Research Note ·
Passive Indexing and the Percentile Comparison
What a personal passive-index portfolio actually achieved versus the median pension, the median endowment, and Yale — with FY2024 and FY2025 data from NACUBO, CPP Investments, Ontario Teachers', the Yale Investments Office, Cliffwater, and the NACUBO-Commonfund Study of Endowments.
Thesis
A retail passive-index portfolio appears to out-rank almost every institutional pension and endowment over the most recent 10-year window — and the gap is largely structural, not skill-based.
Three personal investing documents prepared by the author between 2007 and 2019 make a single, repeated, empirically-grounded claim: a low-cost, broadly-diversified, equity-tilted passive index portfolio — held with discipline through drawdowns — outperforms the great majority of professionally-managed pension funds and university endowments over rolling 10-year periods. Six to twelve years later, the institutional benchmarks have changed but the basic claim still holds. This note updates the comparison to the FY24 and FY25 reporting cycles, re-runs the percentile placement, and explains why institutions under-perform — most of which is structural rather than skill.
The institutional under-performance is not an indictment of the professionals running these funds. Two mechanical factors explain most of it: liability-matching fixed-income allocation, and private-asset valuation lag during the 2022-2024 rate-hiking cycle. A retail investor not facing those constraints captures the equity premium without paying the diversification toll. — from Section 1, Principal findings
Five principal findings
What the data show — distilled
① The decade ending mid-2025 was anomalously generous to US equity
~13%/yrS&P 500 10-year total return through October 2025 ran approximately 13-15% per annum, driven by the post-GFC recovery, zero-rate compounding, and the AI-led mega-cap surge. Every major institutional universe in this study returned 6-10% over the comparable window.
② NACUBO-Commonfund FY25 universe median is 7.7%
7.7%657 US college and university endowments, $944.3B in aggregate AUM, 10-year average annual return through June 30, 2025. The largest endowments ($5B+) reach approximately 9-10%; the smallest cohorts are closer to 6-7%.
③ CPP Investments delivers 8.8% — and is the highest-returning pension on this list
8.8%Canada Pension Plan Investment Board, $793.3 billion AUM, 10-year annualised net return through March 31, 2026. Materially above Ontario Teachers' (7.4%), NY State Common (7.74%), and the Cliffwater 24-year state-pension median (6.34%).
④ Yale, the Yale-model exemplar, returned 9.4%
9.4%Yale Investments Office FY2025 10-year annualised. Above the NACUBO median by 1.4 percentage points but below the US large-cap passive benchmark by approximately 3.5pp. The Yale CIO has explicitly acknowledged private-asset valuation lag as the source of recent under-performance.
⑤ The personal-portfolio claim places at the 95th-99th percentile
>99thUsing NACUBO dispersion approximations, the author's reported 14.2% (10-year to April 2019) sits at z = +4.93 above the FY24 mean, well beyond the 99th percentile of the institutional universe. The expected mean-reversion case (13%) is at approximately the 97th-99th percentile.
Direct ranking
10-year annualised net return — ranked, ending mid-2025
Rather than rely on dispersion assumptions, the cleanest test is to rank a hypothetical 10-year return for the author's portfolio (which, given US large-cap exposure plus Vanguard small-value and emerging-market tilt, would have tracked the S&P 500 within approximately ±100 basis points over 2015-2025) directly against named institutions.
| # | Entity | 10-yr return | Notes |
|---|---|---|---|
| 1 | Author's portfolio (US-heavy passive) | 13-14% | S&P 500 plus small/value/emerging tilt |
| 2 | Yale University endowment (FY25) | 9.4% | Best Ivy-style long-term operator; private-asset drag |
| 3 | Ivy League average (FY24) | ~9.2% | All eight schools |
| 4 | Endowments >$5B (FY23 cohort) | 9.1% | Larger size = more resources |
| 5 | CPP Investments (FY26) | 8.8% | $793B AUM; Maple 8 flagship |
| 6 | CPP Investments (FY25) | 8.3% | Same fund, earlier reporting cut |
| 7 | NY State Common Retirement (FY25) | 7.74% | Largest US state plan |
| 8 | NACUBO-Commonfund 658 endowment universe (FY25) | 7.7% | Median for entire universe |
| 9 | Ontario Teachers' Pension (CY2024) | 7.4% | Mature liability profile |
| 10 | NACUBO-Commonfund universe (FY24) | 6.8% | Prior year reporting |
| 11 | Cliffwater state pension universe (24-yr) | 6.34% | Long-window state pensions |
| 12 | Global 70/30 passive (Cliffwater 24-yr) | 5.71% | Reference passive portfolio |
Sources: NACUBO press release Feb 12, 2026; CPP Investments press releases (May 2025, May 2026, August 2025); OTPP press release March 20, 2025; Yale News October 24, 2025; NY State Office of the Comptroller press release June 2025; Swedroe/Cliffwater Substack June 2025; Ivy Coach 2024 review. Full citations in the PDF.
Mechanism
Why institutions under-perform — five drivers
| # | Mechanism | Annual cost | Skill failure? |
|---|---|---|---|
| 1 | Liability-matching fixed-income allocation (20-30% of pensions; 10-20% of endowments) | 1.5-2.5% | No — structural |
| 2 | Tax-exempt status enabling debt-substitution in operating-company buyouts | +2-3% | No — advantage to institutions |
| 3 | Private-equity / private-real-estate valuation lag (FY22-FY24 specifically) | 1.0-2.0% | No — valuation method |
| 4 | Active-manager under-performance (88% lose vs benchmark over 5yr; SPIVA basis) | 0.5-1.5% | Yes — partly |
| 5 | Manager and consulting fees (1-3% AUM at hedge/PE layers) | 0.5-2.0% | Yes — partly |
Three of the five mechanisms are not failures — they are structural features of running a pool of capital with specific cash-flow obligations and valuation conventions. The retail investor enjoys none of these constraints and captures the equity premium that institutional managers cannot fully harvest.
Antithesis
Five reasons to discount the headline finding
Scientific honesty requires explicit consideration of the counter-arguments. The author's 99th-percentile claim is broadly supported by the data, but five specific corrections reduce it materially.
- Window-selection. Every "passive beats institutions" comparison is computed over a 10-year window, and 10-year windows are too short to draw structural conclusions. The 1990-2000 US passive return was approximately 18%/yr; the 2000-2010 return was approximately -1%/yr. The current window flatter US large-cap to an unusual degree. Yale's 20-year 10.3% return remains the single most credible long-window endowment performance ever recorded.
- Self-reported survivorship. The 11% (10-yr to Feb 2013) and 14.2% (10-yr to April 2019) personal-portfolio returns are not audited. Personal returns frequently appear inflated because failed positions are forgotten, tax drag is excluded, and cash-drag is mis-handled in spreadsheet time-weighting. A Modified Dietz audit would likely show a number 1-3pp lower.
- Sequence-of-returns risk in retirement. The pension fund 7-9% return includes the cost of de-risking that enables reliable monthly benefit payment regardless of market conditions. A personal portfolio at 100% equity in retirement, combined with required withdrawals, can crystallise permanent capital impairment that the institutional pension structurally cannot.
- Tax-arbitrage favouring institutions. Tax-exempt pensions can buy operating companies, substitute debt for equity, and extract profits as tax-deductible interest. This is a 2-3% structural advantage unavailable to a taxable personal investor. On a tax-equivalent basis, the gap closes substantially.
- Behavioural execution. Dalbar's annual study consistently shows that the average mutual-fund investor's realised return is 1.5-3.0% per year below the fund's own time-weighted return, due to buy-high/sell-low timing. The institutional advantage is enforced discipline. A personal investor who claims to be running passive but actually adjusts during drawdowns will realise institutional-style returns or worse.
Conclusion. A more defensible long-run expectation for a tax-exposed Canadian passive-tilt investor is in the 7-9% real range after tax and inflation — still better than the median pension but not the dramatic 99th-percentile gap the original documents claim. The structural advantages of being a retail investor without liability-matching obligations are real and large; the personal-skill premium on top is much smaller and often negative.
Practical implications
Operational conclusions for the personal investor
- Cap active management at 25-50 basis points. The SPIVA base rate (88% under-perform over 5 years, 98% over 10 years) is robust across two decades. Retail-fund expense ratios above 25-50bp are overwhelmingly negative-expected-value.
- Hold for at least 20 years. Rolling 10-year windows have ranged from -1.5% to +20%. Rolling 20-year windows have all been positive (3.1% to 17.7%).
- Tilt toward factor premia — but expect long stretches of under-performance. Value and small premia were largely absent 2015-2024.
- Use an Investment Policy Statement formally. The IPS in the source documents has all the elements of a foundation IPS — loss limits, rebalance thresholds, sell triggers, monitoring procedures. Writing it down and following it is the single largest defence against the behaviour gap.
- Avoid alternatives unless tax-exempt access is available. Hedge funds (17.7% beat replication per Kat & Palaro), private equity (IRR overstatement, illiquidity premium compression), and private real estate do not justify their fee structures for a taxable retail investor.
- Locate high-yield and high-turnover positions in registered accounts. The IPS tax footnote: foreign dividends in a personal account face 46.42% Canadian marginal; the same security held in an RRSP defers to 0% inside but 46.42% on withdrawal; a buy-and-hold position attracts only 23.21% capital gain at realisation. Asset-location is worth 1-2 percentage points of after-tax return per year for typical retiree portfolios.
See also
The companion Monte Carlo retirement optimizer implements the conclusions of this research note. It uses these institutional return ranges as defaults for its return and volatility assumptions, and models the after-tax/after-inflation Safe Withdrawal Rate for a passive-index portfolio under both US and Canadian tax regimes.