An office tower we would not buy today is already for sale
Listed property groups often hold office towers because selling below book value looks like an admission. I argue that the honest test is whether we would buy the building today at its market price, and that a board should act on the answer even when acting costs income.
A listed property group should treat any building it would not buy today, at today's price, as a building it is already in the process of selling. Put like that it sounds obvious. In practice it is one of the hardest disciplines in our industry to keep, because the moment it matters most is the moment selling looks worst: values are falling, transactions are scarce and every sale below book value will be read as a confession. My position is that those are precisely the conditions in which the test has to be applied, and that a board which defers it has made a decision, not avoided one.
I first learnt this in 2008, not at Corvalla. I was running an unlisted logistics fund whose gearing had reached 44% against a lending covenant of 50%. We could have held our estates and hoped the covenant survived. Instead we sold nine of them, about a fifth of the fund, at prices our investors found painful, and asked those investors for new equity. Several who had argued for patience were investing alongside us again by 2011. I have watched the same pattern play out in office property since 2020, on a larger scale and more slowly.
The question our hold reviews had stopped asking
When I became Corvalla's chief executive in 2021, our annual asset reviews were thorough. Every tower had a business plan, a leasing strategy, a capital-expenditure forecast and a valuation. What the reviews did not ask was an investor's question: if this building were offered to us tomorrow at its current market value, would we buy it, and would it be the best use of that money? Without that question a review drifts naturally into the asset manager's plan for improving the building we already happen to own. The plan may be sound. It simply answers something else.
We added the question in 2022, in a crude form. For each asset the investment team set out the forward return at today's value, compared it with our cost of capital and with the best alternative use of the same money, and stated plainly whether we would buy. For several of our office towers the answer was no. Their forward returns depended on incentives falling and occupancy recovering faster than any evidence suggested, while a logistics development on land we already owned could earn a yield on cost above six per cent. The comparison did not prove that office was finished as an asset class. It showed that some of our particular buildings, in their particular streets, were no longer where we would choose to put new capital.
Book value is a record, not a reason
The objection I heard most, inside the company and outside it, was that selling below book value would crystallise a loss. I understand why the phrase persuades. It is also confused. The loss, if there is one, happens when the building's value falls, whether or not anyone writes it down. Selling at the market price records a value that already existed; holding at book value defers the record and, with it, the decision. A securityholder looking at our discount to net tangible assets was already pricing the difference. In late 2022 our securities traded about 30% below the book value of our assets, which told me the market did not believe our office values and was charging us for its doubt.
A second objection is better: the market may recover, and a seller at the bottom hands the recovery to the buyer. That risk is real, and I have paid for it. We sold a Brisbane tower in 2023 because it was the easiest of the four to sell. Brisbane office recovered faster than our other markets, and a comparable building traded eighteen months later for about A$28 million more than we accepted. I would still have sold a tower that year. I would have chosen a different one. The lesson concerns which asset to sell first, not whether to apply the test.
A third objection came from our largest securityholder, who wanted every dollar of sale proceeds spent buying back our own securities at the discount. Buying our own assets at seventy cents in the dollar, they argued, beat any development. The arithmetic was attractive. But a buyback funded while gearing sat above our range would have weakened the balance sheet in the year it most needed strength. We agreed to buy back securities once gearing fell below 32%, and in FY25 we bought back A$90 million. That holder still thinks we were too slow, and the disagreement has been useful to both of us.
What acting early cost us, and what it bought
The cost was income. We sold four towers for A$1.46 billion between 2023 and 2025, and those buildings had been paying rent. Funds from operations per security fell from 25.1 cents in FY22 to 21.4 cents in FY24, and we reset the distribution from 20.4 cents to 17.1 cents. Some of our retail investors held our securities for that distribution, and a number of them wrote to me about it. I answered every letter, and I did not pretend the reset was free.
What it bought was choice. Gearing fell from 36.1% to 29.8%. We funded 186,000 square metres of logistics development and our share of two rental buildings without raising equity at a discount. Funds from operations per security recovered to 22.6 cents in FY26. Office is now 32% of our balance sheet rather than 54%, which means the next movement in office values will matter less to our securityholders than the last one did.
The test also told us when not to sell. In our FY26 review, two towers we had expected to sell passed it: their refurbished floors had leased at lower incentives than we had forecast, and their forward returns now beat the alternatives. We kept them. A rule that only ever produces sales is not a test; it is a disposal programme with a method attached.
I would not present this as a formula. Some owners have good reasons to hold assets the test would reject: a long lease to a strong tenant, a site whose redevelopment value is years away, a mandate that requires exposure to a sector. Those reasons should be stated, not hidden inside a business plan. What I would ask of any property board is to put the investor's question to management, in writing, for every asset at least once a year, and to insist on an answer that does not begin with book value. If the honest answer is that the board would not buy a building today, the conversation about when and how to sell it has already started, and the only open question is what the delay is costing.