What CalPERS, UC, and Oregon's own numbers say about beating a PE benchmark
Three of the largest institutional private equity programs in the country publish, in their own board materials, exactly how their PE portfolio did against its own benchmark. The numbers don't agree with each other — and one of the three openly admits to 25 years of underperformance before a strategy overhaul. If the allocators managing tens of billions of dollars each can't settle this in real time, that's not a knock on them — it's a fact worth sitting with before trusting any single point-in-time PE valuation.
The numbers, in their own words
Every large public pension and endowment discloses its private equity performance against a benchmark in its own regular board materials — a quarterly performance report, an annual program review, a committee memo. We pulled the most recent one from three of the largest: CalPERS, UC Investments, and Oregon's OPERF. Same idea in every case — value added (or excess return) over the program's own chosen benchmark, across multiple horizons — but each one measured as of a different date, against a different benchmark, because that's genuinely how each program reports it. That's not a flaw in this comparison; it's the point.
| Program | As of | 1yr | 3yr | 5yr | 10yr | Benchmark used |
|---|---|---|---|---|---|---|
| Oregon OPERF | 2026‑03‑31 | −19.0pp | −20.5pp | −7.6pp | −5.7pp | Russell 3000 + 300bps (qtr lag) |
| UC Investments | 2023‑12‑31 | −27.4pp | +0.3pp | −0.8pp | +2.2pp | UC's custom Private Equity Policy Benchmark |
| CalPERS | 2025‑12‑31 | +8.9pp | +6.5pp | +3.2pp | +0.3pp | State Street Private Equity Index (SSPEI) — All |
Figures are each program's own reported value added / excess return (portfolio return minus benchmark return, in percentage points) over the trailing period shown, as of the date shown — not a common as-of date across the three, and not a common benchmark. Source documents: Oregon's Q1 2026 OPERF performance report (Oregon Investment Council public meeting materials, March 2026); UC Investments' Q2 FY2023–24 performance memo to the Regents Investments Committee; CalPERS' 2026 Private Equity Annual Program Review (excess return vs. SSPEI All, as reported for the current blended portfolio).
The part that should give anyone pause
CalPERS' own program review doesn't just show a number — it names the reason for it. The review states plainly that prior to a strategy overhaul launched in Q4 2022, CalPERS' PE program had gone through 25 years of inconsistent or subscale commitments, and that the prior strategy "underperformed vs the PE index in all major time horizons." Its own numbers back that up: under the prior strategy, 3‑year IRR was 8.8% and 1‑year IRR was 9.0%; under the new strategy launched in 2022, the same figures are 30.9% and 36.9%. That's not a rounding difference — it's the same institution, the same broad asset class, producing a completely different track record depending on which multi-year strategy window you're standing in.
Oregon's picture looks nothing like either of the other two: its PE segment trails a purpose-built benchmark (public-market-plus-premium, the standard "PE should beat public equity by X" construction) across every horizon shown, by a wide and fairly consistent margin. UC's sits in between — a sharp one-year miss, close to flat over three and five years, and modestly ahead over the decade. None of the three tell the same story, because there isn't one story. There's a benchmark choice, a strategy era, and a point in time, and all three move the number a lot.
Why this matters for pricing a single fund
This isn't a "PE is good" or "PE is bad" post — it's the opposite. If three of the most resourced allocators in the country, each with their own investment staff, consultants, and board oversight, can't agree on whether their own PE program is beating its own chosen benchmark at any given moment — swinging from a −27pp one-year miss to a +9pp one-year beat depending on which institution and which quarter you look at — that is exactly the environment a single point-estimate valuation is built for. It's also exactly why we validate every model we ship out-of-sample and publish the result even when it's unflattering (see why most PE pricing models fail out-of-sample), rather than asking you to trust a number because the institution behind it is large.
Figures sourced from each institution's own public board materials, cited above, not from Equisect models. Research and software, not investment advice.
Where these numbers come from: each program's own published value-added-vs-benchmark disclosure (CalPERS 2026 PE Annual Program Review and equivalents)
Reported by the source, not computed by Equisect — there is nothing
for us to recompute, so we cite it rather than imply otherwise.