On 22 October 2021, Aurizon announced it would pay $2.35 billion in cash for One Rail Australia. We spoke to the company that day, and three days later we wrote to its directors. Over the four years that followed we wrote twice, replied once, drafted a third letter we never sent, met the chairman, presented to the board, and met his successor.
The thing we argued for longest, separating Aurizon's rail Network from the rest of the business, did not happen. The company's 2026 annual report records the conclusion:
The Company also completed its review of Aurizon Network's ownership structure. Following a comprehensive assessment of strategic and capital management alternatives, the Board concluded that retaining 100% ownership of Network remains the option most likely to maximise long-term value for shareholders.
We remain shareholders. Both our funds hold Aurizon.
That is the honest shape of most engagement, and it is why this is worth writing down. Engagement is not a vote you win. It is a way of testing a board, on the record, and the record is useful whichever way the decision goes.
What we asked for
Aurizon's share price fell as the market read the One Rail announcement. Our letter of 25 October 2021 measured the fall at more than $600 million of Aurizon's market value in two days — a quarter of the price being paid for the business it was buying — and put eleven concerns to the directors. Four of them carried most of the weight.
The price. Aurizon was paying about 10.5 times the acquired business's earnings before interest, tax, depreciation and amortisation — a common way of pricing a business against its cash earnings. The market was valuing Aurizon itself at about 7.1 times the same measure. Buying something dearer than yourself needs an argument, and the company's argument was that the deal offered a return in excess of buying back its own shares. We could not reconcile that with anything the company had disclosed.
The disclosure. For an acquisition worth 34% of Aurizon's market value before the announcement, shareholders were given no historical profit and loss statements, no balance sheet, no cash flow statement, and no indication of the target's operating earnings or asset backing.
The absence of a shareholder vote. A transaction of that size was not conditional on shareholder approval.
The absence of any disclosed independent advice on what the transaction meant for shareholder value.
We also noted that Aurizon's eight non-executive directors collectively owned 527,391 shares — worth less, at the time, than the $2,019,000 of directors' fees Aurizon's shareholders paid in 2021. And we told the board we were prepared to lobby other investors to call an extraordinary general meeting, or for a board spill, if the matters were not addressed.
The board answered
This is the part that most engagement never reaches, and it is the part worth paying attention to.
Tim Poole, Aurizon's chairman, confirmed in writing that our letter had been circulated to every director, met us within a fortnight, and replied in writing on 3 November 2021. His letter set out the methodology behind the return figure we had questioned: an unlevered pre-tax return over twenty years of forecasts, beginning in the first full year of ownership, and assuming the East Coast Rail business was demerged rather than sold. It itemised what the $80 million of fees was made up of — stamp duty, advisory and due diligence, financing and divestment costs — against a letter that had put those fees at 3.4% of the transaction's value. It explained the funding arranged for East Coast Rail ahead of its divestment.
On the two governance questions his answer was direct, and both sentences are his:
There is no requirement under the Corporations Act or the ASX Listing Rules for Aurizon to obtain shareholder approval for the proposed acquisition of One Rail Australia.
There is similarly no requirement for Aurizon to obtain and disclose an independent report or advice assessing the shareholder value implications of the acquisition of One Rail Australia.
Those are statements about what the rules require, and they are correct. They are also the answer to a different question from the one we asked, which was not what the law compelled but what the board judged shareholders were owed.
Aurizon disclosed, in the same letter, forecast earnings for the acquired bulk business in the first full year of its ownership — a figure that had not been given when the deal was announced. We had asked for it. Our reply of 10 November questioned why it had not come earlier.
And then the board disagreed with us
Our 10 November letter went through the chairman's answers point by point. We recorded his acknowledgement that the company's communications and financial disclosures "could have been better". We recorded that he had told us he believed Aurizon was undervalued by as much as 50%, while describing One Rail as acquired at a fair price that was neither expensive nor a bargain, and had confirmed the same discount rates were used to compare the two — which left the original claim, in our view, unreconciled. We recorded his confirmation that no independent financial advice had been obtained. And we said we would advocate for future long-term incentives to be weighted more heavily towards Aurizon's shareholder returns relative to other companies.
He wrote back on 12 November and told us we had it wrong:
However, I do not believe that the isolated quotes and paraphrased comments in your letter of 10 November accurately reflects our meeting because they do not have the benefit of the context of the broader discussion. For example, I do not agree with your comments regarding poor financial disclosure. While your interpretation of our communications may have caused you 'uncertainty and consternation' this is not reflective of the view of the majority of investors or analysts.
He added that Aurizon welcomed the feedback and valued the debate, and that it was clear our views on parts of the transaction differed from the board's.
We have not changed our view of the transaction. We publish his because an engagement record that contains only the replies agreeing with us would not be a record of anything. A chairman who writes back to say a shareholder has mischaracterised a meeting is a chairman who is reading the letters, and how a board argues is information you cannot get from a spreadsheet.
The second letter, and the small thing it moved
We wrote again on 17 April 2023, this time to introduce ourselves to a changing board, set out how we think about environmental, social and governance matters, and put four issues to the directors: the company's exposure to decarbonisation, its governance and capital allocation including a large new containerised freight contract, its management of people and safety, and its community relations.
One of our asks was narrow and specific. Aurizon ran regular employee engagement surveys and did not publish the results, and we said that publishing them would reduce a concern the market had about the company. The chairman's reply of 31 May 2023 answered each issue in turn, told us enterprise agreements covering more than 70% of covered employees had been renewed since December 2021, described how the directors had assessed the freight contract — some of them meeting the customer's senior people — and on the surveys said this:
We note the feedback regarding employee engagement survey disclosures and we will consider providing additional information in this year's Sustainability Report.
That is not a commitment, and we did not read it as one. It is a board telling a shareholder, in writing, that it has heard a specific request and will weigh it — which is a position it can later be asked about. Most of what engagement produces looks like this rather than like a victory.
The letter we did not send
In February 2025 Aurizon announced a review of the ownership of its Network business — the thing we had been arguing about since 2021.
By June we had a third letter drafted. It was abrasive, deliberately so, and it was written to be sent.
It never went. Our Advisory Board, which is independent of the investment team, read the draft and advised against sending it as written. Its counsel was to take a friendlier approach, on the plain ground that the point was not to put a position on the record but to be heard by the people conducting the review. What it pushed us towards instead was asking for the meeting and making the case in person.
So that is what we did, and the presentation exists because the letter was stopped. In August we set out our view of the Network to the board, and later met the directors conducting the review. In December we met Aurizon's new chairman and restated the same position.
The presentation was dated 20 August 2025 and ran to sixteen slides. This is the page that states what we were asking for:
The third of those bullets is the one that took the most work, because it answers the company's own objection rather than restating our case. When Aurizon last reviewed the Network's ownership in 2019, it told shareholders that separation would cost roughly $80-100 million a year in lost operational synergies and duplicated functions. Six years later we went back through that argument line by line:
Most of what the company had said would be lost had, by 2025, already been built.
The unsent draft is the part of this story we would not otherwise have had any reason to publish, and it is the most useful part of it. A manager that writes to boards for long enough acquires a reputation for writing to boards, and past some point the letters stop being read. Knowing when a letter is the wrong instrument matters as much as the letters do — and on this occasion that judgement was not ours. It came from the people we had put in the room to tell us things we would not tell ourselves.
The Network, and the answer we did not get
The 2026 annual report, quoted above, is the answer: the board assessed the alternatives and concluded that keeping all of the Network was the course most likely to maximise long-term value. The same report sets out what shareholders got instead:
Total dividends for FY2026 were 23 cents per share at an increased 90% pay-out ratio. During the year, Aurizon also completed a share buy-back program and returned a further $250 million to shareholders. Over the past four years, shareholders have received approximately $1.8 billion through dividends and share buy-backs.
That is worth reading against 2021. Our record of the November meeting has the chairman attributing part of the market's reaction to the expectation of dividends at the lower end of the company's range for the next few years — and our record of that meeting is the one he went on to dispute, so take it as our account rather than his. The arithmetic needs no account: spending $2.35 billion of cash reduces what is available to pay out. Our first letter had argued that the money was better returned to shareholders than spent on the acquisition. A payout ratio lifted to 90% four years later is the company coming back to what its cash generation supports.
We do not claim credit for that. A board that raises a payout ratio four years after an acquisition is allocating its own capital, not answering a letter, and we have no way of separating the two. What we can say is that the question we put to the directors in 2021 — whether this use of shareholders' money beat giving it back to them — is a question the company has now answered twice, once by buying One Rail and once by lifting the payout.
We argued the other way on the Network and we were not persuaded by the outcome. We are also not entitled to it. A manager with a few per cent of a company gets to make its case to the directors who decide, and then it holds, sells, or votes. What we have instead is four years of the board's own reasoning in writing, which is what lets us judge the decision rather than guess at it.
What four years of this actually bought
Not the Network. What it bought was:
Disclosure that was not there. Forecast earnings for an acquired business, the methodology behind a return claim, and an itemisation of $80 million of fees — all provided after shareholders asked, which is late but is not nothing.
A record that can be tested. A board that has written down why it did something can be asked, later, how that reasoning held up. A board that said nothing cannot.
An argument about alignment. We said we would advocate for incentives weighted towards relative shareholder returns, and we have kept saying it.
Information about the board itself. The chairman circulated our letter to all directors, met us twice, answered in detail, and told us plainly when he thought we were wrong. That is a serviceable board, and we learned it from the correspondence rather than from a governance score.
And one decision not to write. The third letter would have been the easiest of the three to send and the least likely to achieve anything.
How these letters get written
Aurizon is not a special case. We have written more than 80 letters to boards of companies we own since 2021, and 37 of them drew written replies on the issues we raised. Letters to boards are not new here — the practice goes back to 2012, and 2021 is simply where our searchable record of it begins, with the Aurizon letter above as the earliest entry in it.
Each letter follows the same path: an analyst drafts it, the investment team reviews it, we discuss the issues with management, our Advisory Board — which is independent of the investment team — reviews the draft and advises on it, the letter goes to the chair, and the board acknowledges it.
The Advisory Board step is not a formality, and the Aurizon draft above is what it looks like when it bites: the Board can tell us a letter we have written and intend to send is the wrong move, and it did. Its role is to advise rather than to decide — it is not a decision-making body and does not direct the portfolio — which is precisely why its advice is worth having on a document we are already committed to.
The step that does the most work, though, is the third one. We talk to the company before we write, and we say what we are going to say. Both Aurizon letters record that the company had engaged with us beforehand and that we had been open about our concerns. A letter that arrives as a surprise is a press release with an envelope.
What this does not tell you
It is one company over four years. One engagement is a story, not evidence. The aggregate — more than 80 letters, 37 replies — is a better measure of whether this is a practice or an anecdote, and it is still a measure of activity rather than of outcomes.
We did not get the outcome we argued for. The clearest result in this piece is a board deciding against us after considering the question.
The chairman's words are the company's, not our analysis. Every passage quoted above is Aurizon's position, written by its chairman, and we disagreed with parts of it at the time.
Engagement is not a reason to own something. It is what we do with a holding, not why we hold it. A company we write to is a company we have valued, and the valuation is the decision.
And none of this promises an outcome. We write because the alternative — owning a business for years and never putting a question to the people running it — makes less sense to us, not because the letters always work. Four years of this one bought better disclosure, a documented board and a no.
.jpg)