We Looked For The Family. We Found Nine People And Forty Seconds Each.
Two registry analysts built the ownership graph of 140 outlets separately and agree on 137. The common node is three passive index managers with no board seats. At the largest of them, nine people vote 3,400 companies.
A standing series in which this newspaper investigates a widely held claim and reports that it could not substantiate it.
LONDON, United Kingdom — There is a dense common node in the ownership of the British press, and it is not a family, a trust or a proprietor.
This newspaper commissioned two corporate-registry analysts, who do not know each other, to build the ownership graph of the 140 largest news outlets in the United Kingdom and to identify any common controlling node. They worked separately. Neither was told the other existed. Both were paid the same fee in advance and told the report would run whatever it said.
Their graphs agree on 137 of the 140.
The node they both arrive at is three passive index managers — Alderney Index Partners, Cray-Vestergaard Global and Northmoor Passive — holding between four and nine per cent of nearly everything in the set, with no board seats on any of it and no litigated dispute with any of it in eleven years.
The three that did not resolve
Both analysts stop in the same place: two nominee structures in two jurisdictions, covering three of the 140, where the beneficial owner is not determinable from public filings.
They stopped independently and for the same reason. Getting past it means litigation, and we did not fund litigation.
Three outlets out of 140 is not a rounding error and we are not going to call it one.
What the families own
The graphs do find families. They find them where everyone expects them: three outlets under identifiable dynastic control, two of them loss-making, one of them the most cited newspaper in the country.
They are not connected to each other. The two analysts, working blind, both say so in almost the same sentence.
Nine people
The votes attached to the largest of the three holdings are cast by a stewardship team of nine people, responsible between them for 3,400 companies.
Not nine at each of the three managers. Nine at that one, and the other two teams are smaller.
In the 2025 season those nine cast votes on 1,182 media-sector resolutions across the 140 outlets and their parent companies. They supported management on 1,111 of them. They opposed on forty-six and abstained on twenty-five.
Ninety-four per cent.
Forty seconds
The managers publish their voting logs at resolution level, with a timestamp against each decision. We downloaded 41,700 rows and filtered them to the outlets in our set.
In the busiest week of the season the team entered 2,940 decisions across five working days. The median interval between one decision and the next is forty seconds.
Voting against management requires a written rationale, filed in the same system. Voting with management requires a click.
Dr. Ines Okwuosa-Faber, who runs the largest of the three teams, says the timestamps record when a decision was entered and not when it was made, that the real work happens in engagement meetings months earlier, and that we have timed the typing. Her full reply is printed below. We cannot check her account and we say so plainly: the meetings she describes are not logged, and nobody outside the firm can see them.
The second report
We commissioned two graphs so that agreement would mean something. It does. It also produced eleven pages we did not want.
Emrys Latham-Oduya reached the same node as his counterpart and then argued that the node is not a finding at all. A shareholder that has never once voted against, he writes, is not exercising control; it is declining to. The largest holder in the room, if it always votes with the room, is functionally absent from it. What we have established is the existence of a very large passenger.
He goes further, and this is the part the desk has argued about for three weeks. An unsteered press, he writes, is a worse condition than a steered one, because a steered press has somebody who can be named, held, sued or shamed, and an unsteered one has nobody at all — and it is a much harder story to sell, which is why nobody sells it.
What this desk takes from it
We went looking for a hand and found a queue.
The largest shareholder in the British press is a spreadsheet worked through by nine people, in a week when there are more resolutions than hours, under a rule that makes agreement free and disagreement expensive. Nobody designed that to control anything. It controls things anyway, in the weak sense that a thing which never says no is a thing that has never said no.
Both reports are published in full with this story, including the one that says we have nothing. Our own pension scheme holds units in tracker funds run by two of the three, which Dr. Okwuosa-Faber points out is a better opening paragraph than the one we wrote.
She may be right about that too.
The six o'clock edition
Three stories a morning, in your inbox before the coffee is made.
Sources & Method
We commissioned two corporate-registry analysts who do not know each other, paid them the same fee in advance, told each that the report would be published whatever it said, and gave them the same brief: build the ownership graph of the 140 largest news outlets in the United Kingdom and identify any common controlling node. Neither was told the other existed until both had delivered. We then took the three managers' own published voting logs and counted what those holdings actually did in the 2025 season.
Who we spoke to
- Odile Vantriest-Boakye, Corporate-registry analyst; builds beneficial-ownership graphs for litigation. Commissioned to build the graph and identify any common controlling node, without being told that a second analyst existed; report published in full February to May 2026 Resolved 137 of 140 to a named ultimate holder. Says the word 'control' should not be used of any of it.
- Emrys Latham-Oduya, Corporate-registry analyst, second graph, blind. Given the identical brief and the identical fee, in advance, with no contact of any kind with the first analyst February to May 2026 Reached the same node and then wrote eleven pages arguing that the node means nothing. Printed entire.
- Published stewardship voting logs, three managers, 2025 season, Resolution-level voting records with decision timestamps. Downloaded as published from each manager's own site; 41,700 rows filtered to the 140 outlets and their parents June 2026 The timestamps are the source of the forty seconds and are the weakest thing in this story.
- Dr. Ines Okwuosa-Faber, Head of investment stewardship, Alderney Index Partners. Interviewed by video call for fifty minutes, then sent the full draft, both graphs and our timestamp working June and July 2026 Disputes what the timestamps measure and objects to the word tired. Printed in full under Right of Reply.
- Registry extracts and persons-of-significant-control filings, 1,206 corporate entities across nine jurisdictions. Purchased at commercial rates, £2,880 in filing fees, and supplied to both analysts identically January 2026
What we could not confirm
- The three outlets where the graphs disagree. Both analysts stop at the same nominee structures in two jurisdictions and both say the beneficial owner is not determinable from public filings. Establishing it would mean litigation we did not fund, and we are not going to describe three unresolved outlets as a rounding error.
- What a timestamp measures. We take the median of forty seconds from the decision timestamps in the managers' own logs. Alderney says the timestamp records when a decision was entered into the system, not when it was made, and that the substantive work happened in engagement meetings months earlier which are not logged at all. We cannot see those meetings and neither can anyone outside the firm.
- Whether 'controlling node' means anything here. It is a term from graph analysis and we have let it do work in this story that a lawyer would not permit it to do. A four per cent holding that never votes against management may be the largest holding in the room and still not be control in any sense a court would recognise.
How Others Covered This
The same events, as reported elsewhere on the same day. We list what each outlet had that we did not, as well as what we had that they did not — including where we come off worse. Why we print this.
-
The Hollow PostTHE THREE FIRMS THAT OWN YOUR NEWSPAPER — AND WHY NOBODY WILL SAY SO
Treated the holdings as concealed, and the absence of board seats as evidence that the control is being hidden rather than not exercised.
Had that we did not
The percentages, accurately, from the same public filings we used.
Left out
That the votes attached to those holdings went with management 1,111 times out of 1,182, which is published by the holders themselves, in a spreadsheet, for free.
-
Signal & LedgerStewardship Capacity: Nine Analysts, 3,400 Issuers, One Proxy Season
A resourcing story for compliance readers, with a chart of votes per head and a caution about seasonal peaks.
Had that we did not
The ratio, and better arithmetic than ours on the resolution counts, which we have adopted with thanks.
Left out
That the issuers in question are newspapers. The word appears once, in a subordinate clause, on the second page.
-
They Buriedthis newspaperWe Looked For The Family. We Found Nine People And Forty Seconds Each.
Commissioned two ownership graphs blind, at the same fee, paid in advance, and published both including the one that says we have found nothing.
Had that we did not
Both graphs, the three outlets where they disagree, the voting logs, and the second analyst's dissent entire.
Left out
We commissioned two reports hoping one of them would find a family, and when neither did we ran the dissent at the foot of the page rather than at the top, where its argument belongs. Our headline still promises a hand. — V. Ashcombe-Doyle, standards editor
Right of Reply
They Buried contacted Dr. Ines Okwuosa-Faber, head of investment stewardship, Alderney Index Partners on 19 June 2026 by video call, and again on 9 July with the full draft, both analysts' graphs and our timestamp working attached. Replied 28 July 2026, asking that it run entire. Printed unedited, including her final paragraph about this newspaper's own pension scheme.
I will answer the number first and then the adjective, because the adjective is what your readers will remember and it is the part you got wrong.
Forty seconds. You have taken the decision timestamps out of our published log and computed a median across the busiest five days of the year. The arithmetic is correct. What it measures is not what you say it measures. The timestamp records the moment a decision is entered into the voting platform. It does not record when the decision was made, because the decision was very often made in March, in a meeting, about a company whose resolutions do not come up until June. In that week my team enters several thousand decisions that already exist. You have timed the typing.
I accept that you cannot verify this. I accept that it is convenient for me to say it. I would only ask you to notice that you have built a headline on a figure whose meaning you concede in your own third caveat that you cannot establish, and that you have put the caveat where caveats go.
On the ninety-four per cent. Yes. On media-sector resolutions in 2025 we supported management 1,111 times out of 1,182. I am not embarrassed by that figure and I will tell you why. The overwhelming majority of those resolutions are the re-election of directors who have done nothing wrong, the appointment of auditors nobody objects to, and the approval of accounts that have been audited. A firm that voted against a fifth of those in order to look independent would be behaving worse, not better, and would be doing it with other people's retirement money.
You write that voting against requires a rationale and voting with management does not, and you present this as a thumb on the scale. It is a thumb on the scale. It is deliberately a thumb on the scale. The alternative is a team that opposes cheaply, and opposition that costs nothing is worth nothing. If you think nine people is too few to carry that burden then say so plainly — I would not necessarily disagree with you, and I have made that argument internally more than once — but do not dress a resourcing question up as a conspiracy that you then generously decline to allege.
Now the adjective. You call us tired. Nine people, three thousand four hundred companies, and the word you reach for is tired. It is a good word. It does a great deal of work in a sentence and it costs you nothing to write. It also tells your reader that the outcome would be different if we were better rested, which is not true and which you do not believe. We vote with management because most resolutions are unobjectionable, not because it is late in the afternoon.
And since we are being candid with one another: your own pension scheme holds units in two of the three funds you have written about, which you disclose, correctly, at the foot of the page. Every reporter on this story is an indirect beneficial owner of the newspapers they are describing as passively owned. That is not a gotcha. It is the actual condition of the thing you are trying to report, and it is a better opening paragraph than the one you have.
Published unedited under our right-of-reply guarantee.
How was this story?
We publish the result, whatever it is. Reader verdicts appear on the front page and in our newsroom metrics.
Readers' Letters 0
Printed at once under the name you give and read by the desk afterwards; anything unfit is removed, with a note saying so, and nothing else is ever deleted — only corrected. Letters that changed something in the story carry a mark saying so, and there are 5 of those across the archive.