The Rebalancing Bonus: How a Volatile Diversifier Can Lift a Portfolio's Return

A diversifier can post an ugly return on its own and still raise the compound return of the portfolio it sits in. That sounds wrong, so this guide works through the math, the historical evidence, and the conditions that have to hold for it to be true. The idea has a name in the academic literature: the diversification return, often called the rebalancing bonus.²

The short version

High volatility helps a portfolio under specific conditions: the asset has an acceptable expected return, a low correlation to what you already own, reasonable costs, and you actually rebalance into and out of it. When those hold, the extra volatility is useful, because the rebalancing bonus grows as volatility rises and as correlation falls. When they do not hold, a volatile asset is just a volatile asset, and it drags the portfolio down.

This is a portfolio-construction idea, so judge each holding by what it does to the whole portfolio’s return and risk, not by how scary it looks in isolation. That is the core of Markowitz’s mean-variance framework: a portfolio’s risk depends on the covariances between holdings, not on the average of their individual volatilities.¹

The mechanism: volatility drag

Start with the gap between two ways of measuring return. The arithmetic return is the simple average of yearly returns. The compound (geometric) return is what your money actually grows at over time, and it is always lower when returns bounce around. The approximation that links them:

g \approx \mu - \tfrac{1}{2}\,\sigma^{2}

Compound return ( $g$ ) is roughly the arithmetic return ( $\mu$ ) minus half the variance ( $\sigma^{2}$ , the square of volatility). That subtracted term is volatility drag. A 20% volatility costs about 2.0 percentage points of compound return a year; a 50% volatility costs about 12.5 points. The real return of stocks guide covers this drag in more depth. It is also the reason a wild asset can look terrible on its own, and the raw material the rebalancing bonus works with.

The diversification return formula

Hold a set of assets at fixed weights, rebalance back to those weights, and the portfolio’s compound return beats the weighted average of the holdings’ compound returns. Start from the portfolio’s expected return and variance:

\mu_p = \sum_i w_i\,\mu_i \qquad\qquad \sigma_p^{2} = \mathbf{w}^{\top}\Sigma\,\mathbf{w}

where the $w_i$ are the weights, the $\mu_i$ the expected returns, and $\Sigma$ the covariance matrix of returns. The portfolio carries the same volatility drag at the portfolio level:

g_p \approx \mu_p - \tfrac{1}{2}\,\sigma_p^{2}

The diversification return is the gap between that portfolio compound return and the weighted average of the holdings’ own compound returns. It works out to half the difference between the weighted-average variance and the portfolio variance:

\mathrm{DR} \approx \tfrac{1}{2}\left(\sum_i w_i\,\sigma_i^{2} \;-\; \mathbf{w}^{\top}\Sigma\,\mathbf{w}\right)

For two assets, write out the portfolio variance in full,

\sigma_p^{2} = w^{2}\sigma_1^{2} + (1-w)^{2}\sigma_2^{2} + 2\,w(1-w)\,\rho\,\sigma_1\sigma_2,

substitute it into the diversification return, and the squared-weight terms collapse into a single product:

\begin{aligned}\mathrm{DR} &\approx \tfrac{1}{2}\big[\,w\,\sigma_1^{2} + (1-w)\,\sigma_2^{2} - \sigma_p^{2}\,\big] \\[4pt] &= \tfrac{1}{2}\,w(1-w)\big(\sigma_1^{2} + \sigma_2^{2} - 2\rho\,\sigma_1\sigma_2\big).\end{aligned}

Here $w$ is the weight of one asset and $\rho$ the correlation between the two. Three things make the bonus large: high volatility in the holdings, low (ideally negative) correlation, and weights that are meaningful but not lopsided. Push $\rho$ to $+1$ and the term in parentheses collapses, so the bonus goes to zero. Lower the correlation and it grows.

A clean example makes it concrete. Take two assets that each have a 5% arithmetic return and 20% volatility, held 50/50. Their standalone compound return is the same in every case. Only the correlation between them changes:

Correlation	Portfolio volatility	Portfolio compound return	Diversification return
+1.0	20.0%	3.0%	0.0%
0.0	14.1%	4.0%	1.0%
−0.5	10.0%	4.5%	1.5%

Same assets, same standalone returns. The only difference is correlation, and it adds a full percentage point of compound return at zero correlation and 1.5 points at −0.5.

The counterintuitive part: a bad-looking asset that helps

Now make the diversifier look genuinely unappealing on its own. Put a stock-like core next to a high-volatility diversifier:

Asset	Arithmetic return	Volatility	Standalone compound return
Stock core	8.0%	18.0%	6.38%
High-vol diversifier	7.5%	50.0%	−5.00%

The diversifier compounds at about −5% on its own, because its 50% volatility costs 12.5 points of drag against a 7.5% arithmetic return. On a fund fact sheet it would look like something to avoid. But put 10% of the portfolio in it, hold the other 90% in the core, assume zero correlation, and rebalance. The blended portfolio has a 7.95% arithmetic return and about 16.95% volatility, for a compound return of roughly 6.51%. That edges out the 6.38% from holding only the core, even though one of the two pieces compounds at −5% by itself.

\begin{aligned}g_d &\approx 7.5\% - \tfrac{1}{2}(50\%)^{2} = -5.0\% \\[4pt]\mu_p &= 0.9(8\%) + 0.1(7.5\%) = 7.95\% \\[4pt]\sigma_p &= \sqrt{0.9^{2}(18\%)^{2} + 0.1^{2}(50\%)^{2}} \approx 16.95\% \\[4pt]g_p &\approx 7.95\% - \tfrac{1}{2}(16.95\%)^{2} \approx 6.51\%\end{aligned}

The line item can look terrible alone and still pull its weight at portfolio scale. That is the whole intuition. The calculator below lets you pressure-test it with your own numbers: raise the diversifier’s volatility and watch its standalone return crater while the portfolio line can still rise, then push the correlation toward +1 and watch the benefit vanish.

Can a volatile diversifier still help?

Set a stock-like core and a single diversifier. Watch the diversifier’s standalone compound return fall as its volatility rises, while the rebalanced portfolio can still come out ahead when the correlation is low.

Core expected return8.0%

Arithmetic expected return of your stock sleeve.

Core volatility18.0%

Diversifier expected return7.5%

Arithmetic expected return, before its fee.

Diversifier volatility50%

Higher here is the whole point: it raises the bonus when correlation is low.

Correlation to the core0.00

The main lever. The bonus grows as correlation falls.

Diversifier weight10%

How much of the portfolio you rebalance into the diversifier.

Diversifier annual cost0.00%

Fees and frictions come straight off the diversifier's return.

Diversifier alone

-5.0%

Standalone compound return, after its cost.

Portfolio compound return

6.51%

At 17.0% portfolio volatility.

Change vs 100% core

+0.13%

Compound return vs holding only the core (6.38%).

Diversification return

1.27%

The rebalancing bonus from holding both instead of their blend.

Portfolio compound return as you add the diversifier

Educational sensitivity tool, not a forecast. Inputs are arithmetic expected returns; the compound figures use the standard g = μ − ½σ² approximation. Correlations and expected returns are unstable, and small changes in your assumptions can flip the result. A diversifier with a poor expected return, high cost, or positive correlation to stocks will not be rescued by volatility.

Where the bonus actually comes from: rebalancing

The diversification return is not a free reward for owning volatile things. It is produced by the act of rebalancing. Eric Willenbrock showed that the source of the return is the discipline of selling assets that have risen above their target weight and buying assets that have fallen below it, which is a systematic way of selling relative strength and buying relative weakness.³ Booth and Fama formalized the same effect and named it the diversification return.²

William Bernstein’s classic essay on the rebalancing bonus adds the caveat that matters most. The bonus is largest for volatile, low-correlation assets with similar long-run returns. When two assets have very different expected returns, such as stocks and bonds, rebalancing tends to lower the compound return versus simply holding the stronger asset, because you keep trimming the winner. You accept that lower return in exchange for lower risk.⁴ That is why this argument works best when the diversifier is expected to earn a real return over time, rather than being a permanent loser. Our guide on rebalancing frequency covers a related trap: a portfolio that drifts to mostly stocks can post a higher return, but that comes from taking more risk, not from a rebalancing bonus.

Capital efficiency: why high volatility can be a feature

Cliff Asness makes a sharper version of the point for alternatives. The logic runs through a simple identity: a strategy’s expected excess return is about its Sharpe ratio times its volatility.

\mathbb{E}[r] - r_f = \mathrm{SR} \times \sigma

If a diversifying strategy keeps the same Sharpe ratio whether it is run at 10% volatility or 25% volatility, the higher-volatility version delivers more diversifying return per dollar you allocate to it. For an investor with a fixed amount of portfolio capital, that capital efficiency matters.⁵

AQR’s stylized example holds the total portfolio volatility fixed at 10% and lets an optimizer choose the mix of stocks, bonds, and an uncorrelated alternative:

Alternative used	Stocks / Bonds / Alts	Expected compound return
None	58 / 42 / 0	8.9%
10%-volatility alternative	62 / 0 / 38	9.3%
25%-volatility alternative	48 / 28 / 24	9.8%

Stylized illustration from AQR, “In Praise of High-Volatility Alternatives.” All three portfolios target the same 10% total volatility.⁵

Read the rows as the same risk budget bought three ways. The higher-volatility alternative earns the most while taking the smallest dollar allocation, 24% versus 38%, because each dollar carries more diversifying punch. This is the same idea behind return stacking and portable alpha: use capital efficiency to fit a diversifier in without selling all your stocks to make room.

The conditions, stated plainly

Volatility is only useful when the rest of the picture cooperates. A high-volatility holding earns its place only when all of these hold:

Acceptable expected return. The drag is real. An asset with a poor arithmetic return will compound badly and stay a drag at portfolio scale.
Low correlation to what you own. A volatile asset that moves with your stocks adds risk and no bonus.
Reasonable cost and liquidity. Fees, financing, and trading frictions come straight off the top, and an illiquid sleeve cannot be rebalanced.
A rebalancing plan you will follow. The bonus is paid for by rebalancing. Buy and hold a volatile diversifier and you keep the drag without the discipline that offsets it.

Asness is blunt that the behavioral condition is the binding one. He notes that many investors, perhaps most, cannot sit through the stretches when a high-volatility diversifier looks foolish, and so they should not own one.⁵ A diversifier you abandon at the bottom is worse than no diversifier at all.

What the evidence says about real diversifiers

Managed futures and trend following

Trend following is the most studied liquid diversifier that fits this mold: volatile, historically low correlation to stocks and bonds, with a real return signal. Moskowitz, Ooi, and Pedersen documented time-series momentum across 58 liquid futures and forward contracts spanning equity indexes, bonds, currencies, and commodities, with returns that were historically strong in extreme up and down markets.⁶ Hurst, Ooi, and Pedersen extended trend-following evidence back to 1880 and found positive average returns in every decade and low correlation to traditional assets.⁷

The caveat is that trend can disappoint for years. Morningstar reported that the typical systematic trend fund lost about 2.3% annualized over the three years through August 31, 2025, even though some higher-volatility managed-futures funds shone in 2022 when stocks and bonds fell together.¹¹ Our managed-futures guide and the guide to AQR funds go deeper on specific vehicles, leverage, and the behavioral problem.

Commodity futures

Commodities show how much the rebalancing framing matters. Gorton and Rouwenhorst found that an equally weighted, fully collateralized basket of commodity futures from July 1959 to December 2004 earned returns and a Sharpe ratio comparable to U.S. equities, with negative correlation to stocks and bonds and positive correlation to inflation.⁸ Erb and Harvey added the nuance that individual commodity futures had average excess returns near zero, yet a rebalanced, diversified basket could earn an equity-like return, much of it from rebalancing and term structure rather than from the spot price.⁹ A decade-later review by Bhardwaj, Gorton, and Rouwenhorst found the original conclusions largely held out of sample.¹⁰ Commodities are close to the textbook case: low-return parts, combined and rebalanced, producing a useful whole.

Why this matters for DIY investors

A simple stock and bond portfolio is still enough for most people. Vanguard’s own diversification guidance makes the standard point that assets which do not move in sync reduce portfolio volatility.¹² The high-volatility diversifier idea matters most for investors who are already equity-heavy and want a portfolio that may hold up better across inflation shocks, equity bear markets, and the occasional stretch when stocks and bonds fall together.

It tends to make sense for investors who:

Already hold a diversified, low-cost core.
Are equity-heavy and want a genuinely independent return stream.
Can tolerate years of a holding looking wrong without selling.
Will rebalance on a schedule rather than by conviction.
Size the sleeve small enough to survive a bad decade.

It tends not to make sense for investors who:

Want the simplest possible portfolio.
Are still building an emergency fund or basic retirement savings.
Cannot stomach a line-item loss without abandoning the plan.
Hold everything in a taxable account with little tax awareness.
Are buying because a recent return looked good.

Taxes deserve a flag. Futures-based strategies often involve Section 1256 contracts, where gains and losses are generally treated as 60% long-term and 40% short-term and marked to market at year-end, regardless of holding period.¹³ Product structures vary, and high-turnover diversifiers usually belong in a tax-advantaged account. Understand the tax reporting before you buy.

What I would actually do

I would not tell a DIY investor to load up on high-volatility alternatives. The defensible version is narrower: a modest allocation to liquid, genuinely diversifying strategies can make sense for some long-term investors, especially those who are equity-heavy, comfortable with tracking error, and disciplined about rebalancing. Managed futures and commodity futures are more defensible candidates than opaque private alternatives because they are liquid, marked to market, and supported by a clearer economic rationale.

A reasonable range to model, not a recommendation, is 0% to 15%. For many readers, 0% is the right answer and a cheaper bond or TIPS allocation does the diversifying job well enough. For curious, advanced DIY investors, something in the 5% to 10% range is enough to see whether the diversifier improves portfolio behavior without letting the tail wag the dog. Size it so a quiet five-year stretch would not make you sell. Bonds remain the first diversifier to reach for; alternatives are a step beyond, for a specific job.

Frequently asked questions

What is the diversification return or rebalancing bonus?

It is the amount by which a rebalanced portfolio’s compound return exceeds the weighted average of its holdings’ compound returns. It grows with the volatility of the holdings and shrinks as their correlation rises, and it is produced by rebalancing back to target weights.²

Can an asset that loses money on its own improve my portfolio?

An asset with a negative compound return but a positive arithmetic return can still raise a portfolio’s compound return when its correlation is low and you rebalance, as the 90/10 example shows. An asset with a genuinely negative expected (arithmetic) return, high costs, or positive correlation to your stocks will not.

Does the rebalancing bonus require rebalancing?

Yes. The return comes from systematically trimming holdings that have run up and adding to those that have lagged. Buy and hold a volatile diversifier and you keep the drag without the discipline that pays for it.³

Is the rebalancing bonus free money?

No. It requires volatile, low-correlation assets with acceptable expected returns, reasonable costs, and the discipline to rebalance through long stretches of underperformance. Miss any of those and the volatility just costs you.

How much should I allocate to high-volatility diversifiers?

For many investors, zero. For those who want the exposure and can hold it, a modeling range of roughly 5% to 10% is enough to matter without dominating the portfolio. Size it so a bad multi-year run would not push you to sell.

Key takeaways

Judge a holding by its effect on the whole portfolio, not by how it looks alone. Volatility drag makes a standalone number misleading.
The diversification return grows with volatility and falls with correlation. A volatile, low-correlation diversifier can raise a portfolio’s compound return even with a poor standalone return.
The bonus is paid for by rebalancing, and it is cleanest for volatile, low-correlation assets with similar long-run returns. Very different expected returns shift the trade toward risk reduction instead.
High volatility can be capital-efficient, delivering more diversifying return per dollar, which is why a small alternatives sleeve can punch above its weight.
The conditions are strict. Acceptable return, low correlation, reasonable cost, and the discipline to hold and rebalance. A volatile asset that fails any of these is just risk.
For most DIY investors, 0% is fine. For those who want it, 5% to 10% in a liquid, marked-to-market diversifier, held in a tax-advantaged account, is plenty.

Related guides

How Often Should You Rebalance? covers the discipline that pays the diversification return, and the drift trap that masquerades as a rebalancing bonus.
Do You Need Managed Futures? is the full evidence base for the most-studied high-volatility diversifier.
Do You Need Return Stacking? shows how capital efficiency lets you add a diversifier without selling your stocks to fund it.
Modern Portfolio Theory for Real Life is the covariance framework behind why correlation, not standalone risk, drives portfolio behavior.
The Real Return of Stocks explains the volatility drag that the rebalancing bonus works against.
Are Bonds Still Good Diversifiers? covers the first diversifier most investors should reach for.

Sources

Markowitz, H. (1952). Portfolio Selection. Journal of Finance, 7(1), 77-91.
Booth, D.G., & Fama, E.F. (1992). Diversification Returns and Asset Contributions. Financial Analysts Journal, 48(3), 26-32. A rebalanced portfolio’s compound return exceeds the weighted average of its components’ compound returns.
Willenbrock, S. (2011). Diversification Return, Portfolio Rebalancing, and the Commodity Return Puzzle. Financial Analysts Journal, 67(4), 42-49. The source of the diversification return is rebalancing.
Bernstein, W. (1996). The Rebalancing Bonus: Theory and Practice. Efficient Frontier. High-volatility, low-correlation, similar-return assets create the largest bonus; for very different returns, rebalancing lowers return but reduces risk.
Asness, C. (AQR). In Praise of High-Volatility Alternatives. Capital efficiency of higher-volatility diversifiers, the stylized optimizer example at a fixed 10% portfolio volatility, and the behavioral condition.
Moskowitz, T., Ooi, Y.H., & Pedersen, L.H. (2012). Time Series Momentum. Journal of Financial Economics, 104(2), 228-250. 58 liquid futures and forward contracts.
Hurst, B., Ooi, Y.H., & Pedersen, L.H. (2017). A Century of Evidence on Trend-Following Investing. Journal of Portfolio Management, 44(1), 15-29. Positive returns in every decade since 1880 with low correlation to traditional assets.
Gorton, G., & Rouwenhorst, K.G. (2006). Facts and Fantasies about Commodity Futures. Financial Analysts Journal, 62(2), 47-68. Equity-like returns and Sharpe, negative correlation to stocks and bonds, positive correlation to inflation (Jul 1959-Dec 2004).
Erb, C.B., & Harvey, C.R. (2006). The Strategic and Tactical Value of Commodity Futures. Financial Analysts Journal, 62(2), 69-97. Individual futures averaged near-zero excess returns; a rebalanced basket can be equity-like.
Bhardwaj, G., Gorton, G., & Rouwenhorst, K.G. (2015). Facts and Fantasies about Commodity Futures Ten Years Later. NBER Working Paper 21243. Original conclusions largely held out of sample.
Morningstar. Managed-Futures Funds Look to Rebound. Can They Help Diversify Your Portfolio? The typical systematic trend fund lost about 2.3% annualized over the three years through August 31, 2025.
Vanguard. Portfolio diversification: What it is and how it works. Assets that do not move in sync reduce portfolio volatility.
Legal Information Institute. 26 U.S. Code § 1256. Section 1256 contracts are marked to market and taxed 60% long-term / 40% short-term.