The Rising Water

Nothing on this chart was always knowable. Each figure entered the record on a date — a filing crossing a threshold, a report published, a Budget read out. Drag the year below and the chart reports only what had been disclosed by the end of it. The water rises; the wall does not move.

The chart, as of 2026

What the slider is actually showing

Two different things rise at once, and the distinction matters. GIC's own disclosures grow slowly and by choice: one annual report a year, a rolling return, an asset mix. The compelled disclosures grow in bursts, driven by a foreign regulator's thresholds and by GIC's own deal-making crossing them. The line between the two is the line this whole chart is drawn along — and one side of it has moved far more than the other.

Note also what never appears at any position of the slider: a size, a single-year return, a full holdings list. Those are not late-arriving figures. They are the ones that never arrive.

The year the water fell

Everything above accumulates. That is the district's argument: disclosure only ever adds, and the slider exists to make the adding visible. The FY2025/26 report is the first entry that runs the other way, and the claim above has to answer for it.

For over a decade GIC published a Reference Portfolio and said what it was made of — GIC’s Reference Portfolio, which comprises 65% global equities and 35% global bonds — and reported its rates of return over 20, 10 and 5 years. That is a yardstick. It is what let a reader ask the only question that matters about an investor whose size is secret: did it beat a naive balanced portfolio, bought once and left alone?

The FY2025/26 report retires it. The refreshed framework is anchored by a Strategic Portfolio, described as three broad asset groups — Equities for growth, Fixed Income for income, Real Assets for inflation — and no weights are given for them anywhere in the document. The Reference Portfolio survives in a single sentence, in the past tense, listing what the 2013 framework used to contain: The Reference Portfolio: A market-based representation of the Client’s risk appetite.

What is still published: the Asset Mix of the GIC Portfolio — Equities 56%, Fixed Income 22%, Real Assets 22% at 31 March 2026. What GIC holds is disclosed as it always was, and in the same three groups the new framework names.

What is no longer published: the composition of the benchmark representing the Client's risk appetite, and that benchmark's returns. The portfolio is still visible. The yardstick it was measured against is not, and with it goes the comparison a reader could previously make unaided.

Whether that is a fund retiring a stale benchmark for one better matched to how it actually invests, or a published anchor being withdrawn, is not a question this chart can settle — GIC gives its reason as a framework refresh, and no figure here contradicts it. What the chart can say is narrower and still worth saying: for the first time in the record, a thing that had been published stopped being published, and the depth of this district's own claim moved.

What else stopped being said

The finding above was made by reading. That should not be what it takes, so the same question is now asked of the whole archive: 17 annual reports, every capitalised term whose head is a domain noun, flagged when a concept that ran for years stops — or collapses.

The three that collapsed together are not three findings. They are one: the 2013 framework was built from exactly three named components, and the FY2025/26 report retires all of them at once — which is what a framework being replaced looks like from outside, and what the hand-read finding above saw only one third of.

The check is deliberately dumb, and that is the point. It was not told what to look for and it does not know what a Reference Portfolio is; it counts defined things and notices when one stops being used. It would have surfaced this in FY2025/26 with nobody watching, and it will run again when FY2026/27 lands. Absence from a report is not proof a fact stopped being true — only that the term stopped being printed, which is why every hit is a lead to read rather than a finding to publish.

Two kinds of disclosure, and only one of them is safe

The counters above are not the same kind of thing, and the year just described is what separates them.

Filings are compelled. They exist because a foreign regulator requires them of anyone crossing a threshold in US-listed stock, and a regulator’s archive does not retract: once a document is filed it stays filed, and the count can only rise. Nothing GIC decides can reduce it.

Reports are voluntary. GIC chooses to publish one each year, chooses what goes in it, and — as FY2025/26 shows — can choose to stop publishing something it published before. The count still rises, because a report was still published. What was in it fell.

So the instrument on this page has a blind spot, and it is worth naming rather than quietly fixing. The “GIC annual reports published” counter cannot see a regression. It went up in the one year the disclosure it measures went down, because it counts documents and not what is in them. Every reading of this slider before FY2025/26 was therefore slightly too generous, and no reading of it will ever be too harsh.

That is the honest statement of what this district measures: the compelled series is a floor that only rises, the voluntary series is a count of publications and not of disclosure, and the difference between them is exactly the space in which a fund can stop saying something without any number here moving.