Alternative Assets in Public Pensions: July 2026 Update

PLUS: Correlations Update! (The correlations are very weak)

Jul 29, 2026

After my recent post on the Equable Institute report on public pensions, I was reminded that I should update my periodic look at U.S. public pensions in alternative assets.

My last look was in December, but many pension plans were missing data for FY2024 in my favorite data source:

Alternative Assets in Public Pensions: December 2025 Update

Mary Pat Campbell

December 17, 2025

Read full story

Links to older posts on this subject, as well as a spreadsheet with the current stats, will be at the bottom of this post. This includes the correlations info.

Public Plans Allocation to Alternative Assets

For simplicity, what is not considered alternative assets:

publicly-traded equities
publicly-traded bonds and similar fixed-income securities
cash

That’s it for “core”, traditional investments.

My data comes from the Public Plans Database, and I grabbed it today.

The types of assets considered “alternative”:

private equity
real estate
hedge funds
commodities
assorted alternatives/other

Trend of Allocation, 2001 - 2024, U.S. Public Pension Plans

I lightened the color on the data points I had from last time. I considered the color was too dark and made it difficult to see the lines.

Key points:

For the plans for which we have allocation information, a number allocate only to “core” asset classes. It’s a stubborn few, alas, as we see even the 25th percentile has increased to even a 20% holding of alternatives. On the other extreme, a few plans have gone to very high allocations, as we’ll see in the ranking table below.
In the December 2025 update, I could blame the decrease seen in the median allocation from 2022 to 2024 on the number of plans I had data for, but I have data for 228 in 2022, 224 for 2023, and 212 in 2024. That can move the median a bit, but not much.
That median for FY2024 sits at 29.5% of assets, not much different from the 27% of “valuation risk” exposure mentioned by the 2026 Equable Institute report, and theirs may have been asset-weighted. Mine is weighted by plan count, not assets. (Being lazy)
Yeah, you see some dots somewhat super-imposed over each other. You’ll see what that is when I show you part of the ranking table below.

Public pension plans with the highest allocations to alternative assets, 2017-2024

You can see the issue with many of the plans in the top 25.

Ranks 1-25

The issue is my data is by plan, not by the shared investment management, which evidently is giving the same allocation to alternatives to, for example, both Maine Local and Maine State & Teacher plans, even if that is not necessarily appropriate from an asset/liability management standpoint.

Bah.

When I look at some of the correlations, that means it’s overdetermined, in that “different” pension funds will have the exact same allocations to alternatives and the exact same investment return experience. They’ll almost definitely have the same investment return targets. However, they won’t necessarily have the same funded ratios.

I added a column this year, in checking how much the allocation to alternatives increased over the full period. In some cases, it didn’t increase much, because the plans always had high allocations… in some cases, HOLY CRAP, Houston Firefighters, what did you do?

I will point out not everybody increased their allocations… and, indeed, when I scrolled down further in the list…

Ranks 26-50

The Arizona Public Safety and State Corrections Officers (#34 and #35) both saw their allocations decrease by about 17 percentage points from 2017 to 2024.

But I also point you to the Wyoming plans immediately below, which increased their allocations 16.5 percentage points over the same period. And not smoothly, either. Just some big ole jumps there. What’s going on?

You can understand why some get a big jumpy about these large allocations to these asset classes.

Correlations: Weak Between Allocations, Returns, Funded Ratios (Either Way), Etc.

There is no reason to waste any time on 1-year allocations or returns.

The problem, as you can see above, is that many of these plans have been changing their allocations to alternatives year-by-year… or, more, likely, their market-valued assets (especially public equity) is volatile on the balance sheet, and thus that nudges the allocations around. They don’t want to keep rebalancing, and given the cash flow status of many of these pensions, they don’t have enough turnover to nudge the allocations to hit targets.

ANYWAY.

To simplify, I took the 10-year average allocation to alternatives and compared against the 10-year returns.

There isn’t much of a correlation.

I also tried some other variables that might be relevant.

The strongest correlation, which is still weak, was between the 10-year return and the 2024 funded ratio. That makes some sense. Here is what that looks like.

Most of the plans are bunched in the middle, so that doesn’t help the correlation much. It’s the low-return plans with not-great funded ratios, and the high-return plans, with somewhat better funded ratios, that give a bump up to the correlation.

There was essentially no correlation between the return assumption and the funded ratio, but no surprise there — there was little variability between return assumptions.

This is a useful graph — checking out how the actual 10-year returns compared against the 10-year return assumptions:

The blue dots “line up” because most return assumptions are chosen by plan sponsors to be nice, round numbers: 7% or 7.5%.

They don’t choose something weird-looking like 5.234234% (usually), though that might be where a calculated actual result lands.

To be sure, I was comparing against the 2024 return assumption — as the Equable Institute noted, and I’ve shown in my own graphs, the return assumptions have been sliding downwards over the last two decades.

But if you weren’t able to hit that downward-sliding target over ten years… maybe you need to slide it even lower.