Jeremy Grantham's asset allocation: GMO's valuation bet

Jeremy Grantham's asset allocation is a valuation bet: GMO's Benchmark-Free Allocation Fund owns what the firm's 7-year forecasts rate cheap, and at 30 September 2026 that meant 43.7% equity exposure, mostly outside the US, 29.7% in alternative strategies and 26.6% in fixed income.

Judged against MSCI World in US dollars to September 2026, the fund was partly set up for different market scenarios and cushioned equity crashes without protecting against them. Its volatility-adjusted returns were partly better, ahead over 20 years and the full window but not over the last 10, and its returns after fees were not better in any window we measured.

Key takeaways

1. Set up for different market scenarios: partly. It lost far less than MSCI World in 2008, 2020 and 2022, but gained 63.4% in the 2023 to 2026 rally against 98.8% and only 2.3% a year from 2014 to 2021.
2. Guard against catastrophic equity falls: cushions, does not protect. Its worst fall was 18.0% against 54.0% for MSCI World, from bonds, alternatives and less equity rather than any contractual payoff.
3. Better volatility-adjusted returns than MSCI World: partly. The Sharpe ratio was 0.60 against 0.50 over 20 years and 0.79 against 0.58 over the full window, but 0.52 against 0.73 over 10 years, and the fund won 73 of 159 rolling 10-year windows.
4. Better returns net of fees than MSCI World: no. It compounded at 6.6% a year against 12.8% over 10 years, 6.0% against 8.6% over 20 years and 7.8% against 9.6% over the full window.
5. Valuation, not an index, sets the weights. The US was 12.8% of the fund's equity holdings at 30 September 2026, against 72.94% of MSCI World, and GMO's forecast put US large caps at -7.3% a year after inflation.

How does GMO Benchmark-Free score on our four tests?

We judge the fund's Class III shares (GBMFX), net of their expenses, against MSCI World Net in US dollars, with one-month Treasury bills as cash, using monthly returns to September 2026. MSCI World is a developed-market equity index, not a balanced benchmark.

Windows: 10 years, Oct 2016 to Sep 2026; 20 years, Oct 2006 to Sep 2026; full window, Aug 2003 to Sep 2026. The fund started in July 2003, so there is no dotcom episode.

What is Jeremy Grantham's role at GMO today?

Grantham co-founded GMO in 1977 and is its long-term investment strategist and chairman of the board, according to GMO's author biography. He no longer runs portfolios: in Waiting for the Last Dance (5 January 2021) he wrote that he was "long retired from the job of portfolio management". The fund's summary prospectus (30 June 2026, revised 20 October 2026) names Ben Inker, on the fund since 2003, and John Thorndike, since 2019, as the people primarily responsible for it.

What does GMO's Benchmark-Free Allocation Fund hold?

At 30 September 2026 the fund's exposures were 43.7% equity, 29.7% alternative strategies and 26.6% fixed income, with $2.7bn in total assets. It aims to return 5% a year above US inflation, net of fees, with 5% to 10% volatility over a complete market cycle, and its fact sheet says it is "constrained only by our unwillingness to overpay for an asset".

At 30 September 2026, % of fund: Equity Dislocation 14.6%; US Nominals 13.3%; Alternative Allocation 13.1%; US TIPS 11.4%; Emerging Markets 8.8%; International Opportunistic Value 8.6%; Japan Fundamental Value 6.1%; Developed ex-US 4.9%; US Opportunistic Value 4.8%; Developed ex-US Small Cap Value 4.8%; Resource Equity 2.8%; Japan Small Value 1.9%; Emerging FX 1.9%; ABS and structured products 1.9%; Domestic Resilience 1.0%.

Source: GMO Benchmark-Free Allocation fact sheet, 30 September 2026, p. 2; chart data (CSV). GMO measures these exposures with its own method; Equity Dislocation is long cheap value stocks and short expensive growth stocks, so its 14.6% is not a net bet on the market.

Japan made up 30.6% of the fund's equity holdings, emerging markets 21.9% and the United States 12.8%. In MSCI World the US weight was 72.94% and Japan 5.91% on the same date, 30 September 2026, per MSCI's fact sheet; its 30 October 2026 edition, now at the same link, shows 72.77% and 5.72%. The mix moves a long way: Inker's 25 Years of Benchmark-Free Investing (September 2026) shows a representative account at 60% equities in July 2003, 25% in August 2008 and 52% in April 2009.

Class III (ticker GBMFX), the share class with the longest record, carries a 1.35% net expense ratio and requires $1m invested with GMO, per GMO's fund page.

What do GMO's 7-year forecasts say in 2026?

GMO's 7-year asset class forecast as of 30 September 2026 ranks cheaper markets far above US large caps.

The normal scenario assumes a 1.3% equilibrium real cash rate, the low scenario 0.0%. GMO says the forecasts cover asset classes, not any fund, and are not a guarantee.

What has Grantham said about bubbles, and what happened next?

Grantham defines a bubble as a price two standard deviations above its long-term real trend, and argues that in large developed equity markets such bubbles have always fallen back to trend. In January 2021 he wrote that the bull market since 2009 had "matured into a fully-fledged epic bubble".

In Let the Wild Rumpus Begin (20 January 2022) he called the US market "the fourth superbubble of the last hundred years", put the S&P 500's trend value at about 2,500 against nearly 4,700, and summarised GMO's advice as to "avoid U.S. equities and emphasize the value stocks of emerging markets and several cheaper developed countries, most notably Japan." In Entering the Superbubble's Final Act (31 August 2022) he wrote that "every historical parallel suggests that the worst is yet to come."

The first leg came; the return to trend did not. On our month-end data MSCI World fell 25.4% from December 2021 to September 2022 and regained its peak by December 2023. In Valuing AI: Extreme Bubble, New Golden Era, or Both (28 January 2026), Grantham wrote that "unlike every bubble before it, we have yet to see this from the December 2021 peak", saying ChatGPT "nipped in the bud" the 2022 bear market; he still calls US stocks a bubble.

Have GMO's valuation forecasts shown skill?

Inker's paper says a 60/40 portfolio was forecast in 1999 to return about 2% a year after inflation over the next decade and delivered 1.3%, while GMO's suggested mixes, forecast at 5.0% to 5.75%, delivered 6.0% to 6.9% to September 2009; it also reports emerging markets proving "a value trap" from 2015 to 2021.

In 2010 CXO Advisory found a 0.94 correlation between GMO's 1999 forecasts and outcomes across 11 asset classes, but called the example "sampled with hindsight for promotional purposes". We found no complete, independently audited record of every forecast vintage.

What does Grantham do with his own money?

We found no published allocation for Grantham's personal portfolio. In January 2022 he gave a preference, not weights: "Speaking personally, I also like some cash for flexibility, some resources for inflation protection, as well as a little gold and silver."

Much of his wealth is in philanthropy: the Grantham Foundation, set up with his wife Hannelore in 1997, lists grants and investments in clean energy, carbon removal, agriculture and chemical toxicity. Inside Philanthropy reported in July 2022 that the couple pledged in January 2019 to commit 98% of a fortune of roughly $1bn to climate action.

Is GMO Benchmark-Free set up for different market scenarios?

Partly: the fund lost much less than MSCI World in each crash it has lived through, but long growth-led bull markets left it far behind.

GBMFX against MSCI World, in USD: financial crisis (November 2007 to February 2009) -18.0% and -54.0%; Covid crash (February to March 2020) -13.9% and -20.6%; inflation and rate shock (2022) -2.3% and -18.1%; rally (January 2023 to September 2026) 63.4% and 98.8%. Source: our calculation from monthly total returns (GBMFX closing prices with distributions reinvested, from Yahoo Finance; MSCI World Net from MSCI); monthly data (CSV); chart data (CSV).

In the 37 months when MSCI World rose 5% or more, the fund averaged 3.1% against 7.2%; in the 23 months when it fell 5% or more, the fund averaged -3.1% against -8.6%. The regime that hurt was the US-led bull market: from January 2014 to December 2021 GBMFX gained 20.1% in total (2.3% a year) while MSCI World gained 125.3% (10.7% a year) and cash 0.7% a year. GMO's 25-year paper calls under-allocating to US stocks after about 2010 its "most damaging misstep".

Does GMO Benchmark-Free guard against catastrophic equity falls at all?

It cushions rather than protects. Nothing in the fund pays out contractually when stocks crash: the shorts in Equity Dislocation are against expensive growth stocks paired with long value stocks, not a hedge on the market, and the prospectus says the fund may write put options on stock indices, which lose value in a fall. What softens falls is less equity, US nominal bonds and inflation-protected Treasuries (TIPS) in a fixed-income book with an effective duration of 2.7, and alternatives that seek returns above cash.

The fund's worst drawdown, 18.0% from October 2007 to February 2009, was recovered by September 2009; MSCI World's 54.0% fall over the same months took until July 2013 to recover (why recovery takes longer than the loss suggests).

In October 2008, the index's worst month, the fund lost 4.3% against 19.0%, and it rose in only one of the 23 months when MSCI World fell 5% or more. Its monthly returns had a correlation of 0.77 with MSCI World and a beta of 0.40, so it moves with stocks, by less.

Diversification depends on different drivers, not labels: Inker says GMO groups exposures by what causes lasting losses: bad recessions, unanticipated inflation and liquidity shocks. But value runs through the equity book and Equity Dislocation, the largest single exposure, so a long run of expensive growth stocks winning hurts several lines at once, as 2014 to 2021 showed.

Are GMO Benchmark-Free's volatility-adjusted returns better than MSCI World?

Partly: better over 20 years and the full window, worse over 10 years, and better in fewer than half of rolling 10-year windows. The Sharpe ratio divides the average monthly return above cash by the volatility of that excess return, annualised. The Sortino ratio divides the same excess return by downside deviation, so only months that fell short of cash enter the denominator.

Of 159 rolling 10-year windows, the fund had the higher Sharpe ratio in 73 and the higher Sortino ratio in 74. The edge sits early: from August 2003 to December 2013 the Sharpe ratios were 1.29 against 0.49, and from January 2014 to September 2026, 0.44 against 0.66.

At equal volatility the fund came out ahead over the full window: MSCI World held at 51.9% with T-bills, rebalanced monthly without costs, had the fund's 7.8% volatility from August 2003, and that mix compounded at only 6.1% a year with a 31.6% worst drawdown, against the fund's 7.8% and 18.0%.

Simple theory says a higher Sharpe ratio raises the growth reachable by scaling a strategy with borrowing, but for GBMFX that would have meant about ten times leverage, at which an 18.0% fall is ruinous (Sharpe ratio versus CAGR). These ratios also depend on the choice of cash rate and say nothing about allocation funds that closed.

Are GMO Benchmark-Free's returns net of fees better than MSCI World?

No, in every window we measured. Fund returns are net of fees; MSCI World Net is an index and carries no costs, while a fund that tracks it does (the iShares Core MSCI World UCITS ETF charges 0.20% a year), so a shortfall smaller than such a fee counts as close rather than a clear loss.

The fund had the higher compound return in 53 of 159 rolling 10-year windows. The gaps, 1.9 to 6.2 percentage points a year, are far larger than that tracker fee. The prospectus says returns before 2012 would have been lower under current fees, and 2024 includes a one-time litigation recovery worth 2.45%. Against its own target the prospectus reports a net return 4.87% a year above US inflation from July 2003 to December 2025, just short of inflation plus 5%.

Grantham's January 2021 letter explains why the lag matters: long bull markets "can easily outlast the patience of most clients", and "there is nothing more supremely irritating than watching your neighbors get rich." Inker's 2026 paper says GMO's clients "have tended to abandon the strategy after a bout of relative underperformance"; anyone who sold after 2021 missed the fund's 22.87% gain in 2025.

Weighing the four results

GMO's fund gave up return to avoid deep falls: it trailed MSCI World after fees over 10 years, 20 years and since 2003, and lost about a third as much at the worst point of 2008. Its volatility-adjusted record beat the index over the longer windows but not over the last decade, when the US-led bull market it judged expensive kept rising.

Whether that trade suits anyone depends on how much a deep fall would cost them and how long they could trail an index; the numbers above do not settle it.

What can an individual investor learn from Jeremy Grantham?

Grantham's lasting point is that the price paid shapes the return available, and acting on it means looking different for long stretches; copying GMO's weights would copy a view that may take years to pay off, or may not.

Questions to examine in your own portfolio:

• What share of your equities is in the US, and is that a decision or simply the index weight?
• What might each asset return from today's prices, and how wrong could that estimate be?
• How long could you trail a world index before abandoning a strategy?
• What did the rest of your portfolio do when stocks fell sharply? See how asset classes behaved in past crashes.
• Do several holdings rely on the same idea, such as value beating growth, under different labels?

Compare Howard Marks on market cycles, Seth Klarman's cash discipline, Cliff Asness on diversification and our conservative asset allocation guide.

How can copio help you examine your own allocation?

Copio is in private beta across Europe, with early access by application; it offers educational analysis, not personalised advice, and does not place trades.

Its fund look-through breaks funds into asset classes, a return-correlation map sorts asset classes into main driver, true diversifiers, dead weight and shock absorbers, a crash test shows how a mix like yours would have fared in past crises, and an efficient frontier plots your portfolio as "YOU" against a modelled frontier of assumed 10-year growth and volatility, labelled not a forecast.

Like GMO's forecasts, that frontier rests on assumptions, so it shows where your mix sits and what a change trades away rather than an optimal allocation. The free portfolio checker examines concentration in your own holdings.