Portfolio Lab

Model portfolios compared: ten well-known allocations

Basics9 min read · updated 16 November 2026 · 9 sources

A model portfolio is a fixed allocation to a few broadly diversified funds that someone has published and anyone can copy. The best-known ones come from the United States, where they are called lazy portfolios because they need almost no work once they are set up. They differ less than their names suggest: almost everything depends on the equity share and a few tilts. This article puts ten allocations side by side and shows what to check before you pick one.

What makes a model portfolio

A model portfolio consists of three things: a list of asset classes, fixed weights and the rule to restore the weights regularly (rebalancing). There are no forecasts and no picking of individual stocks. The author decides once how much goes into stocks, bonds and other building blocks, and sticks to it. That is what separates a model portfolio from a tactical strategy that changes its weights based on signals.

The idea is old. Bogle (2007) argues for buying the whole market cheaply instead of looking for winners, and Bernstein (2000) shows how a few building blocks are enough to spread a portfolio across regions and size segments. In Germany the global portfolio after Kommer (2018) stands for the same idea. Three related concepts that aim to spread risk rather than capital have their own article: All Weather, Permanent Portfolio and Golden Butterfly.

Ten well-known allocations

The table lists the allocations as they are stored as templates in Portfolio Lab. Many authors have published several versions, for example with more or fewer bonds. Read the weights as a common version, not as the only valid one.

PortfolioAuthorAllocation
Buffett 90/10Warren Buffett90% US large caps, 10% short-term government bonds
No-BrainerWilliam Bernstein25% each in US large caps, US small caps, developed markets ex US, short-term government bonds
Core FourRick Ferri48% US stocks, 24% developed markets ex US, 8% real estate stocks, 20% bonds
Three-fund portfolioBogleheads communityDomestic stocks, international stocks, bonds; in the Europe template 60% developed markets, 10% emerging markets, 30% bonds
Ultimate Buy and HoldPaul Merrimanten equity building blocks at 6% each (including real estate stocks), 40% bonds
Swensen portfolioDavid Swensen30% US stocks, 15% developed markets, 5% emerging markets, 20% real estate stocks, 15% Treasuries, 15% inflation-protected bonds
Coffeehouse portfolioBill Schultheissix equity building blocks at 10% each (including real estate stocks), 40% bonds
Ivy PortfolioMeb Faber20% each in US stocks, developed markets ex US, government bonds, real estate stocks, commodities
Larry PortfolioLarry Swedroe30% stocks (small value stocks and emerging markets), 70% bonds
Permanent PortfolioHarry Browne25% each in stocks, long-term government bonds, cash, gold
Ten model portfolios by asset class

Weights of the templates in Portfolio Lab, grouped by asset class and sorted by equity share. Real estate stocks (REITs) are shown separately because they are a building block of their own in several portfolios.

Buffett describes his allocation in the shareholder letter for 2013 as an instruction to the trustee for his wife's inheritance (Buffett 2014). Swensen ran the Yale University endowment and described its approach for institutional investors in Swensen (2000); the version for individual investors comes from his later book "Unconventional Success". The Ivy Portfolio holds the same five asset classes that Faber (2007) examines with a trend rule, but does without the rule.

Where the portfolios really differ

The chart shows the most important difference at a glance: the equity share. It ranges from 25% in the Permanent Portfolio to 90% for Buffett. If you count real estate stocks as stocks, seven of the ten portfolios sit between 60% and 80%. Anyone comparing two model portfolios with very different equity shares is mainly comparing two risk levels and only then two ideas.

  • Number of building blocks: Buffett gets by with two funds, Merriman needs twelve. More building blocks mean more rebalancing work and often higher costs, because not every building block is available as a cheap fund.
  • Equity tilts: Coffeehouse, Merriman and Larry weight small companies and value stocks more heavily than the market. Behind this is the research on factor premiums (see Fama-French factors). The premiums are not guaranteed and can stay away for many years.
  • Real assets: Swensen, Ivy and Core Four hold real estate stocks as a building block of their own, Ivy adds commodities, the Permanent Portfolio gold. These shares are meant to help when inflation rises.
  • Type of bonds: Short-term government bonds (Buffett, No-Brainer) dampen swings but carry little interest rate risk and little extra return. Long-term government bonds (Permanent Portfolio) react strongly to changes in interest rates (see duration).
  • Home market: Almost all of the allocations weight the United States heavily because they were written for US investors. That is a decision you should make deliberately rather than inherit (see home bias).

Worked example: equity share and risk

A calculation with just two assets shows how strongly the equity share drives risk. Assumptions: stocks have a volatility of 16% a year, bonds 5%, and the correlation is 0.1. The volatility of the mix follows from the familiar formula for two assets.

equity share
volatility of stocks and bonds
correlation between the two

A bad year makes it even more tangible. Suppose stocks lose 40% and bonds gain 5%. Then the loss of the portfolio depends on the weights alone. On $100,000:

The calculation simplifies a lot: it knows only two asset classes, fixed volatilities and a fixed correlation. In crises correlations often rise, and building blocks such as real estate stocks or commodities behave differently from broad stocks or bonds. It is still good enough for the order of magnitude, and that is exactly what you should know before you settle on an allocation (see equity share by age).

Putting a model portfolio into practice

The originals name specific index funds, and most of them now have a cheaper ETF counterpart. You translate each building block into a fund you can actually buy in your account. Three decisions come up along the way.

  1. US stocks or the world: Instead of buying "US stocks" and "developed markets ex US" separately, one global fund is often enough. The United States has the largest weight in it anyway (see MSCI World, ACWI and All-World).
  2. Bonds in your own currency: The bond building blocks of the originals are in dollars, which suits a US investor. Anyone who spends in another currency is better off with bonds in that currency or with currency-hedged funds, otherwise the exchange rate drives the result of the safe part (see currency risk).
  3. Building blocks without a cheap fund: Small value stocks outside the United States or broad commodity baskets are available in only a few funds with higher costs. Leaving them out changes the portfolio less than you might think, as long as the equity share stays the same.

Taxes and costs are part of putting it into practice. In a taxable account every rebalancing with sales can trigger capital gains tax, and many building blocks mean many small trades. Two to four funds reproduce most model portfolios well enough.

What the research shows

Two widely cited papers show why the allocation across asset classes matters so much. Brinson et al. (1986) find for large US pension plans that the long-term allocation explains by far the largest part of the variation of returns over time, while security selection and timing contribute little. Ibbotson and Kaplan (2000) put the result into perspective and ask in the very title whether the allocation explains 40, 90 or 100 percent: the answer depends on whether you look at the variation of one fund over time, the differences between funds or the level of return.

Research on optimization also speaks for simple rules. DeMiguel et al. (2009) compare 14 models with plain equal weighting and find that none of them is consistently better out of sample. The reason is estimation error: expected returns and covariances can only be estimated imprecisely from the past (see covariance estimation). A model portfolio with round weights is therefore not a naive solution but a robust one.

Which of the ten allocations was "the best" depends on the period. In decades with strong US stocks, portfolios with a high equity share and a high US share come out ahead, in years with high inflation those with gold and commodities. Anyone who picks the winner from a comparison over the last ten years often just buys the asset class that did well most recently (see backtest overfitting).

Limits and common mistakes

  • Comparing returns alone: A portfolio with 90% stocks has almost always returned more than one with 30%. That says nothing about the idea, only about the risk level. A meaningful comparison uses the Sharpe ratio and maximum drawdown.
  • Allocation does not fit the person: The best allocation is useless if you abandon it in a downturn. The equity share should fit your own time horizon and your own tolerance for losses.
  • Too many building blocks: Twelve funds at 6% each are a lot of work and change little. Small positions barely move the overall result.
  • Copying the home bias: Anyone who copies the originals one to one holds a lot of dollars and a lot of US stocks. That can be intended, but it should be a decision.
  • Constant switching: The advantage of a model portfolio comes from sticking with it. Anyone who switches to last year's winner every year is doing market timing by a detour.

How Portfolio Lab calculates

In the backtest you will find all of the allocations mentioned here as templates, the US originals with long time series from index funds and several Europe versions with UCITS ETFs. You can run up to three portfolios side by side, with regular contributions or withdrawals, rebalancing by date or tolerance band and, if you like, with costs. The comparison shows return p.a., volatility, the largest drawdown and every calendar year. Monte Carlo shows the range of possible paths, and the optimization shows how far a model portfolio lies from the efficient frontier of its building blocks. A backtest describes the past and is not a forecast.

Frequently asked questions

What is a model portfolio?

A published, fixed allocation to a few broadly diversified funds that can be copied with ETFs and is only occasionally reset to its target weights.

Which model portfolio is the best?

That depends on the period and on the risk. Portfolios with a high equity share have mostly returned more over the long run, with much larger drawdowns. More important than the name is that the equity share fits your own tolerance for losses. Not investment advice.

What is the three-fund portfolio?

An allocation of three broad index funds: domestic stocks, international stocks and bonds. There are no fixed weights; equity shares of 60% to 80% are common.

Can I build a lazy portfolio with ETFs instead of the original mutual funds?

Yes. Each building block is replaced by a low-cost ETF on the same market. A few building blocks, such as small value stocks outside the United States, are only available in a limited number of funds.

How often do I need to rebalance a model portfolio?

Once a year or when the weights drift far from their targets is common. Rebalancing more often usually adds little and costs fees and taxes.

What is Buffett's 90/10 about?

In the shareholder letter for 2013, Warren Buffett recommended for his wife's inheritance 90% in a low-cost index fund on US large caps and 10% in short-term government bonds.

Literature

  1. Bernstein, William J. (2000): The Intelligent Asset Allocator. McGraw-Hill.
  2. Bogle, John C. (2007): The Little Book of Common Sense Investing. Wiley.
  3. Brinson, Gary P.; Hood, L. Randolph; Beebower, Gilbert L. (1986): Determinants of Portfolio Performance. Financial Analysts Journal 42 (4), pp. 39–44. doi:10.2469/faj.v42.n4.39
  4. Buffett, Warren E. (2014): Berkshire Hathaway Inc.: Letter to Shareholders 2013 (letter dated February 28, 2014). Berkshire Hathaway Inc., Omaha; accessed October 25, 2026. Source
  5. DeMiguel, Victor; Garlappi, Lorenzo; Uppal, Raman (2009): Optimal Versus Naive Diversification: How Inefficient is the 1/ N Portfolio Strategy?. Review of Financial Studies 22 (5), pp. 1915–1953. doi:10.1093/rfs/hhm075
  6. Faber, Mebane T. (2007): A Quantitative Approach to Tactical Asset Allocation. The Journal of Wealth Management 9 (4), pp. 69–79. doi:10.3905/jwm.2007.674809
  7. Ibbotson, Roger G.; Kaplan, Paul D. (2000): Does Asset Allocation Policy Explain 40, 90, or 100 Percent of Performance?. Financial Analysts Journal 56 (1), pp. 26–33. doi:10.2469/faj.v56.n1.2327
  8. Kommer, Gerd (2018): Souverän investieren mit Indexfonds und ETFs: Wie Privatanleger das Spiel gegen die Finanzbranche gewinnen [Investing confidently with index funds and ETFs: how private investors win the game against the financial industry]. Campus Verlag, 5th edition.
  9. Swensen, David F. (2000): Pioneering Portfolio Management: An Unconventional Approach to Institutional Investment. Free Press.

Full bibliography

This article explains a method; it is not investment or tax advice. Worked examples are simplified and rounded. All terms in the glossary.