There is a version of the Australian property story that I have heard told many times over the course of a decade in this industry, and it contains a genuine truth that deserves acknowledgment before it is examined carefully. Australian residential property has, over the past three decades, delivered compound returns that placed it among the better-performing asset classes available to domestic investors. Capital gains in Sydney, Melbourne, and Brisbane over that period have been substantial by any reasonable measure. Households that entered the market in the 1990s and held through the cycles have, in many cases, accumulated wealth that would not have been replicated in a savings account or a managed fund. I do not dispute this record, and I am not here to retrospectively condemn the individuals who responded rationally to the incentives their environment provided. The record of residential property returns is real. What I am here to do is examine what kind of return it actually was — and what kind of return was being foregone in its favour.
The residential return record is, on careful examination, an appreciation record rather than a yield record. The income return, the rent collected relative to the value of the asset, has not kept pace with the growth in asset values; it has been compressed by it. A gross rental yield of approximately 3.5 per cent on a median capital city residential investment property in 2026 represents the arithmetic consequence of asset prices having grown faster than rents over the same period that produced the capital gains story above. But the 3.5 per cent gross figure is not the number that matters. The number that matters is what arrives at the investor’s bottom line after the actual cost structure of residential investment is applied — and that calculation is materially different from the headline figure. What residential investment has delivered as an income-producing asset, once the architecture of its costs is properly accounted for, is a return that would not be considered competitive with the productive property alternative on yield grounds alone. The competition was never on yield grounds, because yield was not the basis on which residential investment was being evaluated. That is precisely the problem.
The frictional cost structure of residential investment property is substantial and is not always presented in its entirety when the gross yield figure is cited. Property management fees consume eight to ten per cent of gross rental income as a starting point. Vacancy, even a modest vacancy of four to six weeks per year in a well-located property, compounds the degradation. Maintenance and capital works, which accumulate with age and cannot be indefinitely deferred, represent an ongoing liability that conservative estimates place at one to one and a half per cent of property value annually; in older stock or through a renovation cycle, this figure is materially higher. Council rates, insurance, water, and land tax, where applicable, add further. The interaction of these costs with the gross rental income produces, for a well-managed residential investment property in a major Australian capital city, an effective net yield in the range of 1.5 to 2.0 per cent. This is not a theoretical or pessimistic estimate; it is broadly consistent with the aggregate picture visible in Australian Taxation Office data on rental income and deductions, where the gap between gross rental receipts and net rental income across the investment property population has been a persistent structural feature for many years.
Against this, consider the economics of a prime industrial asset acquired on a net lease — the standard contractual structure for warehouse, logistics, and manufacturing facilities in the Australian market. Net lease means precisely what it states: the tenant, not the landlord, pays the outgoings. Council rates, insurance, maintenance obligations, and in many cases capital expenditure requirements are contractually the tenant’s responsibility. The landlord receives a clean income stream against which the headline capitalisation rate is the operative yield figure — not a gross figure to be degraded by a cost structure, but the net income return before financing costs. Prime industrial and logistics assets in established Australian corridors are currently transacting at capitalisation rates above five per cent for quality new or near-new facilities, with established assets in high-demand precincts offering comparable or tighter yields reflecting the structural supply constraint in serviced industrial land. The comparison between the 1.5 to 2.0 per cent effective net residential yield and the five-plus per cent industrial net lease yield is not a marginal difference of presentation. It is the difference between an asset that generates income and an asset whose investment case rests primarily on someone paying more for it in the future.
The durability of the industrial yield is not incidental to the comparison; it is the central feature of the asset class. Net leases for industrial and logistics facilities are typically structured over five, ten, or fifteen year terms, with fixed annual rent reviews at three to three and a half per cent per annum or CPI linkage. The tenant — an established operator whose business depends on the specific facility, its location, its power connection, its logistics connectivity — has embedded itself into the site through fit-out investment, equipment installation, and operational integration that makes relocation genuinely costly. Vacancy risk in industrial property, the primary mechanism through which residential effective yield is destroyed, operates on fundamentally different terms: industrial vacancy events are less frequent, search periods for replacement tenants in constrained supply markets are often shorter, and the structural undersupply of serviced industrial land — documented in detail in the research underlying this series — supports asking yields rather than compressing them. The income return is contracted, predictable, and structurally insulated from the management intensity that characterises residential tenancy.
If the industrial yield advantage is this durable and this measurable, why has Australian domestic capital not moved toward it at scale? The answer returns us to the architecture examined in the preceding piece. The tax framework, until the proposals of the May 2026 Budget, actively favoured residential investment through negative gearing deductibility and the capital gains tax discount — instruments that allowed residential losses to shelter other taxable income and residential gains to be taxed at half rate. These mechanisms tilted the after-tax mathematics of residential investment so decisively that the question of yield was, for many investors, almost beside the point; the combination of leverage, tax treatment, and appreciation produced a return profile that rendered the yield comparison irrelevant. But even as the May 2026 Budget begins to dismantle this tax architecture — and I have acknowledged in the preceding piece that this proposed dismantling is directionally correct — it leaves the deeper layer untouched. The APRA prudential risk-weighting framework, which makes residential mortgage portfolios structurally more capital-efficient for bank balance sheets than productive sector lending, continues to tilt institutional capital allocation toward residential, regardless of what the yield comparison shows. We are reaching for the tax lever. The prudential lever remains locked in the opposite direction. The industrial yield was always there; one layer of the architecture obscuring it from domestic capital is now proposed for removal, and the layer beneath it is not.
What the listed market evidence confirms is that the capital outside the bank balance sheet constraint has already found the yield. The outperformance of industrial property vehicles on Australian listed markets over the past several years is not speculative premium; it is a repricing toward the fundamental income return that net lease industrial assets have consistently offered. The domestic capital that has moved has moved because it was free to respond to the yield signal directly — superannuation funds deploying capital into unlisted industrial assets, foreign investors acquiring logistics precincts, and private capital funding the development pipeline. Their movement is the market’s own confirmation that the yield case is real. What has not moved is the domestic capital most insulated from that signal by prudential design — and its immobility, as the previous piece established, is a regulatory condition rather than an economic judgment. Three decades of architecture directing Australian capital away from this yield has had a cost. What that cost has been — and how it has been documented from outside our own borders — is the question I turn to next.
