Calls Ledger: What We Said, What Held

The calls ledger defined: what it records and what it excludes#

A calls ledger is a dated register of the statements Victaura Research has published about the future or about the state of the world, each scored against evidence. Every entry carries four fields. The article and its date. The exact claim. A status of held, revised, wrong or open. And the dated proof that moved the status. This note applies the instrument to nine calls drawn from four articles, as of 9 October 2026. It is the first ledger, and it will be updated.

The four statuses are defined before the scoring, so the scoring cannot bend to the result. Held means the evidence arrived and matched the claim. Revised means the claim was partly right, or the facts it rested on changed, and the text was updated. Wrong means the claim contradicted facts that were already available on the day, or that arrived later. Open means the test date has not come. A call can be held on timing and revised on magnitude. The ledger then carries two rows, and it does so below.

The reason to publish the register is mechanical. The allocator reading a research house needs to know how often the house moves its position, and why. A correction in a revision note answers that for one article. A ledger answers it across articles. Each entry links to the piece it comes from, so the principal can open the original text and judge whether the scoring is fair.

The scope is deliberately narrow. Only claims about the future, or about the state of the world on the day of publication, are logged. Nothing here concerns returns, prices or the performance of any project, and no internal forecast is recorded. Where an article was later corrected for attribution or sourcing, that correction sits in the article's own revision note and is not repeated here as a market call. The register draws on four articles. Dozens of other pieces are not in it, and the weaknesses section returns to that.

Is Wynn Al Marjan Island now due to open later than Volume 4 assumed?#

Wynn Al Marjan Island is now scheduled to open in September 2027, and that date sits inside the window Volume 4 set out. On its second-quarter 2026 call on 4 August 2026, Wynn Resorts moved the opening of its resort on Al Marjan Island, Ras Al Khaimah (RAK), to September 2027 and raised the project budget by about USD 600 million. The date is Wynn's own and appears in its results release of 4 August 2026 (broker-reported); iGaming Business, CDC Gaming, Covers and World Casino Directory agree on the size of the increase (directional, press-reported).

The call under test sits in the Volume 4 dossier on RAK and the Wynn effect, published on 28 May 2026. The dossier quoted Wynn's chief executive on the first-quarter call of 7 May 2026, saying the company expected a 'modest delay' and would quantify it in the coming months. The dossier then added its own scenario. Opening in the second half of 2027 at the earliest, with 2028 as the fallback. The company's guidance before the delay had been the first quarter of 2027, and its 2022 announcement had scheduled the resort for 2026.

On the date, the call held. September 2027 is in the second half of 2027. It is not 2028. The scenario was a floor and a fallback rather than a point estimate, so the scoring is generous by construction. A range wide enough to catch most outcomes earns little credit for catching one. The ledger records the row as held, with that caveat attached to it. The resort has not opened, so the row rests on company guidance until it does.

On the word 'modest', the record is less clean. The word was Wynn's, quoted rather than endorsed, but a quotation placed in a dossier carries weight with the reader. The new date is roughly six months after the old guidance. The budget moved from about USD 5.1 billion to about USD 5.7 billion in the figures this house used later, and some outlets round the new total to USD 5.8 billion. Management attributed roughly half of the increase to the Iran conflict (directional, press-reported, unaudited). Whether six months and USD 600 million is modest is a judgement. The ledger marks the row revised, because the facts the word rested on changed.

The cost line was never a call, and the ledger does not treat it as one. Volume 4 did not forecast the budget. Scoring the increase as a miss would manufacture an error. Ignoring it would hide the fact that the delay arrived with a cost attached, and an allocator underwriting an anchor-project timeline needs the two figures together.

What did the 26 August Ras Al Khaimah note get wrong about the ceasefire?#

The 26 August 2026 note read the US-Iran de-escalation as settled, and that reading was wrong. The first version of the Ras Al Khaimah forecast re-read after Wynn's reset described the Strait of Hormuz as reopened, the corridor as de-risked and the ceasefire as holding for now. The note's own revision notes say the reading was already wrong on the day of publication. The ledger enters it as wrong and gives it the same space as the rows that held.

The failure was not a surprise event. According to the note's revision, the June memorandum of understanding between Washington and Tehran had broken down on 8 July 2026, seven weeks before publication. A report the note itself cited, from Al Jazeera on 23 July 2026, said that hostilities had resumed. The evidence needed to avoid the error was in the note's own bibliography. It was not weighed against the headline.

The status as of 27 September 2026 is that no truce was in force. During the week of the UN General Assembly, Iran offered to reopen the Strait within seven days if the United States lifted its naval blockade of Iranian ports, waived sanctions on Iranian oil sales and observed a ceasefire that would include Lebanon. President Trump rejected the proposal on 26 September 2026, according to Associated Press and AFP copy. The revised note cites an Al Jazeera report of 27 September 2026 for the same rejection, which is a later report of the same event, not a different date.

The follow-on entry is open, and it is dated. As of 27 September 2026, the note states that no truce was in force. That is a statement of state, not a forecast. It stays open until a signed arrangement or a documented incident changes it, and the ledger will update it with the date and the source. It will not use the words resolved or over, because neither describes a conflict whose status has changed several times since February 2026.

For the allocator, the cost of the error is the useful measure. The correction reached the title, the deck, the geopolitical sections and three takeaways. A status error on a conflict propagates into every conclusion built on it, including insurance cost, supply chains and construction timing. It also bears on Wynn's own explanation of its delay, since management tied part of the budget increase to the same conflict.

How did the Volume 3 branded residence count move from 611 to 903?#

Knight Frank now counts 903 branded residence schemes worldwide at the end of 2025, on a wider survey than the one behind the 611 Volume 3 used. The 903 comes from Knight Frank's Residence Report 2026/27, as set out in the firm's newsroom release of 14 September 2026 (industry estimate). The survey verified nearly 1,800 live and pipeline schemes from more than 200 brands across 90 countries. On Knight Frank's own count, the sector stood at 354 schemes in 2015.

Volume 3 on branded residences and the luxury market, published on 28 May 2026, set Knight Frank's 611 against Savills' 910. It read the 299-scheme gap as the numeric form of its two-markets thesis, that the global count covers two markets sharing a name. The 611 came from a luxury-only screen in Knight Frank's 2025 survey, covering nearly 80 brands in 83 countries. The 910 was the end-2025 total Savills expected, up from 764 recorded in December 2024. The dossier's title carried the 910.

The gap has closed to seven schemes, and the closing came from scope, not from the market. Knight Frank widened its universe from nearly 80 luxury brands to more than 200 brands. Savills' 910 was an expectation rather than a verified count. The two figures sit close together now, but they were never measuring the same thing, and the proximity of 903 to 910 may be coincidence. The ledger enters the row as revised.

The author says the thesis did not rest on the count, and the ledger has not tested that claim. The 27 September 2026 revision note on Volume 3 states that the two-markets argument rested on operator counts, fee stacks and premium bands. That is the author's own account. The ledger can confirm only that the count-based illustration was weaker than it was presented. It cannot confirm that the thesis itself held, because the operator and fee evidence was not re-tested for this note.

The grade on the 903 is industry estimate, the grade the method gives to a property firm's research count. The full report is registration-gated, so this note read the figure through Knight Frank's newsroom release and press coverage. The release does not say whether the 903 counts only operating schemes or includes some in the pipeline. Until the report is read in full, the figure carries that caveat, and so does any comparison built on it.

Open row: where the branded residence count goes after 903#

Two forward counts from Volume 3 remain open, and the new Knight Frank edition does not settle them. Volume 3 cited Knight Frank's earlier forecast of 1,019 schemes by 2030 and a Savills arithmetic ceiling of 1,747 by 2032, which assumed that all 837 contracted projects complete. Knight Frank's 2026 edition now says the sector will pass 1,000 developments in 2026 and, on current pipelines, reach 1,800 schemes and more than 300,000 units by 2031 (industry estimate).

The forecasts are not comparable, which is the point. The 2025 edition's 1,019 was built on the luxury screen. The 2026 edition's 1,800 is built on the wider universe. Savills' ceiling is gross of attrition. A reader who lines up 1,019, 1,747 and 1,800 sees convergence. A reader who checks the scope sees three different instruments, measured on three different dates.

Attrition is the variable none of the published figures shows. Volume 3 put attrition between contracted and completed schemes at ten to twenty per cent. As published, it called that band typical for branded residences. Its 27 September 2026 revision relabelled the band a Victaura Research assumption, because no cited source contains the rate. It remains an assumption. Applied to the 837 contracted projects, it would take the 1,747 ceiling down to roughly 1,580 to 1,663 (directional, our arithmetic on our own assumption). That reads as a range for the ceiling, not as a forecast.

The test dates are known. Knight Frank has set its own marker, which is the sector passing 1,000 schemes within 2026. That is checkable on the next edition. The 2030 to 2032 figures cannot be checked until counts for those years are published, long after a purchase decided on them today. The ledger marks the row open, for both counts, and will move it only when a dated publication lets it, with the scope of the new count written beside the old one.

Where did the ECB and Federal Reserve stand when the 12 September note ran?#

When the 12 September note ran, the ECB deposit facility rate was 2.25% with a rise to 2.50% already decided, and the Federal Reserve's September decision had not been taken. The ECB rate table shows 2.25% effective 17 June 2026 and 2.50% effective 16 September 2026 (measured, ECB). The Governing Council took the September decision on 10 September, two days before the note on the interest rate path and prime property ran. The decision date comes from the note's own revision and from press calendars, since the ECB table gives effective dates only.

The ledger enters the ECB row as wrong, and the error is a state error, like the ceasefire row. The note made no rate forecast. It said in terms that it did not predict the next move. What it got wrong was the state of the world. It treated the ECB's next step as open when the step had already been decided. That is a verification error. Nobody forecast anything, and the primary table was one page away.

The Fed row is different again. On 12 September the Federal Reserve's September decision had not been taken. The note treated it as unknown, which was correct on the day. On 16 September the Committee raised the target range by a quarter point to 3.75% to 4.00%, by a vote of 12 to 0 (measured, Federal Reserve). The text was updated on 27 September. The ledger enters the row as revised, since the original statement of uncertainty was accurate when written and the facts have since changed.

The path into those decisions matters for scoring. The Fed held the range at 3.50% to 3.75% on 29 July 2026 by 9 to 3, with three members dissenting in favour of a quarter-point increase (measured, Federal Reserve). The dissent was a public signal that a hike was live. The ECB's deposit rate had sat at 2.00% since 11 June 2025 before moving to 2.25% in June 2026 (measured, ECB). Neither bank had pre-committed to a path, which is why the note was right not to forecast.

The note's strongest line was its refusal to forecast, and the ledger credits that only as an absence. The note stated that nothing in it should be read as a prediction of the next decision. The ledger cannot score a claim that was not made. It can record that the refusal was consistent across the piece, that the ECB error came from a stale fact and not from the stance, and that the Fed row was overtaken by a later decision.

Open row: the Volume 4 rate assumption for Ras Al Khaimah#

Volume 4's base case assumed US policy rates would fall to 3.00% to 3.50% by 2027, and the first dated evidence runs against it. The 3.00% to 3.50% band stood in the base case as published on 28 May 2026. An October 2026 revision of the dossier replaced it with a qualitative condition, that the Federal Reserve cuts rates, and the ledger scores the call as published. The Fed range was 3.50% to 3.75% when the dossier ran on 28 May 2026. It is now 3.75% to 4.00%. The assumption is for 2027, so the row stays open. The direction since publication, however, is up by a quarter point (measured, Federal Reserve).

An open row can still be running against its call, and the ledger says so. The distance between today's range and the assumed 2027 range is at least 25 basis points, measured from the 3.75% floor of the current range to the 3.50% ceiling of the assumed one. At the widest it is 100 basis points. Either figure is arithmetic on published ranges (measured inputs, directional conclusion).

The test is written down now, so it cannot be adjusted later. If the target range at the last decision of 2027 sits inside 3.00% to 3.50%, the row moves to held. If it sits above, the row moves to wrong. A range straddling the boundary moves it to revised, with the overlap stated. A range below 3.00% also moves it to revised, with the gap stated. The date is chosen because the dossier's assumption was framed as an end-state by 2027 rather than a path.

Rates are one input to the Ras Al Khaimah read, and this row does not stand for the whole. The same base case also assumed that demand holds and that regional escalation does not worsen. The status of the conflict as of 27 September 2026, with no truce in force, bears on the second condition. The ledger does not re-score the base case as a whole. The dossier presented it as a scenario with several conditions, and scoring a conjunction on one leg overstates what is known.

Error types in the calls ledger: forecast, state and scope#

Three kinds of failure sit inside the nine rows, and they need different remedies. A forecast error is a claim about the future that did not happen. A state error is a claim about the present that was false when written. A scope error is a comparison across measures that were never the same. Both wrong rows are state errors. The revised count row is a scope error. No forecast row has yet been scored as a miss.

State errors are the cheaper ones to prevent and the more embarrassing to commit. Both were checkable on the day of publication, one against a primary rate table and one against a report already in the note's bibliography. A pre-publication check that opens the latest dated source for every status claim would have caught both. That check is the remedy. Whether it is applied is something the next ledger will show.

Forecast errors will come, and they are harder to avoid. Four rows are forward-looking claims about variables outside the author's control. They are the Wynn timing, the 'modest' characterisation, the Fed path assumption and the forward scheme counts. One held, one was revised and two are open. A sample of four cannot separate skill from luck, and the ledger does not claim otherwise.

Revised is the status most likely to be abused. A text that is quietly updated and then labelled revised can hide a wrong. The test applied here is whether the original claim was accurate on the day it ran. The Fed row passes that test. The count row passes on its numbers and fails on its framing. The 'modest' row passes as a quotation and fails as an implicit guide. Readers can disagree with each placement, and the table gives them the dated proof to do it.

Weaknesses honestly disclosed: what the calls ledger cannot show#

The sample is small and chosen by the author. Nine rows from four articles, selected because the evidence to score them exists now. Dozens of other pieces are absent, and articles whose calls have not yet been tested do not appear at all. A ledger built from testable calls overstates how testable the house's claims are, and the allocator should read the tally with that bias in mind.

The scoring is self-scored. The house that made the calls also judged them. The dated proof in each row is a primary document wherever one exists, so a reader can re-score every entry, but nobody independent has done so. Five of the nine rows concern Ras Al Khaimah, where Victaura holds an off-plan capital position, as the disclosure at the end of this note states. Readers who place a row differently, particularly the 'modest' row and the count row, have a legitimate case.

Several dates differ between sources, and none changes a status. Knight Frank's newsroom release is dated 14 September 2026, while some press coverage of the report ran later, Khaleej Times on 28 September 2026. AP and AFP copy dates the Trump rejection to 26 September 2026, while the Ras Al Khaimah note cites an Al Jazeera report of 27 September 2026 for it. Both are the kind of detail the ledger exists to surface.

Wynn's budget figures are press-reported and graded directional, and the press differs on the total. The new budget appears as about USD 5.7 billion in some outlets and about USD 5.8 billion in others. The split of the increase between conflict and other causes is management's own attribution and has not been audited. The six-month slip counts from Wynn's first-quarter 2027 guidance; some reports start from spring 2027, a phrase Wynn's releases do not use.

The conflict data is a news-flow reading. The ceasefire status draws on wire reports and daily coverage. It is not a primary government record, and the situation has changed several times since February 2026. Every status line in this note is dated for that reason, and it should be treated as a photograph at that date and not as a standing fact. The latest conflict status in this note is dated 27 September 2026, twelve days before the ledger's 9 October cut-off.

The Knight Frank figure is read second-hand. The full report is gated, so the 903 comes from the newsroom release. Whether it includes pipeline schemes is not stated there. Any conclusion about the sector's size that depends on the distinction between live and pipeline schemes is on hold until the report is read in full.

Reading the calls ledger as an allocator#

The tally is one held, three revised, two wrong and three open. The held row is a scenario floor. The wrong rows are state errors checkable on the day. Two open rows run to 2027 and beyond, and the third waits on the conflict. The principal using this house's work should check any status claim against its primary source and treat forward calls as scenarios until dated proof arrives. That advice applies to every research house, and this ledger is one house applying it to itself.

The next update dates are set by the evidence, not by a calendar. The ledger will move when Knight Frank's next edition appears, when Wynn reports on the opening, when the Federal Reserve and the ECB take their next decisions, or when the Hormuz status changes. Each update will carry its date and source. Rows will not be deleted. A wrong stays wrong, with the correction beside it, and a held stays held only until a later fact says otherwise.

Skin in the game disclosure. Victaura, through its parent Greystone B.V. (Netherlands), develops projects on Lake Como (Italy), in Nungwi, Zanzibar (Tanzania), on Gili Air and in Uluwatu, Bali (Indonesia), and holds an off-plan capital position on Al Marjan Island, Ras Al Khaimah (UAE). Readers should assume commentary may be influenced by, or benefit, these positions. This document is classified as marketing material under MiFID II Article 24(3). It is not investment advice.