Faber GAA

Faber Global Asset Allocation (GAA) composite portfolio

Mebane T. Faber (Cambria Investment Management) · 2015 (book 'Global Asset Allocation: A Survey of the World's Top Asset Allocation Strategies')

Live account on this site. This portfolio is also tracked here with a virtual $100,000 since 21.10.2026. Live account on this site →

Where it stands today

Data through 2026-09. Updated every month.

Since January+4.2 %

Last 12 months+6.4 %

Last 5 years, per year+4.3 %

Now below the peak−2.6 %

Place by 12 months50 / 60

Our calculation, 1984-02 – 2026-09

Per year8.7 %

Real (after inflation)5.7 %

Volatility8.1 %

Worst drop−23.0 %

Worst year−16.7 % (2022)

Best year28.4 %

Worst 12 months−21.5 %

Positive years85 %

9 % of this history is calculated with stand-ins for asset classes that did not exist yet (see “How we model it” and the page about data). Without stand-ins the history starts in 2000.

Growth of $10,000

Faber GAA: $350,677S&P 500: $1,163,109

Where it came from

Meb Faber is the founder and chief investment officer of Cambria Investment Management and a prolific author of free research papers and books. In early 2015 he self-published an e-book, 'Global Asset Allocation: A Survey of the World's Top Asset Allocation Strategies' (priced at a couple of dollars at launch). It compares a dozen or more well known model portfolios (for example the Permanent Portfolio, the endowment style portfolios, Bernstein's and Swensen's) on the same 13 asset classes with data from 1973.

The 13 classes in the book: US large and small stocks, foreign developed and emerging stocks, US corporate bonds, T-bills, 10-year and 30-year US Treasuries, 10-year foreign government bonds, TIPS, commodities (GSCI), gold and REITs. A review by Javier Estrada and others summarises two findings: apparently different strategies have similar exposures, and the cost of investing is a key determinant of performance. Excluding the Permanent Portfolio, the average returns of the strategies were within about one percentage point of each other (a blog review of June 2015).

The book's 'Global Asset Allocation' portfolio is a composite benchmark made from the allocations of prominent investors and endowments; AAII says it resembles the global market portfolio, that is, all global investments combined. It holds 36% stocks, 49.5% bonds and 14.5% real assets (REITs, commodities and gold). Faber had earlier shown a rougher global market portfolio in a blog post of 27 June 2014 (40% stocks, 20% corporate bonds, 30% government bonds, 5% real estate, 2% each TIPS, EM bonds and high yield, rounded to 101%).

Separately, Cambria launched an ETF named Cambria Global Asset Allocation ETF (ticker GAA) on 9 December 2014. It is not the same thing as the static book portfolio: the fund tracked the Cambria Global Asset Allocation Index (to 2018 filings) which targets about 40% equity, 40% fixed income and 20% other assets and picks its ETFs each year by value, momentum and trend; the fund page currently states 45% equities, 45% fixed income and 10% alternatives, 0% management fee and a 0.32% total expense ratio. Because these funds select holdings with a rules-based algorithm, they are not a fixed list of weights, so this model uses the book's static composite.

The requested static buy-and-hold version is therefore the weights of the book composite as reproduced by AAII (with source 'Meb Faber Research') and by two portfolio sites that build it from ETFs (VV/VEA/VWO/VNQ/TLT/LQD/BNDX/TIP and DBC/IAU).

What is in it

US stocks	18%	VV or SPY in replications
Foreign developed stocks	13.5%	VEA / EFA
Foreign emerging stocks	4.5%	VWO / EEM
Corporate bonds	19.8%	LQD (investment grade)
30-year (US Treasury) bonds	13.5%	TLT
30-year foreign (international) bonds	14.4%	BNDX in replications
TIPS	1.8%	TIP / VTIP
Commodities	5%	DBC / GSG
Gold	5%	GLD / IAU
REITs	4.5%	VNQ

Source of the shares: AAII Journal, 'The Trinity Portfolio' (Table 1, source Meb Faber Research); cross-checked with PortfolioDB and Portfolio Einstein ETF builds — www.aaii.com. Table of ten weights summing to 100% (36 stocks, 49.5 bonds, 14.5 real assets). Same ten weights appear at PortfolioDB and Portfolio Einstein, which map them to ETFs. I could not open the book itself, so this is a secondary reproduction of the book's table.

How we model it

Corporate bonds	19.8%
US large companies (S&P 500)	18%
US bond market (aggregate)	14.4%
Developed markets outside the US	13.5%
Long-term government bonds (20+ y)	13.5%
Commodities	5%
Gold	5%
Emerging markets	4.5%
Real estate (REITs)	4.5%
Inflation-protected bonds (TIPS)	1.8%

Direct: US stocks = us_large, foreign developed = intl_dev, emerging = em, REITs = reit, corporate bonds = bond_corp, 30-year Treasuries = bond_long, TIPS = tips, commodities = commodities, gold = gold. Foreign (international) 30-year government bonds: we have no non-US bond series, so the 14.4% is mapped to bond_agg (intermediate aggregate US bonds). That cuts duration and drops currency and foreign-rate effects; the alternative mapping to bond_long would raise long-bond exposure to 27.9%, which is not what the published ETF (BNDX, intermediate duration) does. Bond total in the mapping is 49.5%, matching the published 49.5%. bond_corp starts in 1993-11 and tips in 2000-07, so the full portfolio can only be tested from 2000-07.

For a euro investor

The same portfolio bought with euros, without currency hedging: monthly returns converted at the end-of-month rate (German mark before 1999), inflation of the euro area. History in euros from 1984-02 to 2026-09.

Per year in euros7.4 %

After euro inflation5.2 %

Worst drop in euros−30.4 %

Worst year in euros−17.2 % (1987)

Safe withdrawal, 30 years, dollars6.46 % (1987-03)

Safe withdrawal, 30 years, euros5.13 % (1985-03)

The largest first-year withdrawal, then growing with inflation, that lasted 30 years from every monthly start. Test withdrawals for this portfolio →

How to build it with European funds

Examples of UCITS funds sold in the EU for each part of the portfolio. They track similar markets, not exactly the indices of our calculation.

Part	Share	Example funds (ISIN)
Corporate bonds	19.8%	iShares USD Corporate Bond UCITS ETF (Dist) `IE0032895942`
US large companies (S&P 500)	18%	iShares Core S&P 500 UCITS ETF USD (Acc) `IE00B5BMR087` Vanguard S&P 500 UCITS ETF (USD) Distributing `IE00B3XXRP09`
US bond market (aggregate) (closest match)	14.4%	iShares Core Global Aggregate Bond UCITS ETF EUR Hedged (Acc) `IE00BDBRDM35` iShares Core Global Aggregate Bond UCITS ETF USD (Dist) `IE00B3F81409`
Developed markets outside the US (closest match)	13.5%	iShares Core MSCI Europe UCITS ETF EUR (Acc) `IE00B4K48X80` iShares Core MSCI Japan IMI UCITS ETF `IE00B4L5YX21`
Long-term government bonds (20+ y)	13.5%	iShares USD Treasury Bond 20+yr UCITS ETF USD (Dist) `IE00BSKRJZ44`
Commodities	5%	iShares Diversified Commodity Swap UCITS ETF `IE00BDFL4P12`
Gold	5%	Invesco Physical Gold ETC `IE00B579F325` Xetra-Gold `DE000A0S9GB0`
Emerging markets	4.5%	iShares Core MSCI Emerging Markets IMI UCITS ETF (Acc) `IE00BKM4GZ66`
Real estate (REITs)	4.5%	iShares US Property Yield UCITS ETF `IE00B1FZSF77` iShares Developed Markets Property Yield UCITS ETF `IE00B1FZS350`
Inflation-protected bonds (TIPS)	1.8%	iShares USD TIPS UCITS ETF USD (Acc) `IE00B1FZSC47`

These are examples, not recommendations. Check the fund's key information document, costs and taxes with your broker before buying. Names and ISINs checked on justETF on 4 November 2026; Udak receives nothing from these funds.

The idea

Hold roughly the world's market portfolio, but tilted by what the famous investors do: about a third in stocks, half in bonds, with a slice in real assets as inflation hedge; the aim is a benchmark that is hard to beat after costs (Faber's survey finds that costs, not the allocation choice, matter most).

How people buy it

Third-party ETF builds use VV (or SPY), VEA, VWO, VNQ, TLT, LQD, BNDX, VTIP (or TIP), DBC or GSG, IAU or GLD. The Cambria GAA ETF is a different, rules-based fund of ETFs.

Rebalancing as published

Static weights with periodic rebalancing (PortfolioDB describes buy-and-hold with periodic rebalancing; the book's rebalancing rule was not confirmed). Cambria's index is rebalanced and reconstituted annually.

Variants

Cambria GAA ETF / Cambria Global Asset Allocation Index: about 40% equity, 40% fixed income, 20% other (2018 filing); fund page now says 45/45/10.
Faber's rounded global market portfolio of 27 June 2014: 20% US stocks, 20% foreign stocks, 20% corporate bonds, 30% government bonds, 5% real estate, 2% each TIPS, EM bonds and high yield.
Faber's Ivy Portfolio (five asset classes, equal weights) is a separate model (m_faber).

Criticism

The composite contains 28.8% in long-duration government bonds (US and foreign 30-year) and corporate bonds, so rising rates hurt: it fell 16.06% in 2022 in one replication and lost 20.6% from Dec 2021 to Sep 2022 in another.
A static copy of what 'prominent investors' own is not an independent view; weights are not derived from a clear rule (my reading).
Third-party backtests of the ETF build are mediocre against US 60/40: CAGR 7.51% vs 9.5% over 32 years in one site's test (US equity bull market period).
The Cambria ETF itself has a 0.32% acquired-fund expense ratio and a rules-based, evolving holding list; the static book weights are not what the fund holds.

Through the eras

Era	Dates	Total	Per year	Worst drop	S&P 500
Black Monday	1987-08 – 1987-12	−8.3 %	—	−12.6 %	−21.8 %
Savings-and-loan crisis and 1990 recession	1990-07 – 1991-03	+6.1 %	—	−6.3 %	+7.4 %
1994 bond market massacre	1994-02 – 1994-12	−4.2 %	—	−5.3 %	−3.2 %
Asian and Russian financial crises	1997-07 – 1998-10	+9.3 %	+6.9 %	−6.0 %	+22.4 %
Dot-com bubble and crash	2000-03 – 2002-10	+1.1 %	+0.4 %	−6.4 %	−37.2 %
Recovery and housing boom	2002-11 – 2007-09	+89.1 %	+13.8 %	−4.2 %	+97.0 %
Global financial crisis	2007-10 – 2009-03	−17.2 %	−11.8 %	−23.0 %	−44.7 %
The decade after the crisis	2009-04 – 2019-12	+156.4 %	+9.2 %	−6.7 %	+411.2 %
Fourth-quarter 2018 sell-off	2018-09 – 2018-12	−5.7 %	—	−5.7 %	−14.4 %
Covid crash and stimulus boom	2020-02 – 2021-12	+21.2 %	+10.6 %	−8.9 %	+53.6 %
Inflation shock and rate hikes	2022-01 – 2022-12	−16.7 %	−16.7 %	−21.0 %	−19.9 %

Calendar years

Any N-year period

Any N-year period	Worst	10 % worse than	Median	10 % better than	Best
1 y (501)	−21.5 %	−0.8 %	+9.4 %	+18.6 %	+38.7 %
3 y (477)	−1.8 %	+3.1 %	+8.5 %	+13.2 %	+23.9 %
5 y (453)	+2.5 %	+4.8 %	+8.1 %	+12.0 %	+17.6 %
10 y (393)	+3.9 %	+5.9 %	+8.4 %	+10.6 %	+13.3 %
15 y (333)	+4.8 %	+6.4 %	+8.3 %	+9.8 %	+12.2 %
20 y (273)	+6.4 %	+6.7 %	+8.3 %	+9.9 %	+11.0 %
30 y (153)	+6.9 %	+7.4 %	+8.3 %	+9.4 %	+10.3 %

How it behaved in history

1929-1949 and 1969-2009. Stocks underperformed bonds for 20 years (1929-1949) and 40 years (1969-2009), the argument for holding half the portfolio in bonds and some real assets (Faber's text as quoted by AAII). www.aaii.com
1973-1974 and 1970s inflation. The book's data start in 1973, so the oil-shock years are included in the 1973-2015 test (9.48% a year, maximum drawdown -26.72%). Year-by-year figures not seen. www.aaii.com
2000-2002 (dot-com bust). In an ETF-based backtest the portfolio returned +2.4% a year through the dot-com period while US 60/40 lost 4.3% a year. www.portfoliodb.com
2007-2009 (GFC). Maximum drawdown -23.54% from Oct 2007 to Feb 2009, recovery by Nov 2009; GFC-period CAGR +3.3% against -0.4% for US 60/40. www.portfoliodb.com
2022. Reported -16.06% (one site) and a -20.6% drawdown from Dec 2021 to Sep 2022 with recovery in Aug 2024 (another site); the long bond and corporate bond weights drove the loss. www.portfolioeinstein.com

Published results (by others)

Faber's backtest of the Global Asset Allocation composite (1973-2015): 9.48% a year, volatility 7.93%, maximum drawdown -26.72% (as reproduced by AAII) www.aaii.com
The book's strategies earned within about 1 percentage point of one another, except the Permanent Portfolio about 1 point lower (1973-2013): qualitative mymoneyblog.com
PortfolioDB backtest of the ETF-based static portfolio vs US 60/40 (about 32 years to 2025): CAGR 7.51% vs 9.5%; volatility 7.8% vs 10.0%; max drawdown -23.54% (Oct 2007 - Feb 2009) vs -29.7% www.portfoliodb.com
Portfolio Einstein ETF-based build (1989-2023): CAGR 7.29%; 2022 return -16.06% www.portfolioeinstein.com

What we could not verify

The weights are verified from three secondary reproductions (AAII, PortfolioDB, Portfolio Einstein) that agree; I could not open the book, Faber's blog post with the table, or a Cambria document that lists these exact numbers.
It is not confirmed that the book names this composite 'GAA' the same way the ETF is named; the ETF does not hold these weights.
The Cambria prospectus text was read only through a 2018 497K (40/40/20, index-based, annual reconstitution) and the fund page (45/45/10); the current prospectus was not read.
Whether the 30-year foreign bond slice in the book is hedged or unhedged is not known.
Rebalancing rule in the book is not confirmed.
Publication date: early 2015 (reviews on 28 Feb 2015 and in March 2015), exact release date not confirmed.
The performance figures of the replications (PortfolioDB, Portfolio Einstein) are third-party model results of unknown methodology and differing end dates.

Sources

www.aaii.com https://www.aaii.com/journal/article/the-trinity-portfolio-combining-diversification-tilts-and-trend-following?printerfriendly=true
www.portfoliodb.com https://www.portfoliodb.com/portfolios/global-asset-allocation-gaa-portfolio-faber
www.portfolioeinstein.com https://www.portfolioeinstein.com/meb-faber-portfolios-backed-by-solid-research/
mymoneyblog.com https://mymoneyblog.com/global-asset-allocation-book-review.html
mebfaber.com https://mebfaber.com/2014/06/27/this-is-the-global-market-portfolio/
cambriafunds.com https://cambriafunds.com/gaa
www.sec.gov https://www.sec.gov/Archives/edgar/data/0001529390/000143774918005515/cambria20180324h_497k.htm
blog.iese.edu https://blog.iese.edu/jestrada/files/2016/03/JOIM-BR-Faber.pdf
alphaarchitect.com https://alphaarchitect.com/book-review-global-asset-allocation-a-survey-of-the-worlds-top-asset-allocation-strategies/

Open in the builder

Put this card on your site →

Calculated on historical data. Past results do not predict the future. Models before the 1990s use stand-ins for classes that had no index yet; fees, taxes and fund costs are not included. This is mathematics, not an investment recommendation.