The Reservoir · a chart of GIC
Survey 2026 · soundings in disclosure
legend

The Foreign Harbour

The same instrument, run in the open: Norway

Everything else on this chart measures GIC against what GIC chose to say. This district measures it against another country. Norway runs a sovereign wealth fund built on the same logic — save the windfall, invest it abroad, spend only the returns — and publishes almost everything GIC withholds. The comparison settles a question the rest of the site can only raise: secrecy of this depth is a choice, not a requirement of the job.

What Norway publishes about the same instrument official

DisclosureNorway (GPFG)GIC
Total size of the fund Published continuously, to the krone Withheld by policy
Only the Government may disclose it; it never has
Every individual holding Published annually, company by company, with voting record Not published
Visible only where a foreign regulator compels a filing
Annual return Published each year, with the single-year figure Published as a rolling 20-year average only
No single-year return has ever been disclosed
Return by asset class Published Not published
Asset mix percentages are given without returns
Management costs Published, in basis points Not published
Real-time fund value Published on a public ticker Not published
There is no ticker to point at
Compiled from each fund's own published material. Norway's figures: nbim.no, read 2026-09-06. GIC's practice is documented district by district across this chart.

The shape of a fund that shows its work official

Because Norway publishes its allocation, its portfolio can simply be drawn — no inference, no bounding, no italics. It reports NOK 23.0 trillion spread across 69 countries, and names the amount in each:

CountryNOKShare
United States12.1 trn52.9%
Japan1.4 trn6%
United Kingdom1.2 trn5.2%
Germany0.8 trn3.7%
France0.8 trn3.3%
Canada0.7 trn3.2%
Taiwan0.6 trn2.7%
South Korea0.6 trn2.6%
Country allocation published by NBIM, in NOK, read 2026-09-06. Every figure on this page is roman type: nothing here had to be inferred.

There is one more row worth reading slowly. Norway discloses that it holds NOK 223.5 billion in Singapore — a figure Singaporeans can look up about their own market. The reverse lookup does not exist. GIC publishes no country allocation at all, so no Norwegian can ask the same question in the other direction.

Norway, restated in GIC's own metric inferred

GIC discloses exactly one performance number: a rolling twenty-year annualised return. Because Norway publishes its annual returns, that same statistic can be computed for Norway — and it is the only like-for-like answer available to the obvious question, which is whether GIC's number is any good. Neither fund publishes this comparison; it follows from combining two published series.

GIC financial yearGIC, 20y nominalNorway, 20y nominalDifference
FY2020/21 · vs NBIM 2020 6.8% 6.0% -0.8
FY2021/22 · vs NBIM 2021 7.0% 6.9% -0.1
FY2022/23 · vs NBIM 2022 6.9% 6.3% -0.6
FY2023/24 · vs NBIM 2023 5.8% 6.5% +0.7
FY2024/25 · vs NBIM 2024 5.7% 6.7% +1.0
FY2025/26 · vs NBIM 2025 5.6% 6.9% +1.3

The two funds tracked each other closely through 2020/21, when GIC was ahead by 0.8 points. Since then they have separated: GIC's twenty-year figure has fallen from 6.8% to 5.6%, while Norway's has risen to 6.9%. On the one measure GIC has chosen to publish, the transparent fund is now 1.3 points ahead.

Read this with its limits. Currency basis differs from GIC: NBIM reports in the fund currency basket, GIC in US dollars. Period differs: NBIM reports calendar years; GIC's financial year ends 31 March. These are nominal returns; GIC's headline disclosure is a real return, and its nominal figure is published only from FY2020/21. The current year is excluded until it has ended: NBIM's chart shows it as a year-to-date figure, which is not an annual return. The currency-basis difference is the largest of these and cannot be corrected from public data — the comparison is indicative, not decisive, and is set in italic for that reason. NBIM annual returns retrieved 2026-09-06; the twenty-year windows are computed here from that series, not taken from any summary figure. NBIM's chart also shows 2026 so far (9.44%); that is a year-to-date figure, not an annual return, and is excluded until the year ends.

What the contrast does and does not prove

Norway's openness is possible partly because of what its fund is: a single pension vehicle, invested almost entirely in listed markets, with a mandate that is itself an act of parliament. GIC's stated case for secrecy — that publishing size would expose a small country's defences and invite speculation against it — is a real argument, made in public, and this chart does not pretend to settle it. What the comparison does establish is narrower and harder: the technical obstacles are not the reason. A fund of comparable scale publishes to the krone, and functions.

The honest summary is that GIC and Norway resolve the same tension differently, and only one of them can be checked by the people whose money it is. GIC's rolling 3.4% is a real disclosure; it is also, by construction, the least falsifiable number a fund can publish. Norway's 69-country table can be wrong in 69 ways, and that is the difference.

GIC's own reasoning is set out in The Dam; the 141 companies a foreign regulator forced into view are in The Glass Pool.

Against doing nothing clever estimated

For over a decade GIC published a Reference Portfolio — 65% global equities, 35% global bonds — and reported what it returned, so a reader could ask the only question that really tests an investor: did it beat buying the market and leaving it alone? The FY2025/26 report retires that benchmark, so the comparison can no longer be made from GIC’s own disclosures. It can still be made from Norway’s.

NBIM publishes its benchmark index return as well as its own — a roughly 70/30 index, the closest public analogue to the anchor GIC dropped, run by a sovereign fund for the same purpose. Benchmark is fund return less relative return, both as NBIM prints them, deflated by the inflation NBIM reports on the same basis.

3.4%
GIC, 20-year annualised real, FY2025/26
Naive index, NorwayReal
since 1998 (28y)4.2%+0.8pp vs GIC
last 15 years5.4%+2pp vs GIC
last 10 years5.4%+2pp vs GIC

GIC’s twenty years falls between Norway’s fifteen- and twenty-eight-year windows. It is below the naive index on both, so the answer does not depend on which window is chosen — which is why both are shown rather than whichever one reads better.

The honest reading is narrow. NBIM’s benchmark is roughly 70% equities / 30% fixed income. GIC’s retired Reference Portfolio was 65/35. Close, not identical. Currency basis differs: NBIM reports in the fund’s currency basket, GIC in US dollars. Nothing here is like-for-like enough to call a verdict on GIC’s management, and this district does not call one. What it can say is that the question has a public answer, that the answer is not flattering on any long window, and that it stopped being answerable from GIC’s own pages in the same year the numbers looked like this.

Norway’s figures: NBIM, derived in scripts/fetch-benchmark.mjsbenchmark.json, which carries every caveat in full. GIC’s figure is its own published rolling 20-year real return.