Level III Core · Overview of Asset Allocation

Risk Budgeting and Asset Allocation Alternatives for CFA Level III

Updated 9 October 2026 · Fact-checked

Risk budgeting sets how much total risk a portfolio may take and divides it among asset classes, factors or managers by their risk contribution, not by capital. Alternatives include factor-based allocation, 1/N, risk parity and Monte Carlo simulation. You solve questions by matching the method to the client's goals and constraints.

Understand Risk Budgeting and Asset Allocation Alternatives

Traditional allocation starts with capital: put 60% in equities, 40% in bonds. But 60% of capital in equities can mean the large majority of portfolio risk (for example, 97% in the worked example), because equities are far more volatile. Risk budgeting fixes this by starting with risk. You decide how much total risk the investor can bear, then split that risk among asset classes, factors or managers where it is expected to earn the best reward.

The risk budget can be set in absolute terms (for example, portfolio volatility of 8%) or relative terms (tracking error against a benchmark of 2%). Each component uses risk budget according to its contribution to risk, which is its weight × its marginal contribution to risk. A good budget compares each component's expected excess return per unit of risk. Capital is moved until the return per unit of risk is similar across components, given the constraints.

Factor-based allocation treats asset classes as bundles of underlying risk factors, such as equity market risk, interest-rate (duration) risk, credit spread, inflation, liquidity, value, size, momentum and carry. Two asset classes with different labels may carry the same factor exposure, so apparent diversification can be poor, especially in a crisis. You allocate to factors you want paid for, and the factor view can be combined with a risk budget. Its limits: factors need to be well defined, can be hard to implement and can behave differently from expectations.

Other approaches are simpler or different in kind. 1/N gives each of N assets an equal weight. It needs no forecasts and so avoids estimation error, but it ignores risk and the client's needs. Risk parity sets weights so each asset contributes equally to portfolio risk. It does not need return forecasts, but it usually overweights low-volatility assets such as bonds, so it often needs leverage to reach a return target. It depends on volatility and correlation estimates, which can shift. 60/40 is a simple capital-based rule. Monte Carlo simulation generates many random paths of returns for the assets, using assumed distributions, correlations and cash flows. It shows the probability of meeting goals, shortfall and the effect of spending, contributions and rebalancing over time. It is flexible and can handle path dependency, but its answers are only as good as its inputs (garbage in, garbage out).

No method is best. Mean-variance optimization uses return forecasts and is sensitive to them. Risk budgeting, risk parity and 1/N lean less on return forecasts. Monte Carlo is used to test a chosen allocation. On the exam, always tie your choice back to the client's objectives, constraints, and the quality of the inputs.

Key rules to remember

Marginal contribution to risk (MCTR): MCTR_i = (Cov(R_i, R_p)) ÷ σ_p = β_i,p × σ_p, where β_i,p = Cov(R_i, R_p) ÷ σ_p²; Change in portfolio risk from a small increase in asset i. β_i,p is asset i's beta to the portfolio, defined as Cov(R_i, R_p) ÷ σ_p².
Absolute contribution to risk: CTR_i = w_i × MCTR_i = w_i × Cov(R_i, R_p) ÷ σ_p; The CTR values of all assets add up to portfolio standard deviation σ_p.
Percentage contribution to risk: %CTR_i = CTR_i ÷ σ_p = w_i × Cov(R_i, R_p) ÷ σ_p²; These sum to 100%. A risk budget is stated in this form.
Risk-budget optimality (reward per unit of risk): (E(R_i) − R_F) ÷ MCTR_i is equal for all assets i, which is equivalent to (E(R_i) − R_F) ÷ β_i,p being equal across assets; At the optimum, no shift of risk between assets improves return per unit of risk. The two forms are equivalent because MCTR_i = β_i,p × σ_p and σ_p is the same for every asset. Unconstrained case.
Risk parity condition: w_i × MCTR_i = w_j × MCTR_j for all i, j; Equal risk contribution. If assets are uncorrelated, weights are proportional to 1/σ_i.
1/N weight: w_i = 1 ÷ N; Equal capital weight for each of N assets.
Sharpe ratio: SR = (E(R_p) − R_F) ÷ σ_p; Used to compare risk-budgeted portfolios.

How to solve Risk Budgeting and Asset Allocation Alternatives questions

Use this order for most questions on risk budgeting and allocation approaches. It works for calculation and for recommend-and-justify questions.

1Read the client's objective and constraints first: return target, risk limit, leverage, liquidity, horizon and any rules on shorting.
2Identify the command word. Calculate, Identify, Explain and Recommend each need a different answer length.
3If the question asks for a risk contribution, compute the portfolio standard deviation, then the weight times the marginal contribution for each asset, then divide by portfolio risk.
4Check that percentage contributions add up to 100%, and that absolute contributions add up to portfolio standard deviation.
5If the question asks about an approach, state its input needs (return forecasts, volatilities, correlations, none), then its main strength and weakness.
6Match the approach to the client: low forecast confidence suggests 1/N or risk parity; a goal-probability question suggests Monte Carlo; hidden concentration suggests factor analysis.
7Write the answer in the fewest words that earn the point: the approach, one reason from the vignette, and one limitation if asked.

Quickest way: Fast risk-contribution check and approach match

When to use it: Use when time is short and the question asks for risk contributions or which approach fits the client.

For two assets, find portfolio variance: w1²σ1² + w2²σ2² + 2w1w2ρσ1σ2.
Find each CTR directly as w_i × (w_i σ_i² + w_j ρ σ_i σ_j) ÷ σ_p. Check they sum to σ_p.
For risk parity with uncorrelated assets, set weights proportional to 1/σ and scale to 100%.
For approach questions, remember a one-line tag: 1/N = no inputs; risk parity = no return forecasts, may need leverage; factor = look through labels; Monte Carlo = probability of goals and path dependence.

Common mistakes in Risk Budgeting and Asset Allocation Alternatives

Treating risk budgeting as a capital allocation
Candidates assume 60% capital in an asset means 60% of risk.
Fix: Compute risk contribution with weight × marginal contribution. A volatile asset's share of risk is usually larger than its share of capital.
Saying risk parity equals equal weights
Both sound like equal treatment, so they get mixed up.
Fix: 1/N equalizes capital. Risk parity equalizes risk contribution, so low-volatility assets get larger weights.
Claiming risk parity needs no assumptions
It does not need return forecasts, so it seems assumption-free.
Fix: It still needs volatility and correlation estimates, and it often uses leverage, which brings its own costs and risks.
Believing Monte Carlo predicts the future accurately
Thousands of simulated paths look rigorous.
Fix: Say it estimates outcome distributions under assumed inputs. Its results are only as reliable as the assumptions on returns, correlations and distributions.
Treating asset-class diversification as factor diversification
Different labels look like different risks.
Fix: Look at the underlying factors. Assets with a common factor, such as equity risk, can fall together in a downturn.
Forgetting to tie the answer to the client
Candidates describe the method but not why it fits.
Fix: Add one phrase from the vignette, such as a leverage ban, a spending need or low confidence in forecasts.

Worked examples

Example 1

A portfolio holds 60% in Asset E (σ = 20%) and 40% in Asset B (σ = 5%). The correlation is 0. Calculate portfolio standard deviation and the percentage contribution to risk of each asset.

Show the solution

Variance = (0.6 × 0.20)² + (0.4 × 0.05)² = 0.0144 + 0.0004 = 0.0148.
σ_p = √0.0148 = 12.17%.
With zero correlation, each asset's contribution to variance is its own term: E = 0.0144, B = 0.0004.
%CTR_E = 0.0144 ÷ 0.0148 = 97.3%.
%CTR_B = 0.0004 ÷ 0.0148 = 2.7%. Total is 100%.

Answer: Portfolio standard deviation is 12.17%. Asset E contributes 97.3% of risk and Asset B 2.7%, despite a 60/40 capital split.

Example 2

Using the same assets (σ_E = 20%, σ_B = 5%, correlation 0), find risk parity weights. Expected returns are 10% for Asset E and 4% for Asset B. Then explain why a client with a 10% return target and a ban on leverage may be unable to use them.

Show the solution

With zero correlation, risk parity weights are proportional to 1/σ: E = 1/0.20 = 5; B = 1/0.05 = 20.
Total = 25. w_E = 5 ÷ 25 = 20%; w_B = 20 ÷ 25 = 80%.
Check: variance = (0.2 × 0.20)² + (0.8 × 0.05)² = 0.0016 + 0.0016 = 0.0032. Each asset contributes 50%.
Portfolio risk is √0.0032 = 5.66%, with 80% in the lower-return asset.
Expected return = 0.2 × 10% + 0.8 × 4% = 2.0% + 3.2% = 5.2%.
5.2% is below the 10% target. Reaching the target would need leverage, which this client bans.

Answer: Risk parity weights are 20% Asset E and 80% Asset B, each contributing 50% of risk. The expected return is 5.2%, below the 10% target. Closing the gap would need leverage, which this client bans, so risk parity is hard to apply unmodified.

Exam tips

For risk contribution calculations, show the variance, the portfolio standard deviation and each contribution so a slip still earns method credit.
Command words matter. For Compare, address both approaches; for Recommend, give one choice and a reason tied to the vignette.
Memorize one strength and one weakness for each of 1/N, risk parity, factor-based, risk budgeting and Monte Carlo. Item sets often test these.
Check that percentage contributions sum to 100% before moving on.
When the vignette mentions low confidence in return forecasts, think 1/N or risk parity; when it asks for the probability of meeting a goal, think Monte Carlo.

Risk Budgeting and Asset Allocation Alternatives in other exams

The same ground in other exams, if you are preparing for more than one or want another angle on it.

Risk Budgeting and Asset Allocation Alternatives: frequently asked questions

What is risk budgeting in CFA Level III?

It is a process that sets the total risk an investor can take and divides it among assets, factors or managers. The split is based on risk contribution, not capital. The aim is to put risk where it is expected to earn the best reward.

How is risk parity different from mean-variance optimization?

Mean-variance optimization needs expected return, volatility and correlation inputs and can be very sensitive to the return forecasts. Risk parity needs only volatilities and correlations and equalizes risk contributions. It tends to overweight low-volatility assets and often needs leverage.

Why use factor-based asset allocation?

Asset classes can share the same underlying risk factors, so diversification by label can be misleading. Factor-based allocation looks through to those exposures, which helps you see concentration. Its limits are definition, implementation and unstable factor behavior.

How does Monte Carlo simulation help in asset allocation?

It simulates many possible return paths using assumed distributions and correlations, along with cash flows such as spending and contributions. You can then estimate the probability of meeting goals and the size of shortfalls. The results depend entirely on the quality of the inputs.