Global market portfolio
Global Multi-Asset Market Portfolio (Doeswijk, Lam, Swinkels)
Where it stands today
Our calculation, 1926-07 – 2026-09
Growth of $10,000
Where it came from
The idea of a 'market portfolio' is the theoretical benchmark of modern finance: if everyone holds the same risky portfolio, it must be all risky assets in proportion to their market value. In practice most studies use a stock index such as the S&P 500 as a stand-in, which Doeswijk and Swinkels point out is a markedly smaller object (32 trillion dollars) than the 150 trillion dollars of assets held by financial investors worldwide. The most cited earlier attempt to measure the entire world portfolio of investable assets is the 'world wealth portfolio' of Ibbotson and Siegel (1983, updated 1985), which covered 1960-1980 (as recounted in the 2024 Doeswijk-Swinkels paper).
In 2013-2014 Ronald Doeswijk, Trevin Lam and Laurens Swinkels (the Robeco data page lists Swinkels as Head of Solutions Research at Robeco, the Rotterdam asset manager, and he is also at Erasmus School of Economics; the employers of the other two in 2014 were not checked) published 'The Global Multi-Asset Market Portfolio 1959-2012'. It first appeared on SSRN in 2013 and was published in the Financial Analysts Journal in 2014 (volume 70, issue 2); a 2014 Graham and Dodd Readers' Choice Award is mentioned by Robeco. They estimated the market capitalisation of eight asset classes held by financial investors: equities, private equity, real estate, high-yield bonds, emerging-market debt, investment-grade credit, government bonds and inflation-linked bonds. The full eight classes are available from 1990 to 2012, and four main categories (equities, real estate, non-government bonds, government bonds) from 1959. Their headline result: the global invested portfolio was worth about 90.6 trillion dollars at end-2012; equities' share fell from about 52 percent in 1990 to about 36 percent in 2012, while government bonds were about 30 percent, investment-grade credit rose from about 11 to 18 percent, and the five smallest classes together rose from about 6 to 16 percent.
The authors argue that the market portfolio is a 'natural benchmark' for strategic asset allocation and a sensible starting point for approaches like Black-Litterman or adaptive allocation, because an investor's own deviations from it become explicit active bets. A letter to the editor by Westerling (2014) on central bank asset holdings and the authors' reply were also published in the Financial Analysts Journal.
In 2019-2020 Doeswijk, Lam and Swinkels extended the data to 1960-2017 and studied the returns in 'Historical Returns of the Market Portfolio' (Review of Asset Pricing Studies 10(3), 2020, pp. 521-567). They found a compounded real US-dollar return of 4.45 percent a year with a standard deviation of 11.2 percent over 1960-2017 (Robeco data page, January 2020). The dataset 'Data update: The Global Multi-Asset Market Portfolio, 1959-2012' is refreshed on a data repository of Erasmus University Rotterdam and, in its latest version (posted on DataverseNL in August 2025), contains annual data until 31 December 2023.
In 2024 Doeswijk (now independent) and Swinkels published 'The Risk and Reward of Investing', a version at monthly frequency from 1970 to 2022 and worth 150 trillion dollars, with five asset categories (equities broad incl. private equity; real estate; non-government bonds; government bonds broad incl. emerging-market and inflation-linked bonds; commodities broad, mostly gold and also cryptocurrencies). The authors found that over 1970-2022 the global market portfolio had a geometric average excess return of 0.30 percent a month and monthly volatility of 2.90 percent, a Sharpe ratio not much higher than that of equities (0.408 against 0.394), but much more stable over rolling ten-year windows and with shallower and shorter drawdowns. Robeco is also cited as the source of a free spreadsheet of annual weights and returns.
What is in it
| Equities (listed, excluding listed REITs) | 43% | End-2023, computed from the published USD amounts in Table A1 of the dataset (74.07 of 172.4 trillion dollars). End-2012: 36.3 percent. |
| Private equity | 7.8% | End-2023 (13.42 trillion dollars; Preqin 'private capital'). End-2012: 3.6 percent. |
| Real estate | 4% | End-2023 (6.98 trillion dollars). End-2012: 5.1 percent. Includes listed and an estimated part of unlisted investable real estate; directly held property is excluded. |
| High-yield bonds | 1.1% | End-2023 (1.85 trillion dollars). End-2012: 1.7 percent. |
| Emerging-market debt | 3.9% | End-2023 (6.68 trillion dollars). End-2012: 3.0 percent. |
| Investment-grade credit | 14.7% | End-2023 (25.43 trillion dollars; derived as Multiverse index minus government and high yield). End-2012: 18.5 percent. |
| Government bonds | 23.9% | End-2023 (41.15 trillion dollars). End-2012: 29.5 percent. |
| Inflation-linked bonds | 1.6% | End-2023 (2.82 trillion dollars). End-2012: 2.3 percent. |
How we model it
| US large companies (S&P 500) | 26% |
| Intermediate government bonds (10 y) | 23.9% |
| Corporate bonds | 15.8% |
| Developed markets outside the US | 12% |
| US small companies | 7.8% |
| Emerging markets | 5% |
| Real estate (REITs) | 4% |
| US bond market (aggregate) | 3.9% |
| Inflation-protected bonds (TIPS) | 1.6% |
For a euro investor
How to build it with European funds
| Part | Share | Example funds (ISIN) |
|---|---|---|
| US large companies (S&P 500) | 26% | iShares Core S&P 500 UCITS ETF USD (Acc) IE00B5BMR087Vanguard S&P 500 UCITS ETF (USD) Distributing IE00B3XXRP09 |
| Intermediate government bonds (10 y) | 23.9% | iShares USD Treasury Bond 7-10yr UCITS ETF (Dist) IE00B1FZS798 |
| Corporate bonds | 15.8% | iShares USD Corporate Bond UCITS ETF (Dist) IE0032895942 |
| Developed markets outside the US (closest match) | 12% | iShares Core MSCI Europe UCITS ETF EUR (Acc) IE00B4K48X80iShares Core MSCI Japan IMI UCITS ETF IE00B4L5YX21 |
| US small companies | 7.8% | SPDR Russell 2000 U.S. Small Cap UCITS ETF IE00BJ38QD84iShares MSCI USA Small Cap ESG Enhanced CTB UCITS ETF USD (Acc) IE00B3VWM098 |
| Emerging markets | 5% | iShares Core MSCI Emerging Markets IMI UCITS ETF (Acc) IE00BKM4GZ66 |
| Real estate (REITs) | 4% | iShares US Property Yield UCITS ETF IE00B1FZSF77iShares Developed Markets Property Yield UCITS ETF IE00B1FZS350 |
| US bond market (aggregate) (closest match) | 3.9% | iShares Core Global Aggregate Bond UCITS ETF EUR Hedged (Acc) IE00BDBRDM35iShares Core Global Aggregate Bond UCITS ETF USD (Dist) IE00B3F81409 |
| Inflation-protected bonds (TIPS) | 1.6% | iShares USD TIPS UCITS ETF USD (Acc) IE00B1FZSC47 |
The idea
If markets are in equilibrium, the market portfolio is the one that all investors in aggregate must hold; its weights are set by the sum of all investors' decisions rather than by any manager's forecasts. It therefore needs no return forecasts, has low turnover (weights move with prices) and shows what an investor with no special view would own. The authors present it as a natural benchmark for strategic asset allocation against which every deviation is an explicit bet, and as an anchor for approaches that adjust it with a view.
How people buy it
There is no single fund. A practical replica combines a global all-cap equity index fund, a global government and credit bond index fund (possibly currency hedged), inflation-linked bonds, global REITs, and optionally private-equity exposure through listed vehicles; weights should follow the latest published market value shares (updated about once a year in the authors' dataset). Exact replication is not possible: directly held property, human capital and family businesses are excluded from the invested market portfolio, and private equity is hard to hold.
Rebalancing as published
The authors do not give a rebalancing rule: the portfolio is defined by market capitalisation, so weights drift with prices without trading (a buy-and-hold property). The dataset is updated once a year using end-of-year market values; the 2024 version is monthly.
Variants
- Doeswijk-Lam-Swinkels (2014): eight asset classes, 1990-2012 (four main categories from 1959).
- Doeswijk-Lam-Swinkels (2020): annual data 1960-2017, includes private equity, inflation-linked bonds and emerging-market debt but excludes privately held real estate.
- Doeswijk-Swinkels (2024): five asset categories, monthly 1970-2022, adds cryptocurrencies and commodities held for investment (mostly gold); average weights equities broad 49.7%, real estate 4.4%, non-government bonds 15.4%, government bonds broad 28.1%, commodities 2.4%.
- Ibbotson and Siegel (1983): earlier 'world wealth portfolio' for 1960-1980.
Criticism
- Only invested assets are measured. The authors exclude human capital, family businesses, durable goods and directly held property; they note that all developed-world real estate (including mortgages) was about 217 trillion dollars in 2015 against about 104 trillion dollars for their entire invested portfolio, so a large part of wealth is outside the 'market portfolio'.
- Data limits: the dataset note warns that countries such as China have entered developed-market bond indexes, which leads to some double counting with emerging-market debt; the private-equity figure is Preqin 'private capital' (which includes dry powder) and the 2023 update uses a June 2023 figure back-filled for 2022.
- Measurement choices matter: Westerling (2014) raised the issue of central-bank asset holdings in a published comment, and the authors replied; the 2024 paper shows that when returns are expressed in currencies other than the US dollar the risk of investing looks larger.
- Estimation of drawdown risk with annual data was imprecise, which is why the authors moved to monthly data in 2024.
- The market-portfolio concept itself is an equilibrium idea that cannot be observed exactly (general academic critique; not checked in this research).
Through the eras
| Era | Dates | Total | Per year | Worst drop | S&P 500 |
|---|---|---|---|---|---|
| Roaring Twenties boom | 1926-07 – 1929-08 | +75.1 % | +19.4 % | −3.5 % | +153.7 % |
| Wall Street crash and Great Contraction | 1929-09 – 1932-07 | −51.4 % | −21.9 % | −58.3 % | −78.2 % |
| New Deal recovery | 1932-08 – 1937-02 | +154.1 % | +22.6 % | −18.8 % | +268.7 % |
| Recession within the Depression | 1937-03 – 1938-03 | −29.9 % | −28.0 % | −30.1 % | −49.5 % |
| Rearmament and World War II | 1938-04 – 1945-09 | +123.8 % | +11.3 % | −14.2 % | +205.8 % |
| Postwar expansion | 1945-10 – 1965-12 | +474.1 % | +9.0 % | −14.8 % | +1212.9 % |
| Golden age of the 1950s | 1950-01 – 1959-12 | +188.9 % | +11.2 % | −8.6 % | +438.8 % |
| Go-go years and nifty-fifty | 1966-01 – 1968-12 | +28.3 % | +8.7 % | −9.7 % | +34.0 % |
| The Great Inflation (stagflation) | 1969-01 – 1982-07 | +114.1 % | +5.8 % | −26.2 % | +84.6 % |
| 1973-74 oil shock and bear market | 1973-01 – 1974-12 | −21.9 % | −11.6 % | −26.2 % | −41.7 % |
| Second oil shock and the Volcker disinflation | 1979-01 – 1982-08 | +48.5 % | +11.4 % | −10.2 % | +59.2 % |
| Great bull market in stocks and bonds | 1982-08 – 1999-12 | +970.3 % | +14.6 % | −17.7 % | +2060.4 % |
| Black Monday | 1987-08 – 1987-12 | −12.6 % | — | −17.7 % | −21.8 % |
| Savings-and-loan crisis and 1990 recession | 1990-07 – 1991-03 | +6.2 % | — | −10.5 % | +7.4 % |
| 1994 bond market massacre | 1994-02 – 1994-12 | −5.0 % | — | −6.4 % | −3.2 % |
| Asian and Russian financial crises | 1997-07 – 1998-10 | +11.6 % | +8.6 % | −8.7 % | +22.4 % |
| Dot-com bubble and crash | 2000-03 – 2002-10 | −5.1 % | −1.9 % | −9.7 % | −37.2 % |
| Recovery and housing boom | 2002-11 – 2007-09 | +85.5 % | +13.4 % | −4.1 % | +97.0 % |
| Global financial crisis | 2007-10 – 2009-03 | −23.0 % | −16.0 % | −28.7 % | −44.7 % |
| The decade after the crisis | 2009-04 – 2019-12 | +189.3 % | +10.4 % | −7.7 % | +411.2 % |
| Fourth-quarter 2018 sell-off | 2018-09 – 2018-12 | −7.7 % | — | −7.7 % | −14.4 % |
| Covid crash and stimulus boom | 2020-02 – 2021-12 | +25.4 % | +12.5 % | −10.7 % | +53.6 % |
| Inflation shock and rate hikes | 2022-01 – 2022-12 | −17.6 % | −17.6 % | −21.9 % | −19.9 % |
Calendar years
Any N-year period
| Any N-year period | Worst | 10 % worse than | Median | 10 % better than | Best |
|---|---|---|---|---|---|
| 1 y (1192) | −39.7 % | −5.6 % | +9.9 % | +23.1 % | +88.0 % |
| 3 y (1168) | −23.4 % | +0.7 % | +8.9 % | +15.9 % | +29.8 % |
| 5 y (1144) | −8.6 % | +3.6 % | +8.4 % | +13.7 % | +23.6 % |
| 10 y (1084) | +0.7 % | +5.4 % | +8.5 % | +13.3 % | +16.0 % |
| 15 y (1024) | +3.0 % | +6.1 % | +8.6 % | +13.0 % | +15.2 % |
| 20 y (964) | +4.0 % | +6.8 % | +8.9 % | +12.6 % | +14.3 % |
| 30 y (844) | +6.6 % | +7.7 % | +9.1 % | +11.0 % | +12.4 % |
How it behaved in history
- 1968. Equities' share of the invested portfolio peaked near 64 percent (1959-2012 data). www.cxoadvisory.com
- 1972-1982 (high inflation, rising yields). Government bond share fell to a low of 19.8 percent in November 1972 and rose to a maximum of 35.8 percent in July 1982; non-government bonds had a maximum drawdown of 38.4 percent in June 1982 and government bonds broad 28.9 percent in September 1981 (excess return basis). d1e00ek4ebabms.cloudfront.net
- 1974. Real estate's largest drawdown (65.8 percent in excess terms) ended in December 1974. d1e00ek4ebabms.cloudfront.net
- 1979-1980. Commodities' share peaked at 11.1 percent in February 1980, after the Hunt brothers' attempt to corner silver. d1e00ek4ebabms.cloudfront.net
- 1999-2000 (dot-com). Equities' share peaked at 63.5 percent in March 2000. d1e00ek4ebabms.cloudfront.net
- 2008-2009 (credit crisis). The global market portfolio's maximum drawdown (36.0 percent of excess return index) bottomed in February 2009 when equities' share hit its low of 35.4 percent. d1e00ek4ebabms.cloudfront.net
- 2012-2023. Total invested assets almost doubled from 90.6 to 172.4 trillion dollars; the listed-equity share rose from about 36 to 43 percent, private equity from 3.6 to 7.8 percent and government bonds fell from 29.5 to 23.9 percent (our division of the published table). dataverse.nl
Published results (by others)
- Compounded real return and volatility of the global market portfolio in US dollars (1960-2017): 4.45 percent per year real (compounded), standard deviation 11.2 percent www.robeco.com
- Monthly excess return of the global market portfolio over Treasury bills (1970-2022): geometric 0.30 percent per month, arithmetic 0.34 percent, standard deviation 2.90 percent; Sharpe ratio 0.408 versus 0.394 for equities broad d1e00ek4ebabms.cloudfront.net
- Maximum drawdown of the excess-return index (relative to cash) (1970-2022): global market portfolio 36.0 percent (bottom February 2009); equities broad 57.7 percent (February 2009); real estate 65.8 percent (December 1974); non-government bonds 38.4 percent (June 1982); government bonds broad 28.9 percent (September 1981) d1e00ek4ebabms.cloudfront.net
- Calendar-year excess return of the global market portfolio (1970-2022): compounded 3.65 percent, average 4.33 percent, standard deviation 11.73 percent per year (equities broad: 4.97, 6.65, 18.08) d1e00ek4ebabms.cloudfront.net
- Market value of the invested multi-asset portfolio (end-2012 and end-2023): about 90.6 trillion dollars (2012) and 172.4 trillion dollars (2023) dataverse.nl
- Equity share of the market portfolio over time (1990 and 2012; 1959-2012): 52 percent (1990) to 36 percent (2012); range 37 percent (2011) to 63-64 percent (1999 and 1968), average 52 percent www.cxoadvisory.com
- Real return in inflationary and disinflationary eras (working-paper version, data to 2015) (1960-1979 and 1980-2015): 2.27 percent versus 5.57 percent compounded real quantpedia.com
- Dependence on the base currency (2020 paper) (1960-2017): compounded real returns 2.87 to 4.93 percent and maximum drawdowns 34.9 to 56.6 percent across ten currencies papers.ssrn.com
What we could not verify
- The authors are Ronald Q. Doeswijk with co-authors Trevin Lam and Laurens Swinkels; a publication titled exactly 'What Is the Market Portfolio?' was not found.
- Percent weights for 2023 are our arithmetic on the published USD amounts in the dataset; the authors' own percentage table for 2023 was not seen. The 1990 weights for all eight classes were not retrieved (only equities 52 percent, investment-grade credit about 11 percent and the five small classes combined about 6 percent from a secondary summary).
- Weights of the 2024 monthly version at end-2022 (including gold/commodities and crypto) were not extracted; only 1970-2022 averages were found.
- The regional split of listed equities (60/28/12) is our assumption; MSCI ACWI IMI composition at end-2023 was not checked.
- Dates of the SSRN first posting (2013) and the 2014 Graham and Dodd award are from search results and a Robeco page; the FAJ publication details are from the dataset readme.
- Practical implementation advice is general; no specific funds were verified.
- The Roll-type critique that the true market portfolio is unobservable is general academic knowledge and was not verified against a source in this session.
Sources
- dataverse.nl
https://dataverse.nl/dataset.xhtml?persistentId=doi:10.34894/CBXAIY - doi.org
https://doi.org/10.2469/faj.v70.n2.1 - papers.ssrn.com
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2352932 - www.robeco.com
https://www.robeco.com/en-int/insights/2020/01/data-sets-historical-returns-of-the-market-portfolio - papers.ssrn.com
https://papers.ssrn.com/abstract=2978509 - quantpedia.com
https://quantpedia.com/historical-returns-of-the-market-portfolio/ - d1e00ek4ebabms.cloudfront.net
https://d1e00ek4ebabms.cloudfront.net/production/uploaded-files/Doeswijk%20Swinkels%202024%20-%20Risk%20and%20Reward%20of%20Investing%20(1)-51bbafa8-05e1-425a-9587-87fa6c331fc6.pdf - www.eur.nl
https://www.eur.nl/en/news/ultimate-market-portfolio - www.cxoadvisory.com
https://www.cxoadvisory.com/strategic-allocation/global-benchmark-portfolio
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.