Risk Contribution: Decomposing Where Portfolio Risk Comes From

What is Risk Contribution?

Risk contribution measures how much of a portfolio's total risk is attributed to a particular asset, strategy, or exposure under a specified risk model. In a volatility-based framework, component risk contribution combines the position's current weight with its marginal effect on portfolio volatility.

For asset i:

1Risk contributioni = wi × MCRi

where wi is the portfolio weight and MCRi is the asset's Marginal Contribution to Risk.

Under the standard differentiable volatility decomposition, the individual component contributions add up to total portfolio volatility.

That property makes risk contribution useful for answering a question that capital weights alone cannot answer:

1Which positions are actually driving the portfolio's measured risk?

The math connects weights and covariance

Portfolio volatility is:

1σp = sqrt(w'Σw)

The marginal contribution from asset i is:

1MCRi = (Σw)i / σp

The component contribution is therefore:

1RCi = wi × (Σw)i / σp

For this volatility measure:

1Σ RCi = σp

This decomposition is sometimes associated with Euler's theorem for homogeneous risk measures.

The important practical point is that contribution depends on three things at once: the position weight, its own volatility, and its covariance with the rest of the portfolio.

Capital diversification can hide risk concentration

Consider a portfolio with equal capital in four assets:

1Asset A 25%
2Asset B 25%
3Asset C 25%
4Asset D 25%

That looks diversified by dollars.

Now suppose Asset A is much more volatile and highly correlated with B and C. It may account for far more than 25% of portfolio volatility.

The equal-weight presentation can therefore hide a concentrated risk structure.

This is why Risk Budgeting often monitors component contributions in addition to capital allocations.

A simple two-asset intuition

Suppose stocks are 60% of a portfolio and bonds are 40%.

If stock volatility is high and stock-bond correlation is modest, stocks may contribute the majority of total volatility.

A stylized decomposition might look like:

1Capital weights:
2Stocks 60%
3Bonds  40%
4
5Modeled volatility contributions:
6Stocks 88%
7Bonds  12%

Those percentages are only an example. The actual result depends on the covariance matrix.

The point is that a 60/40 capital split is not a 60/40 risk split.

Contribution can be expressed in risk units or percentages

Suppose portfolio volatility is 10% and an equity sleeve contributes 7 percentage points of that total volatility under the decomposition.

Its absolute contribution is 7% volatility.

Its percentage contribution is:

17% / 10% = 70% of total portfolio volatility

Both presentations can be useful.

The percentage view is intuitive for a Risk Parity portfolio because the manager can compare each sleeve's share of total risk.

The absolute view is useful when checking a total volatility target or risk limit.

Negative contributions are possible

A hedging asset can have a negative risk contribution under some volatility models.

If an asset's covariance with the rest of the portfolio is sufficiently negative, holding it may reduce total modeled volatility.

That does not mean the hedge is guaranteed to make money when other assets lose.

Correlations can change. The hedge can have basis risk, liquidity risk, financing costs, or nonlinear behavior that the covariance estimate does not capture.

A negative risk contribution is a model statement about the current portfolio, not a guarantee about every future scenario.

Risk contribution is model-dependent

There is no single risk contribution number attached permanently to an asset.

Change the portfolio and the number can change.

Change the covariance window and it can change again.

Change the risk measure from volatility to Expected Shortfall and the decomposition may tell a different story.

This dependence matters most when the portfolio contains nonlinear instruments or assets whose relationships behave differently in stress.

A normal-period volatility contribution can understate the importance of a short option position that produces small gains most days but large losses in rare events.

Risk contribution versus standalone risk

Standalone volatility asks how variable an asset's own returns are.

Risk contribution asks what the asset contributes inside this portfolio.

An asset with 20% standalone volatility could contribute little to a portfolio if its weight is small and its correlations are low.

Another asset with 8% standalone volatility could contribute heavily if the position is large and closely aligned with existing exposures.

This is why portfolio construction should not rank assets by standalone volatility and assume the ranking describes portfolio risk.

Risk contribution changes as markets move

Even without trading, risk contributions drift.

Prices change portfolio weights. Volatility changes. Correlations change.

Suppose a crisis causes equity volatility and cross-asset correlations to rise. The equity sleeve can consume more of the portfolio's risk budget even before its capital weight changes much.

This creates a potential trigger for Portfolio Rebalancing in portfolios governed by risk targets.

But mechanically trading every model change can create turnover exactly when liquidity is poor. Governance should specify how quickly risk-budget deviations require action.

Risk contribution can be decomposed by different dimensions

The same portfolio may be viewed by:

• security;
• asset class;
• sector;
• country;
• currency;
• manager;
• strategy;
• factor; or
• another risk bucket.

The decomposition should match the decision being managed.

An asset-class view may show a balanced portfolio while a factor view reveals that equities, high-yield bonds, and private credit all depend heavily on economic growth and financing conditions.

Grouping choices therefore matter.

Risk contribution and risk parity

Risk parity uses contribution explicitly.

A simple equal-risk-contribution portfolio adjusts weights so major sleeves contribute approximately the same amount of the selected risk measure.

For four sleeves, a target might be:

1Equities        25% of risk
2Government debt 25% of risk
3Credit          25% of risk
4Commodities     25% of risk

The capital weights required to achieve those percentages are usually unequal.

Low-volatility sleeves may need more capital, while high-volatility sleeves need less.

Equal risk contribution is one portfolio design. It is not a universal requirement for a well-diversified portfolio.

Risk contribution and expected return are separate

A position can consume a large share of risk and still be worth holding if its expected return justifies that risk.

Conversely, an exposure that contributes little risk may still be unattractive if its expected return is poor, its costs are high, or it creates an unwanted tail exposure.

Risk contribution describes allocation of risk. It does not determine the attractiveness of that risk.

That distinction is essential when using contribution numbers in an investment committee or optimizer.

A practical risk-contribution review

When a risk report says "equities contribute 70% of portfolio risk," ask:

1. What risk measure is being decomposed?
2. What time horizon and return frequency are used?
3. What covariance or scenario model produced the estimate?
4. Are derivatives represented linearly or with nonlinear risk?
5. Are contributions shown in absolute units or percentages?
6. Can contributions be negative?
7. How stable are the estimates across reasonable lookback windows?
8. What happens under a stress covariance matrix?
9. Does the grouping hide common risk factors?
10. Is the current contribution consistent with the portfolio's stated risk budget?

Those questions turn the decomposition into a decision tool rather than a decorative chart.

What risk contribution cannot tell you

Risk contribution does not forecast return, guarantee diversification, or identify the worst possible loss.

It is only as informative as the selected risk measure and the assumptions used to estimate it.

A precise decomposition of a poor model is still a poor description of economic risk.

Use risk contribution to understand the portfolio you have, then combine it with expected return, stress testing, liquidity, leverage, and investor objectives.

Grizzly Bulls' Models can provide context for systematic portfolio research, while the Macroeconomic Conditions Index can help frame changing market regimes. Neither route publishes a canonical live risk-contribution decomposition for a user's holdings.

Sources and further reading

• CFA Institute: Principles of Asset Allocation, 2026 curriculum
• CFA Institute: Measuring and Managing Market Risk, 2026 curriculum
• CFA Institute: Portfolio Risk and Return: Part I, 2026 curriculum

Explore more topics in the Financial Research Encyclopedia.