Does a bigger fund publish more? Size against transparency

The short answer: somewhat, and mostly among funds that publish their own size. Across the 166 operating sovereign and state funds where we have both a size figure and a full transparency score, larger funds publish more on average. The link is moderate at best, the middle size band scores highest, and several of the biggest funds in the world publish very little.

What we measured

Every fund profile on this site carries a transparency score out of 10. It is our own rubric, built from 10 items a reader would want to find in public: total assets, annual return, long-term return, asset allocation, a geographic or sector split, a holdings or deal list, audited statements, governance and mandate documents, named leadership, and a regular reporting cadence. Each item scores 0, 0.5 or 1, backed by a cited source in the profile. It is not any outside index or score. It measures what a fund discloses, not how well it is run.

For this study we used 206 scored funds. That is every profile, each with all 10 items scored on the 0, 0.5 or 1 scale and a total that matches the sum of its items, minus the 25 proposed or not yet operating funds that hold no money yet. Of the 206, 166 also have a size figure we can express in US dollars.

Finding 1: the middle band scores highest

We put the 166 funds into four size bands and took the mean and median score in each.

Source: our dataset. Basis: our 10-item score; size bands use the headline size figure shown in each fund profile, in US dollars.

Scores climb from the smallest band to the US$10bn to US$100bn band, then fall back. The over-US$100bn group averages less than the band below it, because it holds a cluster of very high scorers and a tail of low ones.

Every band runs from near the bottom of the scale to near the top. A fund of under US$1bn can score 9.5. A fund of over US$100bn can score 1.5. Size alone tells you little about what you will find on a fund's website.

Finding 2: the statistical link is real but modest

We ran a Spearman rank correlation between size and score for the same 166 funds. The result is rho 0.292 with a p-value under 0.001.

In plain words, there is a genuine tendency for bigger funds to score higher, but it is not strong. A rho of 0.292 means size explains only a small part of the ordering. Most of the variation in scores comes from something else.

Source: our calculation from our dataset. Basis: rank correlation, so it uses the order of funds by size, not the dollar amounts themselves.

Finding 3: where the size figure comes from matters more

Our size figures come from two kinds of source, and the split tells a clear story.

Source: our calculation from our dataset.

Among the 134 funds whose size comes from their own reports or from government, bigger funds do publish more, and the link is stronger than in the full group. Among the 32 funds whose size we can only show from a press report or an estimate, the link is weak and not statistically significant, and the average score is far lower.

That makes sense. A fund that does not publish its own size is already missing the first item on our checklist, and it tends to hold back on other items too. Funds that publish nothing about their size at all sit outside this table, and we come back to them below.

This data does not prove why bigger self-reporting funds publish more. Larger funds often have bigger teams, outside stakeholders and listed holdings that bring reporting duties, but we have not measured those things here.

Finding 4: the biggest funds split into two camps

Of the 28 funds in the over-US$100bn band, 14 score 8 or more and 9 score under 5 (source: our dataset). Only 5 sit in between.

Funds in this band that publish audited figures, long-run returns and detailed holdings include Norway's Government Pension Fund Global, Temasek, CPP Investments, Japan's GPIF, the Future Fund and Mubadala.

Others in the band do not publish their total assets at all, so the size figure in their profile is a press estimate. Our profiles record that the Kuwait Investment Authority and the Qatar Investment Authority do not disclose their assets.

Finding 5: smaller funds are not all closed

Of the 33 funds under US$1bn, 11 score under 5 and 5 score 8 or more (source: our dataset). That is a weaker showing than the bigger bands, but the range is wide: the top score in this band is 9.5.

A further 40 scored funds have no comparable size figure in their profile, so they sit outside the bands. Their average score is 3.51 (source: our dataset). Not knowing a fund's size is itself a strong sign that it publishes less.

What this means

If you are trying to judge how much you will be able to learn about a fund, its size is a poor guide. Better guides are whether it publishes its own size, whether it belongs to the IFSWF, and which region it sits in. We look at those in which regions publish the most.

Caveats

1. Our score, our rubric. The score measures public disclosure, on the date shown in each profile. It is not a rating of governance, performance or quality.
2. Mixed size bases. The size figures behind the bands mix audited accounts, official statements and outside estimates, and they mix portfolio values, balance sheets and net assets across sovereign funds, pension funds and holding companies. We use them only to sort funds into bands and to rank them. We never add them up.
3. Correlation is not cause. Nothing here shows that growing makes a fund publish more, or the reverse.
4. Counts move. Profiles are updated as funds publish, and this page is refreshed with them.

Methods note

Population: 206 operating funds (every profile except the 25 proposed or not yet operating funds), each scoring all 10 items on the 0, 0.5 or 1 scale with a stated total equal to the item sum; 166 with a size figure in US dollars. Score: sum of the 10 item scores. Bands: under US$1bn, US$1bn to 10bn, US$10bn to 100bn, over US$100bn, using the headline size figure in each profile. Where a profile gives only a local-currency figure, we convert it at the rate the profile states or at a reference rate for the figure's date; the result is used only to place the fund in a band and to rank it. Figures that are not a measure of assets, such as announced capital, commitments or activity totals, count as no comparable figure. Correlation: Spearman rank, two-sided p-value. Full method: methodology.