts trying to be finished. Fatigue is the most expensive emotion in property. It does not feel like recklessness. It feels like decisiveness, which is what makes it dangerous.

The discipline that answers all of this is unglamorous: make the decision before the moment. Every client of mine has a written brief, and later a written thesis for each acquisition, set down when we are clear-headed and no property is on the table. The role the asset must play, the numbers it must meet, the price above which it stops making sense. The document is not bureaucracy. It is a decision made early, when you were calm, standing guard over you later, when you are not. On auction day, the thinking is already done. All that remains is the execution, and execution is easy when the judgement has been separated from the adrenaline.

People sometimes assume that a psychological approach to buying means being coldly unemotional. It is closer to the reverse. Emotion is information. The vendor's attachment tells you where the price will crack. The agent's choreography tells you what the campaign fears. Your own restlessness tells you when you have stopped assessing and started wanting. Buying well is not refusing to feel any of it. It is refusing to let any of it sign the contract.

The numbers matter, and mine are checked to the decimal point. But twenty years of watching property change hands has settled the question for me. The spreadsheet describes the deal. People decide it.

Ask an investor why they bought their last property and you will usually get a description, not a reason.

Three bedrooms. North facing. Four hundred metres to the station. Good school catchment. All of it description. The reason, the actual investment case, the job the asset was hired to do and the terms on which it would be judged, very often never existed. The purchase felt right, the market moved, and the story was written afterwards to fit.

I run my practice on the opposite discipline, and it is the closest thing Valentine Investment Partners has to a founding rule. No acquisition without a thesis. Before contracts are exchanged, every property I buy for a client is recorded against a written thesis, about a page long, that states five things in plain language.

First, the role. What job this asset does inside the portfolio that the existing holdings cannot do. Income to be banked now. Growth on a ten-year horizon. Land exposure the portfolio lacks. A yield engine to carry the holding costs of something else. If I cannot name the role in a sentence, we do not proceed.

Second, the income today. The rent as a number, not an estimate dressed as one, and the net position after the costs that investors prefer not to total.

Third, the growth case. Not “the area is going well”, but the specific things the ten-year case rests on: the land component, the depth of tenant and buyer demand, the infrastructure that is funded rather than promised, the planning settings and where they are heading. Things that can be checked later, by someone other than me.

Fourth, the structure. Whose name, what entity, and why, with the land tax and capital gains consequences worked through before the purchase rather than discovered at the first assessment after it. The same property held in the wrong structure can be a materially worse investment, and the time to know that is before exchange.

Fifth, the exit. When we expect to sell, to whom the asset is likely to appeal at that point, and what would make us sell earlier. An asset you cannot describe selling is an asset you have not finished thinking about.

The thesis, on one page, before contracts

1. 01The role. The job this asset does that the existing holdings cannot.
2. 02The income today. The rent as a number, net of the costs investors prefer not to total.
3. 03The growth case. The specific things the ten-year case rests on, checkable later by someone else.
4. 04The structure. Whose name, what entity, and the land tax and capital gains consequences worked through first.
5. 05The exit. When we expect to sell, to whom, and what would make us sell earlier.

The thesis is the contract between my judgement and the client's capital. That is the sentence I use with new clients, and I mean it precisely. It binds me as much as it guides them. Every year we mark the portfolio against the theses on file. Is the rent where the thesis said it would be. Is the growth case intact, strengthening, or eroding. Has the structure done its work. When I am right, the record shows it. When I am wrong, the record shows that too, and we deal with it early, while the correction is cheap.

The written part matters more than people expect, and the reason is human rather than administrative. Memory is a flattering historian. Left to ourselves, we all remember our reasons as better than they were, and a purchase that drifted sideways gets quietly reclassified as “a long-term hold”. A thesis on paper cannot be reclassified. It sits there, dated, in plain language, and it can be judged. Most of the industry would find that uncomfortable. I think it is the entire point. Accountability is what separates a portfolio practice from a person who sells you houses.

A thesis also changes the buying itself, because it changes the question. The question stops being “is this a good property”, which is the question the whole industry is built to answer, and becomes “what does this do that the portfolio cannot already do”. Those questions produce different shortlists. A second unit, nearly identical to the first, two suburbs away, can pass the first question easily and fail the second completely. It is concentration dressed up as progress. The same tenant profile, the same risks, the same corner of the land tax ledger, bought twice. Nothing about the property is wrong. Its role is already taken.

Sequence matters just as much as selection, and this is the part almost nobody plans. Borrowing capacity is finite. Land tax aggregates across holdings. Equity builds on its own timetable. Every purchase spends capacity that the next purchase needed, which means a portfolio is not a collection of good decisions. It is an ordered series, and a strong asset bought at the wrong point in the order can close more doors than it opens. Part of each thesis is what the purchase does to the next one: what it does to serviceability, where it leaves the buffers, and when the portfolio should be ready to move again.

That word, ready, deserves care. As equity and income build, there are identifiable moments when the next acquisition becomes possible: the serviceability is there, the buffer is intact, the structure has room. I track my clients' portfolios for those moments, and I tell them when one is approaching. What I will not do is dress that up as market timing. I cannot tell you the right month to buy. Nobody can, and the ones who claim to are selling something. What I can tell you, with evidence, is when you will be ready to buy, and there is more value in that sentence than in any forecast I have ever read. Readiness is about the investor. Timing is about the market. Only one of them can actually be known.

This discipline has always been sound. This year it stopped being optional. The May budget stripped away the tax scaffolding that used to rescue the undocumented purchase, and the cost of money has made every improvised decision more expensive to carry. An investor could once buy on instinct and let negative gearing and a rising market tidy up the reasoning. That era is closing. What replaces it rewards exactly one thing: knowing why you own what you own, in writing, before you own it.

A property judged on its own terms answers a small question. A portfolio with a thesis answers the only question that matters, which is what all of this is for. The house is not the point. The station and the school catchment are not the point. The point is the position you are building across a decade, and nothing you buy means anything until you can say, on paper, what it is for.

I spent years being paid to live with other people's buying decisions.

That is what property management is, once you strip away the software and the inspection reports. The buyer gets the Saturday. The manager gets every Tuesday afterwards: the hot water system that fails in the first winter, the tenant who leaves after eight months, the strata meeting where the special levy lands. Running rent rolls, and later co-founding a property management business, taught me more about what makes an investment property perform than anything I learnt in a sales office. It is the reason I buy the way I buy.

Properties are bought on Saturdays and owned on Tuesdays. On Saturday the property is staged, lit, and holding flowers, with an agent to narrate it. On Tuesday it is a ledger. The distance between Saturday's story and Tuesday's ledger is where investment returns quietly die, and the only people who see that distance clearly are the ones who manage the years after settlement. Very little of what they know makes it back to the moment of purchase. It should. Here is some of it.

Tenants and buyers pay for different things. Buyers pay for the renovated facade, the styling, the street theatre of a good campaign. Tenants pay for light, storage, parking, a second bathroom, quiet, and a landlord who fixes things quickly. The two lists overlap less than you would hope. A rent roll shows you the result in plain numbers: the striking terrace that re-lets slowly at a soft rent because it is dark and cold, three doors from the unremarkable house that never sits vacant for a fortnight because it is warm, practical, and near the school gate. An investor who buys on buyer preferences funds the difference between the two lists every year they hold the property.

Vacancy is arithmetic, not weather. Every week a property sits empty costs close to 2% of the year's rent. Three weeks of vacancy and a re-letting fee together can take 7 or 8% of the annual income, which for most residential investors is the entire difference between a good year and a poor one. Operators know that vacancy is not bad luck. It is predictable, street by street and building by building, and it clusters around the same features every time: the wrong floor plan for the area's tenants, the dark aspect, the awkward access, the rent set by pride rather than the market. When I assess a property now, the first question is not what it will rent for. It is how fast it re-lets in a soft month, because that number is the one that compounds.

What an empty week actually costs

2%

of the year’s rent, for every week the property sits empty.

7–8%

of annual income gone once three weeks of vacancy and a re-letting fee are counted together: for most residential investors, the whole difference between a good year and a poor one.

The ledger lives behind the paint. Any building can be made beautiful for a campaign. What it cannot hide is its minutes. Strata records read like financial statements once you know the dialect: the waterproofing report that keeps being deferred, the lift nearing the end of its service life, the original plumbing behind the new vanities, the capital works fund that is far too thin for the age of the building. Two blocks on the same street can ask the same price and sit a decade apart in future cost. An operator reads the minutes before admiring the kitchen, because the kitchen is the vendor's story and the minutes are the building's confession.

A tenant who stays is an asset nobody prices. Five years of the same good tenant means no re-letting fees, no vacancy, no turnover damage, and rent reviews that arrive without drama. That stability is worth a meaningful slice of yield, and it is partly a property decision, made at purchase. Family-sized homes near schools hold tenants. Practical layouts in quiet streets hold tenants. Properties that attract a revolving door of short stays give some of their headline yield straight back in churn. Managers can tell you which properties in their portfolio hold people and which leak them. Almost nobody asks.

Gross yield is the brochure number; net is the truth. Buyers and marketers quote gross because it flatters. Operators live in net, after management, maintenance, insurance, strata, land tax, and the vacancy the brochure assumed away. Two properties with identical gross yields can sit a full percentage point apart in net, which across a decade of holding is not a detail. It is the investment. In commercial property this discipline is even more central, because the lease itself carries so much of the value: who pays the outgoings, when the reviews fall, how strong the tenant's business is. An operator reads the lease before the building. The building is what you insure. The lease is what you own.

All of this shapes how I buy for clients now, in ways that sometimes surprise them. I inspect like a manager, not a purchaser: I want the strata minutes and the outgoings history before I want the second viewing. I ask how the property re-lets, not just what it rents for. I have a strong preference for the boring building on the better street over the striking building on the lesser one, because after enough Tuesdays you learn that boring is a compliment. And I price the ownership years, not the purchase day, because the purchase day is one transaction and the ownership is a decade of them.

There is a structural reason this perspective is rare in buying advice, and it is worth naming. Most of the people who advise property buyers are paid on the day of purchase and are gone by the first Tuesday. The feedback that would correct their advice arrives years later, in someone else's inbox. The operators hold the evidence and are never in the room when the buying decision is made. I have tried to build a practice where those two rooms are the same room.

A property is bought in a day and operated for decades. Most buying advice comes from the day. The best of what I know comes from the decades.

The rules around property in superannuation moved twice in the space of six weeks this year, and the two moves point in the same direction.

The May budget restricted negative gearing on established residential property and left commercial property untouched in every entity type. Then in June, Parliament passed legislation that ends new borrowing by self-managed super funds against residential property. From the tenth of August 2026, a fund can no longer enter a new limited recourse borrowing arrangement unless the property qualifies as business real property. Existing residential arrangements are grandfathered, contracts exchanged before the deadline are protected, and refinancing continues. But the direction of policy is not subtle. The residential borrowing door in super has closed. The commercial door has been deliberately left open.

Established investors have read the signal, and self-managed fund acquisitions of commercial property have surged since May. Some of that is sound strategy. Some of it is a rush towards an asset class people do not yet understand, inside the most unforgiving compliance structure in Australian investing. Both halves of that sentence deserve attention, so let me take them in order: where the yield genuinely is, and what the obligations actually require.

The yield case is real. Gross residential yields in Sydney sit around 3.1%, the lowest of any capital city. Commercial yields across Sydney currently run between 6.25% and 7.75%. That is not a rounding difference. It is a different kind of asset doing a different job, and for an investor whose priority is income, particularly income inside superannuation as retirement approaches, it is the difference between a portfolio that produces a livelihood and one that produces a valuation.

Gross yields, Sydney, 2026

Residential3.1%

Commercial6.25–7.75%

Scale to 8%. Residential is the lowest of any capital city; the commercial range covers the neighbourhood assets described here, where the tenant commonly pays some or all of the outgoings.

Within commercial, my interest is the neighbourhood end: the chemist, the medical suite, the accountant's office, the small supermarket, the strip of six shops that serves the surrounding streets. These are unglamorous assets with three durable strengths. The tenants are often essential services that trade through every cycle. The leases run longer than residential tenancies, commonly three to five years with options, and the tenant typically pays some or all of the outgoings, which moves the headline yield closer to a net one. And the supply is constrained: Australia is estimated to be between 150 and 250 shopping centres short of demand, while the institutions that once dominated the sector have stepped back, leaving assets at prices private investors can reach.

The honest caveat is vacancy. A flat in a tight Sydney market re-lets in a fortnight. A shop can sit dark for six months, and an investor who has not planned for that is an investor in trouble, particularly inside a fund that must keep paying rates, insurance, and loan repayments while the windows stay papered. In neighbourhood commercial, the strength of the tenant's business matters more than the condition of the paint. The lease and the covenant are the asset. Read them before you admire the building.

For business owners, there is a further fit worth naming. The business real property exception is one of the very few related party doors left open in superannuation law. A fund can buy the premises a member's business operates from and lease it back to that business at market rent. Done properly, the business pays rent to its owner's retirement fund instead of to a stranger's. Done improperly, it is a compliance failure with an audit trail. The difference is entirely in the discipline.

That brings me to the part of this subject most commentary skips, which is what the compliance obligations actually require. In my experience they are not fine print. They decide whether the strategy works at all.

The sole purpose test comes first. Everything the fund owns must exist to provide retirement benefits, nothing else. A residential property held by a fund can never house a member or a relative, at any rent, even a market one. There is no version of the beach house in the SMSF that ends well.

Business real property has a precise meaning: property used wholly and exclusively in one or more businesses. Commercial zoning is not enough, and this now matters more than ever, because since June this definition decides whether a fund can borrow at all. The shop with the flat above it can fail the test. The consulting room in a partly residential building can fail it. Whether an asset qualifies is a legal question about use, answered before contracts, not after.

The leaseback must be genuinely at arm's length. A real lease, market rent supported by evidence, reviews conducted on time, rent actually paid in full, every year. Auditors check, and the penalty regime for non-arm's length arrangements is severe enough to defeat the purpose of the structure. Paying your own fund a mates' rate, in either direction, is the most expensive favour in Australian tax.

Borrowing, where it is still available, is structurally strict. The asset sits in a separate holding trust. The borrowing attaches to a single acquirable asset, and borrowed money cannot fund improvements. Lenders in this market want more equity and charge more for the debt; loan-to-value ratios of 60 to 70% are common for commercial, and the fund must hold genuine liquidity beyond the deposit, because it has to carry vacancy, outgoings, and repayments without a forced sale.

The fund's investment strategy document has to do real work. It must contemplate the purchase honestly: what the asset does to diversification when one property is most of the fund, how the liquidity holds up, how members' insurance needs are met. And the asset must be valued at market annually, which for commercial property means defensible evidence, not the trustee's optimism.

None of this is a reason to avoid the strategy. It is the strategy. The May and June changes have created a genuine opening for established investors, and Division 296, which from the first of July this year applies higher tax to realised earnings on the largest balances, has only sharpened the value of getting the structure right early. But the opening is narrow, and it rewards exactly one kind of investor: the one who treats the compliance obligations as the terms of the deal rather than the paperwork after it.

The yield, for some time now, has been on the commercial side of the ledger. The door is open. Properly or not at all is the only way through.

In 2016 I co-founded a property technology business called Rent360, Australia's first online marketplace for property management.

Investors listed their property, independent managers competed for it, and fees fell by about a third. The platform did what good software does: it collapsed an information problem. Finding a capable manager had been slow, opaque, and priced accordingly, and we made it fast, transparent, and cheaper.

What has stayed with me is the question the technology never touched. However efficient the marketplace became, every landlord eventually asked the same thing: who, exactly, will be looking after my property? Software compressed the search. It did not create an ounce of the trust. That distinction, between the problems technology dissolves and the ones it cannot reach, is the most useful lens I know for what is happening to property right now, because AI is running the same experiment across the whole industry at a far greater speed.

Every task in property buying is one of two kinds. It is an information problem or a judgement problem. The line between them decides everything.

Where the line falls

Information problems, and falling fast

• Search and aggregation
• Suburb research
• Comparable sales analysis
• Feasibility modelling
• The reporting layer around all of it

Judgement problems, and not moving

• Negotiation, where the price actually moves
• Access, earned one settled deal at a time
• The particular house, not the median
• Counsel, with something at stake

The information problems are falling fast, and I want to be honest about what that includes. Property search and aggregation. Suburb research. Comparable sales analysis. Feasibility modelling. The reporting layer that wraps around all of it. Work that took a research analyst a week now takes an afternoon, and the afternoon version is often more thorough. I use these tools daily and they are remarkable.

Here is the uncomfortable part for my own industry. A meaningful share of what buyers' advocates have historically charged for was information asymmetry. The adviser had the sales data, the suburb knowledge, and the analytical templates; the client did not; the fee lived in the gap. That gap is closing on the timescale of software releases, not careers. Anyone can now produce a handsome, confident, forty-page suburb report in minutes. Investors should welcome this, and then draw the obvious conclusion: when a report costs nothing to make, the report is no longer the value. If you are paying a professional mostly for information, you are overpaying, and you soon will be by a wider margin.

What does not automate is just as specific, and it is worth being precise about why.

Negotiation does not automate, because the price of a property is not set by data. It is set by two parties with different needs, and the movement happens in what the money means to each of them. The model has never met the vendor. It was not in the hallway when the agent mentioned, off the record, that settlement timing matters more than headline price. It cannot hear the fatigue in a campaign that has run three weeks too long. Every dollar I have ever negotiated off a purchase price came from information that was never written down anywhere a model could read it.

Access does not automate. The call that comes before a listing exists is the product of years of settled transactions, of being the buyer whose finance never falls over, of a reputation an agent can lend to their vendor without risk. There is no interface to that. It is earned slowly, between people, and it is spent one phone call at a time.

The physical world does not automate, at least not yet. A model can tell you the suburb median. It cannot tell you which house the median is hiding: the one on the flood-affected side of the street, the one with the unapproved extension, the one where the afternoon light dies at two o'clock, the one whose strata minutes contradict the agent's cheerful summary. Property is decided at the level of the particular, and the particular still requires standing in the room. The greatest risk AI has introduced to investors is not bad information. It is polished, plausible, generic information mistaken for judgement.

And counsel does not automate. The most valuable conversations in my practice are the ones where a client is talked out of something: the purchase that fits the spreadsheet and not the portfolio, the auction where the number stopped making sense two bids ago. A model has no duty to you, no memory of what you said you were building towards, and no willingness to lose a fee to protect a strategy. Advice with nothing at stake is content.

The way through, for a practice like mine, is not to bolt AI onto the old workflow and call it innovation. That produces a brief efficiency gain competitors match within a year. The better answer is to rebuild the delivery model around the new economics: let the machines compress the research layer completely, and spend the recovered time where the value has concentrated, in inspections, negotiation, and counsel. At Valentine Investment Partners the division of labour is explicit. AI does research and administration. Humans do judgement. The moment those roles blur in either direction, something is being done badly.

I want to be clear that none of this is nostalgia. I built property technology; I have no romance about manual work, and I have watched software make an industry measurably better at its job. The point is narrower and more practical. Technology keeps repricing the industry, and the repricing always runs in the same direction: the information gets cheap, then free, while the judgement gets scarcer relative to the noise, and more valuable. It happened to property management fees a decade ago. It is happening to research now.

For investors, the practical test is one question, asked of anyone you pay: which side of the line is this fee for? If the answer is information, the price should be falling towards zero, and you should let it. If the answer is judgement, then judge the judgement: the record, the reasoning, the willingness to say no.

The research is becoming free. The judgement is the fee.