Does the total portfolio approach make a pension a better allocator — or just a faster one into private markets?

UAO Research · May 27, 2026 · Pensions & Retirement · Feeds: Pension Strategy Watch

On July 1 CalPERS retires the strategic asset allocation model it has run for decades and adopts a total portfolio approach, lifting its private-equity target to 17% and standing up a new 8% private-credit bucket along the way (CalPERS PERSpective, Nov 17, 2025). The move lands three weeks after the Financial Stability Board published its first dedicated warning on private credit. So the question every large owner is now watching CalPERS to answer is uncomfortably precise: when you let all opportunities compete for capital under a single objective, do you allocate better — or do you simply find it easier to lean into illiquidity at exactly the moment a regulator is telling you to be careful?

What the evidence says

Start with what TPA actually is. The Thinking Ahead Institute — the WTW-affiliated think tank that has done the most to codify the model, in research co-authored with Australia's Future Fund, CPP Investments, New Zealand Super and GIC — defines it by three features: a clearly specified total-fund objective, a single contested pool of capital in which every investment competes against every other, and dynamic, real-time governance rather than periodic rebalancing toward fixed weights (Total Portfolio Approach hub, Thinking Ahead Institute, 2025). The Institute's headline claim is that, implemented well, TPA can add 50 to 100 basis points of annual return relative to a conventional strategic-asset-allocation policy portfolio.

The mechanism behind that number is the interesting part. Under strategic asset allocation, a fund sets fixed bands — say 13% private equity — and the band itself becomes the constraint. Capital flows to fill the bucket regardless of whether the marginal private-equity deal is better than the marginal infrastructure or credit deal available that quarter. TPA removes the bucket. Every dollar is allocated to wherever it most improves the total fund relative to the reference portfolio, which CalPERS has set at 75% equities, 25% bonds. GIC describes its own version as organising the portfolio around risk and return drivers rather than asset-class labels, precisely so that capital is not trapped by a taxonomy (GIC's Total Portfolio Approach, GIC ThinkSpace). In principle, that is a discipline upgrade: it forces an explicit, like-for-like comparison the old model let funds avoid.

The trouble is that the same machinery cuts the other way, and the FSB has just described the terrain it cuts into. Its May report sizes private credit at $1.5–2.0tn, notes that roughly three-quarters of borrowers carry EBITDA below $100m, and warns that reported leverage of five to six times may understate true leverage closer to seven times once EBITDA adjustments are stripped out. It flags the rising share of semi-liquid vehicles that offer redemptions — a structure that can turn procyclical under stress — and "complex interlinkages" with banks, insurers and private-equity sponsors that supervisory data cannot yet fully map (Report on Vulnerabilities in Private Credit, FSB, May 6, 2026). A model designed to let capital flow frictionlessly to its highest total-fund use is, by construction, a model that can flow capital into a leveraged, opaque, increasingly redeemable asset class faster than a bucketed model ever could.

Where it is contested

Two things are genuinely unresolved. The first is whether TPA's claimed return premium survives contact with a downturn. The 50–100bps figure is drawn largely from funds — Future Fund, GIC, CPP — that adopted the approach in benign conditions and have strong internal governance and in-house teams (Total Portfolio Approach: A New Way to Construct Portfolios, Chief Investment Officer, 2026). Whether a US public plan with political oversight, a board of varying expertise and external-manager dependence can capture the same premium is untested. The Future Fund's own CIO has called the approach "taxing" to run well — it demands far more judgement, and far better data, than filling fixed bands does.

The second is a measurement problem the FSB itself names: private-credit valuations are "stale and potentially subjective." TPA depends on comparing the risk-adjusted contribution of each opportunity in something close to real time. If private-credit marks lag reality — as they do almost by definition in an asset class that does not trade — then the very comparisons TPA runs on are fed inputs that flatter illiquid assets during calm periods and lurch only after stress arrives. A model that allocates dynamically on smoothed inputs can systematically over-weight the thing whose risk is hidden. That is not an argument against TPA; it is an argument that TPA without independent valuation discipline can amplify exactly the vulnerability the regulator is worried about.

From the allocator's seat

For a CIO weighing whether to follow CalPERS, the lesson is that TPA relocates the control, it does not remove the need for one. Under SAA, the allocation cap did your discipline for you, crudely. Under TPA, the discipline has to live in governance: in the reference portfolio you choose, in the risk budget you set against it, in the independence of your valuation function, and in the explicit hurdles a private-credit deal must clear to win capital from a liquid alternative.

Concretely, that means three questions to ask before adopting the model. Does our valuation and risk function report independently of the investment teams whose deals it marks — or will smoothed private marks quietly win every internal contest for capital? Have we set an explicit illiquidity budget at the total-fund level, so "no fixed buckets" does not become "no limit"? And is our board equipped to govern a model where the headline number is a single total-fund return and the asset-class detail is buried a layer down — the same opacity the CalMatters critique seized on (The CalPERS gamble, CalMatters, May 22, 2026)? TPA is a better tool for an owner that already has those answers. For one that does not, it is a faster route to the FSB's worry.

What to watch next

July 1, 2026: CalPERS' TPA goes live. Watch the first reporting cycle for how — and how granularly — it discloses private-market exposure under the new model.
CalPERS' FY2025–26 results (typically reported July): the first read on the portfolio as it transitions, and on the pace of private-credit build-out toward 8%.
FSB action-plan milestones: the body committed to closing private-credit data gaps; the next deliverables will shape what supervisors — and LPs — can actually see.
Peer board agendas (Q3 2026): watch whether mid-sized US public plans put TPA on their own agendas. CalPERS adoption tends to set the template for the segment.

Sources

UAO Research. AI-assisted monitoring and drafting; reviewed and edited by the UAO editorial desk before publication. Not investment advice.

Continue the briefing. Read the daily brief · watch the daily video briefing · listen to The Universal Owner · 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.