Australia’s superannuation system reached A$4,437.9 billion at 31 March 2026. The figure is accurate. This brief decomposes the three hidden architectures inside it — an ownership split, a stock-vs-flow paradox, and a performance label that measures relative gaps, not absolute member outcomes.
When annual net contribution flows are only ~1.7% of the stock, a single quarter's market revaluation (~A$44bn on the -1.0% move, RAOSCAFF arithmetic) is comparable to most of a full year's net inflow. The A$4.44tn super number is more sensitive to a revaluation of its asset stock than to a year of net cash flow. RAOSCAFF arithmetic: 74.5 ÷ 4,437.9.
Australia’s total superannuation assets reached A$4,437.9 billion at 31 March 2026 (APRA Quarterly Superannuation Statistics). Up +7.9% year-on-year from A$4,111.4bn. Down -1.0% over the March quarter. The figure is correct. It is also a container — for four ownership regimes with different regulators, for a system where quarterly market returns dominate annual cash flows, and for a regulatory performance label that most coverage shortens to “7 products failed.”
Each of those three stories changes the analytical frame. None of them appears in the headline.
Sources: APRA Quarterly Superannuation Statistics March 2026 (§H URL 1); APRA Quarterly Superannuation Performance Statistics Highlights March 2026 (§H URL 2); APRA 2025 Performance Test Results (§H URL 3).
Annual net cash contributions of A$74.5bn represent ~1.7% of the A$4,437.9bn stock. A single quarter of market softness can rival, in dollar terms, most of the entire year’s positive net inflow.
Axis 1 — Ownership regime. The A$4,437.9bn headline aggregates four structurally distinct ownership regimes. The two largest — APRA-regulated pooled funds (A$3,141.1bn, +8.7% y/y) and self-managed super funds (A$1,057.6bn, +7.0% y/y) — are governed by different regulators, face different annual tests, and grew at a 170bp differential in the year to March 2026 (RAOSCAFF arithmetic: 8.7% − 7.0%). The four categories sum exactly to the headline (RAOSCAFF arithmetic: 3,141.1 + 1,057.6 + 178.4 + 60.8 = 4,437.9bn). None of the internal growth rates matches the 7.9% aggregate.
Axis 2 — Stock vs flow. Annual net cash contributions added A$74.5bn — approximately 1.7% of the stock (RAOSCAFF arithmetic: 74.5 / 4,437.9). The March quarter saw the stock fall -1.0% despite those positive flows. The annual rate of return for APRA-regulated funds was 7.4%; the quarterly compression was a market-return effect, not a cash-flow event. Member contributions surged: employer A$159.8bn (+8.4%), member A$66.3bn (+19.1%). Those signals are real — but not large enough relative to the stock to buffer against mark-to-market quarterly moves.
Axis 3 — Performance test design. The 2025 YFYS annual test assessed 563 products. Seven failed — all in the platform trustee-directed category (137 tested); zero in MySuper (52 tested); zero in non-platform trustee-directed (374 tested). The test measures investment performance against a strategic-asset-allocation benchmark over up to ten years, with an administration fee and costs component. A product “fails” if it underperforms the benchmark by the legislated threshold — not if members lost money in absolute terms. Products with fewer than 7 years of history automatically meet the test. The CPPP metric (a separate APRA analytical instrument, also 10-year) found approximately 40% of platform trustee-directed products with a 10-year history showing significant investment underperformance — but the CPPP is not the pass/fail test.
| Dimension | Published figure | What it conceals |
|---|---|---|
| Headline stock | A$4,437.9bn · +7.9% y/y · -1.0% quarter | Aggregates four ownership regimes; growth rates differ across all four; none matches the 7.9% aggregate |
| APRA-regulated pool | A$3,141.1bn · +8.7% y/y | ~71% of headline; pooled, trustee-managed; its MySuper and trustee-directed products subject to the annual performance test; grew 170bp faster than SMSFs (RAOSCAFF arithmetic) |
| SMSF pool | A$1,057.6bn · +7.0% y/y | ~24% of headline; self-directed, ATO-regulated; not subject to YFYS performance test |
| Annual net cash flow | +A$74.5bn · ~1.7% of stock | The stock is far larger than its annual net flow, so it tracks asset revaluation more than cash flow; net flow is ~1.7% of the stock (RAOSCAFF arithmetic) |
| Performance test (2025 cycle) | 7 of 563 products failed | “Failed” = underperformed SAA benchmark + fees over up to 10 years — not absolute member loss; all 52 MySuper products passed; products with <7 years of history meet the requirements absent an APRA determination |
If you read one thing: the A$4.44tn figure is more sensitive to a single quarter's revaluation of its asset stock than to a full year of net cash contributions.
The A$4,437.9bn superannuation headline is the correct published stock at 31 March 2026. Its analytical content lies in three hidden architectures. First, the ownership-regime split: the aggregate conceals a 170bp growth-rate differential between the APRA-regulated pooled pool (A$3,141bn, +8.7%) and the SMSF pool (A$1,058bn, +7.0%) — two structurally different systems, different regulators, different tests, both invisible in the 7.9% headline. Second, the stock-vs-flow structure: annual net contributions of A$74.5bn represent ~1.7% of the stock (RAOSCAFF arithmetic), which means the system at this scale is far larger than its annual net flow, so it tracks a revaluation of the asset stock more than cash in and cash out — a single quarter’s revaluation (~A$44bn here, RAOSCAFF arithmetic) is comparable to most of the year’s net inflow. Third, the performance label: “7 products failed” means 7 platform trustee-directed products underperformed a 10-year strategic-asset-allocation benchmark by the legislated threshold including fees — not that members in those products lost money in absolute terms; all 52 MySuper products passed; products with less than 7 years of history meet the requirements absent an APRA determination. Mirror Brief AU-13 makes no prediction on the direction of superannuation assets, no recommendation on product or fund selection, and no comment on compulsory superannuation policy. Every figure traces to a primary APRA source; derived figures are labelled RAOSCAFF arithmetic.
Mirror format — RAOSCAFF anchors on the publisher’s own filed documents and decomposes the figure they print. AU-13 uses five APRA primary sources: the March 2026 Quarterly Superannuation Statistics news release (headline + category splits), the March 2026 Quarterly Superannuation Performance Statistics Highlights (contributions, benefit payments, net flow, annual rate of return), the 2025 Annual Superannuation Performance Test Results page (563 products, 7 failed, category breakdown, CPPP insight), the YFYS FAQ (10-year horizon, verbatim; less-than-7-year auto-pass rule; fee components), and the Annual Superannuation Performance Test overview. No analyst estimates; no extrapolation; no survey data.
Industry vs retail sub-split omitted: not re-fetched from primary APRA data. No individual fund or trustee named. No directional view on returns or flows. The -0.50pp legislated underperformance threshold cited in third-party commentary was not confirmed in plain-language APRA web pages accessed and is not stated here; it is treated as medium confidence until verified against SIS Regulations. The YFYS test horizon is “ten years” per APRA FAQ verbatim — not eight years as sometimes cited. RAOSCAFF arithmetic register: category split sum (3,141.1 + 1,057.6 + 178.4 + 60.8 = 4,437.9bn); net flow as % of stock (74.5 / 4,437.9 = ~1.68%); APRA-regulated vs SMSF growth differential (8.7% − 7.0% = 170bp). Every computed figure so labelled.
All five sources are APRA primary publications. Full citations and the verified figures behind every number in this brief are listed in the companion FACTS.md, §H. Headline snapshot: 31 March 2026. Next APRA quarterly (June 2026) will supersede. YFYS results: 2025 annual test cycle; next cycle = 2026 annual test.