What is equal weighting (1/N)? Equal money, unequal risk
On this page
Key takeaways
- Equal weighting puts the same amount of money in every holding and resets to equal shares when prices drift apart; nothing about the holdings has to be estimated.
- Equal money is not equal risk: the most volatile holdings, and those that move together, carry most of the portfolio's swings, while a calm holding that moves on its own can carry almost none.
- DeMiguel, Garlappi and Uppal (2009) found that a long list of optimized portfolios did not consistently beat simple equal weighting on data they had not seen, because errors in the estimated inputs, above all expected returns, cost more than optimizing gained.
- The rule moves the real decision to the list of holdings: equal shares of four stock funds and one bond fund make mostly a stock portfolio.
- On MarketHeist, equal-weight templates such as the Ivy Portfolio rebalance back to equal weights every December.
Definition
Equal weighting, or the 1/N rule, puts the same share of the money in each of N holdings and rebalances back to those shares from time to time. It needs no estimate of returns, volatilities or correlations, which makes it immune to estimation error, but it balances money rather than risk.
Also known as: 1/N portfolio, equal-weight portfolio, naive diversification, equally weighted portfolio, 1 over N.
How it works
The recipe is one line: with N holdings, put 1/N of the money in each. Over time the winners grow into larger shares and the losers shrink, so the portfolio is rebalanced, on a calendar or when the weights drift too far, by selling some of what rose and buying what fell.
Its appeal is what it leaves out. Mean-variance optimization needs an expected return for every holding, a volatility for every holding and a correlation for every pair, all estimated from noisy history. The optimizer then leans hardest on the inputs that look most favourable, which are often the ones most overstated by chance. Equal weighting cannot make that mistake because it uses no inputs at all.
In 2009 Victor DeMiguel, Lorenzo Garlappi and Raman Uppal put this to a test. Across several data sets of US and international stock portfolios, they compared 1/N with a long list of optimized strategies, each re-estimated on a rolling window and judged on the following months. None beat 1/N consistently on the Sharpe ratio or the other measures they used. Their estimate of the history needed for sample-based mean-variance weights to win was measured in centuries of monthly data, far more than any investor has.
The rule is like splitting a restaurant bill evenly: it is quick, fair-looking and avoids arguments over who had what, but whoever ordered the expensive dishes is quietly subsidized. In a portfolio the expensive dishes are the volatile holdings: they get the same money as the calm ones and contribute far more of the swings.
That is where risk-based methods start. Inverse volatility weighting sizes each holding by one over its volatility; equal risk contribution also accounts for correlation until every holding carries the same share of risk. Both need volatility and correlation estimates, but not expected returns, the input DeMiguel and co-authors found hardest to estimate.
Formula
- wᵢ
- weight of holding i at each rebalance
- N
- number of holdings
- RCᵢ
- share of portfolio risk carried by holding i (the shares sum to 1)
- Σ
- covariance matrix of the holdings' monthly returns
- σᵢⱼ
- covariance between holdings i and j (σᵢᵢ is holding i's own squared volatility)
- w
- vector of weights
With equal weights the 1/N cancels: a holding's risk share is its row of the covariance matrix over the sum of the whole matrix. Volatile holdings and holdings that move with the rest have large rows; a calm holding uncorrelated with the others has a row close to zero.
- σₚ
- volatility of the equal-weight portfolio
- σ²̄
- average of the holdings' squared volatilities
- cov̄
- average covariance over all pairs of different holdings
As N grows, the first term shrinks toward zero and the portfolio's risk approaches the average covariance: adding holdings removes what is particular to each one, never what they share.
Worked example
Five asset classes in equal parts, the mix Mebane Faber's Ivy Portfolio uses: US stocks (VTI), international stocks (VEU), intermediate Treasuries (IEF), commodities (DBC) and US real estate (VNQ), weighted from their monthly returns from Jun 1996 to Sep 2026.
With equal weights each fund holds 20.0% of the money, but the risk is nowhere near equal. IEF carries just 0.5% of the portfolio's risk; the four other funds carry almost all of it, the largest share being 27.8%.
Two things put IEF at the bottom. It is the calmest of the five, and its correlations with the other four have hovered around zero, so its swings mostly offset rather than add to theirs. Stocks at home and abroad and real estate, by contrast, tend to rise and fall together.
Balancing risk instead of money takes very different weights. Equal risk contribution gives every fund 20.0% of the risk by raising IEF's weight to 54.0%, and the portfolio's volatility falls from 11.2% with equal weights to 7.0%; inverse volatility lands between, at 8.5%. On the co-movement groups count of independent bets, equal weights read 1.73 and equal risk contribution 2.59.
None of this tests what DeMiguel, Garlappi and Uppal tested. They asked how 1/N performed out of sample against portfolios optimized on estimated returns; the weights here are fitted with hindsight to the whole window, information no investor had at its start. What the lab shows is the other side of 1/N's simplicity: the money is balanced, the risk is not, and the list of holdings decides which risks dominate.
Data through Sep 2026. Reproduce id 427a76e6a013b3b1 (this guide, its lab setup and the data month).
Interactive lab
Money share against risk share for equal weights, with inverse volatility and equal risk contribution for contrast. Swap in other funds: equal weights come closest to equal risk when the holdings have similar volatility and similar correlations.
| Method | Volatility | Largest weight | Largest risk share | Bets (co-movement groups) |
|---|---|---|---|---|
| Equal weight | 11.2% | 20% | 28% | 1.73 |
| Inverse volatility | 8.5% | 41% | 27% | 2.01 |
| Equal risk contribution | 7.0% | 54% | 20% | 2.59 |
| Method | VTI | VEU | IEF | DBC | VNQ |
|---|---|---|---|---|---|
| Equal weight | 20.0%risk 23.7% | 20.0%risk 26.6% | 20.0%risk 0.5% | 20.0%risk 21.4% | 20.0%risk 27.8% |
| Inverse volatility | 16.6%risk 25.3% | 15.5%risk 26.5% | 41.4%risk 7.4% | 13.2%risk 16.6% | 13.3%risk 24.2% |
| Equal risk contribution | 12.3%risk 20.0% | 10.7%risk 20.0% | 54.0%risk 20.0% | 13.2%risk 20.0% | 9.7%risk 20.0% |
Weight, with the holding's share of the risk beneath it. Tree order of the holdings: IEF, DBC, VNQ, VTI, VEU. Jun 1996 to Sep 2026.
How MarketHeist computes it
Every choice an implementation has to make, as the code makes it today. The page is checked against the code each time it is built.
| Choice | How it is computed | In the code |
|---|---|---|
| Data | Monthly total returns (dividends reinvested). Funds younger than the period shown are extended backward with older funds, spliced on returns, not prices. | |
| Window | The worked example and the lab use the common window of the holdings, Jun 1996 to Sep 2026 (364 months), ending at the data month of this page. | |
| Equal weights on this page | 1/N for each fund, held fixed; the risk shares and volatility are computed at those target weights from the funds' monthly returns over their full common window. | weightingDatacontent-engine/lab/compute.ts |
| Risk share and volatility | Risk share = wᵢ(Σw)ᵢ / w′Σw and volatility = √(12 · w′Σw), with Σ the population covariance (divides by N) of monthly returns. | weightingStatspackages/portfolio/engine/risk.ts |
| Independent bets | The co-movement groups count: funds are clustered by average linkage on their correlations, and clusters merge into one group while their average cross correlation is at least 0.5; the count is the exponential of the entropy of the groups' risk shares. | exposureOverlappackages/portfolio/engine/overlap.ts |
| Equal-weight portfolios in the app | A template or custom portfolio with equal target weights is simulated month by month: holdings drift with their returns and are reset to the targets at each rebalance, every December for the templates. | isRebalanceBoundarypackages/portfolio/engine/portfolio.ts |
| Look-ahead in this page | Equal weights need no data, so they carry no look-ahead. The inverse-volatility and equal-risk weights shown beside them are estimated from the whole window and describe it with hindsight. | |
| Engine functions | The functions the portfolio app and the MarketHeist MCP tools run, called here on the same monthly returns. | inverseVolWeightspackages/portfolio/engine/risk.tsequalRiskContributionWeightspackages/portfolio/engine/risk.tshierarchicalRiskParitypackages/portfolio/engine/hrp.tshierarchicalEqualRiskContributionpackages/portfolio/engine/hrp.tsweightingStatspackages/portfolio/engine/risk.tseffectiveBetspackages/portfolio/engine/overlap.ts |
Equal weighting (1/N) compared with its neighbours
| Method | Inputs it needs | Uses expected returns | What it equalizes |
|---|---|---|---|
| Equal weight (1/N) | None | No | Money |
| Market-cap weight | Market values | No | Nothing; follows the market |
| Inverse volatility | Each holding's volatility | No | Stand-alone risk (weight × volatility) |
| Equal risk contribution (ERC) | Volatilities and correlations | No | Share of portfolio risk |
| Mean-variance optimization | Expected returns, volatilities, correlations | Yes | Nothing by design; maximizes expected return for the risk |
Limitations
- Risk follows the most volatile and most correlated holdings. With a mix of stock and bond funds, 1/N is in practice a stock portfolio with a small bond buffer.
- The result depends entirely on the list. Adding a third stock fund to a list of two stock funds and one bond fund changes the mix far more than any weighting rule would; with many similar holdings, equal weighting concentrates risk in what they share.
- Rebalancing back to equal weights sells recent winners and buys recent losers. That helps when holdings mean-revert and costs when one keeps outperforming, and every rebalance has trading costs and, in taxable accounts, taxes.
- The out-of-sample evidence for 1/N comes mostly from portfolios of stocks and stock portfolios, where the holdings are broadly similar. Across asset classes with very different volatility, the case is weaker, and later work disputes how general the result is.
- Long histories are extended backward with older funds or indexes where a fund is younger than the window, so early months approximate the fund rather than reproduce it.
Frequently asked questions
- Why is the 1/N portfolio hard to beat?
- Because optimized portfolios rely on estimates, especially of expected returns, that are very noisy, and optimizers amplify the errors. DeMiguel, Garlappi and Uppal (2009) found that this estimation error cancelled the theoretical advantage of optimizing in the data sets they studied. Equal weighting has no estimates to get wrong.
- Is an equal-weight portfolio diversified?
- In money, yes; in risk, only if the holdings have similar volatility and similar correlations. A calm holding that moves independently of the rest can carry almost none of the risk at the same weight as the others.
- How often is an equal-weight portfolio rebalanced?
- There is no fixed rule. MarketHeist's templates rebalance every December; quarterly and monthly are common alternatives, and the rebalancing guide compares them. More frequent rebalancing keeps the weights closer to equal at the cost of more trades.
- What is the difference between equal weight and equal risk contribution?
- Equal weight gives each holding the same money. Equal risk contribution gives each holding the same share of the portfolio's volatility, which means more money for calm holdings and for those that move against the rest.
- Is an equal-weight S&P 500 the same idea?
- It is the same rule applied to the stocks of an index instead of to asset classes: each company gets the same weight instead of a weight in proportion to its market value, which tilts the portfolio toward smaller companies.
References
- DeMiguel, V., Garlappi, L. and Uppal, R. (2009) Optimal Versus Naive Diversification: How Inefficient Is the 1/N Portfolio Strategy?. Review of Financial Studies 22(5).
- Kritzman, M., Page, S. and Turkington, D. (2010) In Defense of Optimization: The Fallacy of 1/N. Financial Analysts Journal 66(2). Argues that optimization with sensible inputs does beat 1/N.
- Benartzi, S. and Thaler, R. H. (2001) Naive Diversification Strategies in Defined Contribution Saving Plans. American Economic Review 91(1). How savers apply the 1/n heuristic to whatever funds they are offered.
- Faber, M. T. and Richardson, E. W. (2009) The Ivy Portfolio: How to Invest Like the Top Endowments and Avoid Bear Markets. Wiley. The equal-weight five-asset-class portfolio used in the example.
- Maillard, S., Roncalli, T. and Teiletche, J. (2010) The Properties of Equally Weighted Risk Contribution Portfolios. Journal of Portfolio Management 36(4). Places ERC between 1/N and minimum variance.
Try it on your own holdings
Open the Ivy Portfolio, five asset classes in equal parts →
Or ask an assistant connected to the MarketHeist MCP server (https://marketheist.io/mcp), which computes the same figures:
Using MarketHeist.io, analyze an equal-weight portfolio of VTI, VEU, IEF, DBC and VNQ: each fund's share of the money and share of the risk, and how equal-risk-contribution weights would differAsk an AI assistant
The prompt opens in a new tab, already filled in. An assistant without the MarketHeist connector answers from its own knowledge, not from the figures on this page.
Historical simulation, not advice. Computed by the MarketHeist portfolio engine from monthly total returns, data through Sep 2026. Markdown version of this page: equal-weight.md.