When the biggest pension deletes the policy portfolio: does the total portfolio approach actually work?

UAO Research · June 10, 2026 · Pensions · Feeds: Pension Strategy Watch

On July 1, CalPERS retires the strategic asset allocation model and begins running nearly $600 billion against a single 75/25 reference portfolio. It is the first US pension to do so, and the question its move forces is not philosophical but practical: across the funds that have already abandoned fixed asset-class targets, did the total portfolio approach (TPA) actually deliver higher returns — or did it mostly relocate where investment discretion and accountability sit? A universal owner deciding whether to follow needs to separate the evidence from the enthusiasm.

What the evidence says

The headline case is real, and it comes from a small, elite group. The Thinking Ahead Institute, working with Australia's Future Fund, studied 26 of the world's most sophisticated asset owners — together representing roughly $6 trillion — and found that the roughly one-third running a total portfolio approach outperformed those using strategic asset allocation by a meaningful margin over a decade. CalPERS, in adopting the model, cited a figure of 1.3% per year over ten years; the Future Fund's own updated work put the edge at 1.8% per year, revised down from 2.3% in the 2017 version of the study (Total Portfolio Approach hub, Thinking Ahead Institute, 2024–2025; Study gives evidence of value-add from TPA over SAA, Top1000funds, Aug 2024). Even at the low end, 1.3% compounded on a fund of CalPERS' size — about $598 billion as of March 31, 2026 — is nearly $8 billion a year, which is why the board moved.

The mechanism is intuitive. Under SAA, a board sets a target for each asset class every four years and measures staff against eleven separate benchmarks; capital sits where the grid says it should, not where the best risk-adjusted opportunity is. Under TPA, staff size every position by its marginal contribution to the whole portfolio's risk and return, judged against one reference portfolio — for CalPERS, 75% equities and 25% bonds (CalPERS, Nov 17, 2025). The model is not new: the Future Fund, New Zealand Super, Singapore's GIC and Canada's CPP Investments have run reference-portfolio or total-portfolio frameworks for years, and CalPERS has said it expects the switch to add 50 to 60 basis points a year.

Where it is contested

The uncomfortable part is selection. The 26 funds in the study were not a random sample; they were chosen for strong governance, scale and long horizons. The funds that adopted TPA were already among the best-run in the world — so the 1.3-to-1.8% gap may measure governance quality as much as the model itself. The fact that the estimate has fallen from 2.3% to 1.8% as the study was refined is itself a caution: the size of the prize is not settled, and a fund expecting 50–60bps should treat that as a hope, not a forecast.

There are structural objections too. A 75/25 reference portfolio is an equity-heavy benchmark for a liability-driven investor; in a sharp drawdown it can look like the fund took more risk than its members' obligations warrant, and the board has fewer asset-class scorecards to interrogate. Concentrating discretion in the investment team is the point of TPA — but it also concentrates accountability, which is why CalPERS' June 15–17 vote on tying staff pay to total-fund results matters as much as the model (Pensions & Investments, June 2026). The same logic — measure real contribution, not box-ticking targets — is visible elsewhere in the channel: Ontario Teachers' replaced its portfolio emissions-intensity target with a $70 billion climate-investment goal in February 2026, betting that what you build matters more than what a metric says (ESG Today, Feb 2026). The shared risk is the same: when you delete the silo target, you also delete the easy scorecard.

From the allocator's seat

For a CIO weighing the same move, three questions matter more than the outperformance figure. First, governance depth: TPA works where the board trusts staff with discretion and can still hold them to one number — a board that needs eleven scorecards to feel in control is not ready. Second, compensation: incentives have to track the total fund, or staff will quietly optimise their old silo; CalPERS is rewiring pay precisely because the model fails without it. Third, benchmark honesty: a reference portfolio must be defensible against the fund's liabilities, not just against equities in a bull market, or the first serious drawdown will reopen the whole debate. The commercial implication is not "adopt TPA." It is that the policy portfolio is becoming a contestable choice rather than a default — and that the funds best positioned to exploit cross-asset opportunities (AI infrastructure, energy, private credit sized by contribution rather than target) will be the ones that already have the governance to run without the grid.

What to watch next

June 15–17, 2026: the CalPERS board's vote on staff incentive changes — the real signal of whether the model will hold.
July 1, 2026: TPA goes live; the first reference-portfolio performance numbers will land in the FY2026–27 reporting cycle.
This autumn: Washington State's decision on the size of its standalone private-credit allocation — a test of how the channel sizes a contested asset class.
Whether CalSTRS or another large US plan follows — the difference between a CalPERS experiment and a channel-wide shift.

Sources

*UAO Research. AI-assisted monitoring and drafting; reviewed and edited by the UAO editoria

Continue the briefing. Read the daily brief · watch the daily video briefing · listen to The Allocator Briefing · view the chart of the day.

Produced and edited by the UAO editorial desk. Not investment advice.

What the evidence says

Where it is contested

From the allocator's seat

What to watch next

The morning briefing for the people who allocate long-horizon capital.