–>
The 2023 version of this article explained capitalisation rates reasonably well and contained a number that was plainly wrong, which nobody including me picked up. It also attributed office vacancy data to the wrong source and described how residential property is valued in a way that is not how it is valued. Those are the easy fixes. The harder update is that the arithmetic at the centre of the piece has changed: the risk-free rate you compare commercial property against is no longer 4.5%, it is above five, and on prime Sydney assets the premium for taking property risk has very nearly disappeared. I would rather set them out plainly than quietly fix them, because two of them are the kind of mistake that makes a reader doubt everything else on the page. Wrong, and obviously so There has never been a national office vacancy rate of 0.9%. The figure at the time was in the low teens and the national rate today is 16.1%, with CBD markets collectively at 14.9%. It was almost certainly a decimal point that slipped and nobody caught it, including me, for three years. A single number like that undermines a whole article, which is the lesson rather than the arithmetic. Wrong source Office vacancy in Australia is published twice a year by the Property Council of Australia in its Office Market Report. Domain publishes residential market and auction data. Attributing the figure to the wrong organisation is the sort of thing that is invisible to most readers and immediately disqualifying to the ones who matter. Half right, and the second half is not how it works The first half is correct and is the single most useful idea in the article. The second half is not. Residential property is valued by direct comparison with recent sales of similar properties, not by valuing the land. Land value matters enormously to what a house is worth, particularly in the Eastern Suburbs, but it is not the valuation method. The clean way to say it is that commercial property is priced off its income and residential off its comparables. Close on the first, nowhere near on the second Sydney CBD vacancy rose from about 11.3% to a peak near 13.8% in January 2026 and has since eased to 13.3%, with positive demand for a fourth consecutive reporting period. The 14.7% figure was a reasonable call that the market did not quite reach. The 20% worst case never looked like happening. Quoting a worst case alongside a base case, with no probability attached to either, is a habit worth dropping. A sentence that means nothing The intended comparison was with commercial property, which is the point being made in that paragraph. It is a small thing and I am including it because unedited sentences like this are what a reader uses to decide how carefully the rest was written. Premium vacancy at 7.7%, secondary near 15%: the office market scored. Thirty per cent of CRE debt maturing by 2028: commercial refinancing in 2026. Comparing a 2023 cap rate with a 2026 one is less useful than it looks, because the average now contains two markets that have separated. Switch between the views.
Value equals net income divided by the cap rate. That is the whole formula and the 2023 piece got it right. What it did not show is the sensitivity, which is the only reason the formula matters: on the same income, a shift of one percentage point in the cap rate moves the value by hundreds of thousands of dollars without anything happening to the building at all. Move the slider. Net income means income after landlord outgoings, not gross rent; the gap between the two can be 1.5 to 2.5 percentage points on a Sydney asset depending on the lease structure. A cap rate is an observed market rate derived from comparable sales, not a return you choose. Indicative arithmetic only, and not a valuation. That was the most useful paragraph in the original and the reasoning still holds exactly. What has changed is the inputs. The Australian 10-year government bond yield was 5.12% at 30 October 2026, having fallen 0.28% over the month but sitting about 0.8% higher than a year earlier, with the cash rate at 4.85%. Apply the original’s own test and the answer is uncomfortable. If a prime Sydney asset yields less than a government bond, nobody is buying it for the income. They are buying scarcity, rent growth, a tenant covenant and an asset that will still be lettable in fifteen years. That is a legitimate investment case and it is a completely different one from the case the 2023 article was making. Sydney CBD secondary office at 7.25% to 9.0% clears the original’s 2.0 point hurdle comfortably. It is also the segment with vacancy near 15%, the capital expenditure requirement to change grade, and the refinancing problem. A high cap rate is not a bargain; it is the market pricing a risk you have to be willing to carry. Before comparing anything to a bond, strip out landlord outgoings, vacancy allowance, incentives amortised over the lease, leasing fees and capital expenditure. On a Sydney asset the gap between a gross and a net figure can be 1.5 to 2.5 percentage points, which is more than the entire risk premium the original was arguing about. Every item in the original list is still on a lender’s list. Here it is sequenced, with the reason each one is there, because knowing why a lender asks for something is what lets you answer it well. Term remaining, options, the tenant’s covenant and whether the rent is at, above or below market. A commercial facility is secured against an income stream, so the lease is the asset as far as the lender is concerned. Two years remaining and six years remaining produce very different answers from the same credit committee on the same building. The original listed capacity to repay, which is right, but commercial facilities are usually tested on an interest cover ratio: net income divided by interest cost. At a cash rate of 4.85% that test has tightened considerably since most existing facilities were written, and it frequently binds before the loan-to-value ratio does. With no tenant there is no income to capitalise and no cover ratio to test, so the lender is effectively lending against your other income and the land. Expect a materially lower loan-to-value ratio, a shorter term and a higher rate. This is why an owner with a looming vacancy should be solving the lease before solving the loan. The original mentioned property type and a valuer’s report. What it did not say is that grade is now the dominant variable: premium Sydney CBD office vacancy is about 7.7% against secondary near 14.9%, and valuers price that difference. The same floor area in two buildings a block apart can produce valuations that differ by a third. This part of the original was right and is worth repeating plainly: commercial lending generally carries higher rates, shorter terms and lower loan-to-value ratios than residential. A three or five year facility on a fifteen year hold means you are refinancing several times, and each of those is a point at which the market gets to re-price your asset. The equity gap, and what to do eighteen months out: commercial refinancing in 2026. Contract, title, planning certificate and reports: due diligence in NSW. That is the part worth saying out loud. A national office vacancy rate of 0.9% is not a subtle error; it is off by more than an order of magnitude, and it sat there through every time the page was read. Nothing else in the piece depended on it, which is exactly why it survived. I have started going back through older articles for this reason, and this is the fourth one where I have found something I would not publish today. If you read any of my older pieces and something looks wrong, tell me, because I would rather fix it than defend it. On the substance, the thing I would most want a reader to take away is the one the original nearly got to. Commercial property is priced off its income and residential off its comparables, and that single difference explains why the two markets moved at different times for the same reason. When the cost of money rose, every commercial asset was revalued within a quarter by a formula. Housing had to wait until a buyer turned up who could not borrow what the last one did, which took about two years. Same cause, two completely different clocks, and if you own both you felt them at opposite ends of the cycle. The last thing is the bond comparison, because it is genuinely uncomfortable and I would rather put it on the page than leave it out. A ten year Commonwealth bond is paying 5.12% and prime Sydney CBD office is quoted from about 5.0%. On income alone that is not a case anyone can defend. The case for prime commercial is rent growth, scarcity and the fact that the best buildings in the country will still be fully let in twenty years, which is a growth argument rather than a yield one. Anyone selling it to you as income in 2026 has not done the sum. Market scorecards, suburb guides and planning analysis: Alan’s articles. What the site is worth to a residential buyer or a developer against what it is worth as an income asset. Those two numbers have diverged a long way, and under the new density rules the answer surprises people. Thirty-five years in the Eastern Suburbs. Cap rate and yield figures differ materially between providers depending on whether they are gross or net, prime or blended, asking or transacted, and on the reference date; the ranges given here are drawn from several 2026 sources and should be treated as indicative rather than as a market rate for any particular asset. Calculator outputs are indicative arithmetic, not a valuation or an appraisal. Alan Weiss is a licensed residential real estate agent, not a commercial agent, registered valuer, mortgage broker or financial adviser, and nothing here is a recommendation about any property, asset class or investment. Includes Alan Weiss’s personal views. General information only; not financial, tax or legal advice.Cap rates in 2026: what a commercial property is actually worth, and the typo I left in for three years.
Three of these are my errors. The fourth is a forecast that nearly landed.
The headline numbers barely moved. The spread inside them moved enormously.
The original did this calculation once, at one rate. The point is what happens when the rate moves.
The 2023 article used 4.5% term deposits and asked for 6.5% from property. Run the same logic today and prime does not clear the bar.
Comparison
2023 article
November 2026
Risk-free rate used
4.5% term deposit
5.12% 10-year bond
Premium the article wanted
2.0 points
2.0 points
Cap rate that implies
6.5%
7.1%
Sydney CBD prime office
4.8%
About 5.0% to 5.7%
Sydney CBD secondary office
Not separated
About 7.25% to 9.0%
You are buying growth and safety, not income
And the yield is telling you something
Gross yield is not what you receive
The lending checklist held up. What it was missing is the order and the reason.
The lease, before anything else
Interest cover, not just loan-to-value
Vacant, and the question changes entirely
Grade and location, through the valuer
Expect lower leverage and shorter terms than residential
Own a shop, a suite or a small commercial holding in the East? The residential question is usually the one worth asking.
35+ years in Sydney’s Eastern Suburbs
Commercial valuation · November 2026
5.12%
Australian 10-year government bond yield at 30 October 2026, with the cash rate at 4.85%. Prime Sydney CBD office is quoted between roughly 5.0% and 5.7%. That is the entire investment case for prime commercial property in one comparison.
01 · Four corrections and one forecast
“Sydney CBD’s vacancy rate is higher than the national average of 0.9%.”
“According to the latest data from Domain.”
“Commercial property is valued on rental income, whereas residential property is valued on the value of the land.”
“Vacancy could rise to 14.7% in three to five years, or even 20% in worst-case scenarios.”
“Government bonds are more liquid than government stocks.”
What happened to the offices themselves
And to the debt against them
02 · Cap rates then and now
03 · The sum itself
Net income a year
Cap rate
Implied value
—
If the cap rate softens 25 basis points—
If it firms 25 basis points—
Value per $1,000 of extra net income—
04 · Against the risk-free rate
What that means
Prime is not a yield play
Where the yield is
Secondary
The honest version
Add the costs
05 · Borrowing against it
01
02
03
04
05
When the facility matures
Before you buy anything
Weiss Real Estate · My read
A wrong number sat in that article for three years and the only reason it is being fixed now is that I went back and checked.
Everything else I have written
Alan Weiss
Principal, Weiss Real Estate
Sources