How does risk budgeting work using marginal VaR and component VaR?
I'm studying risk budgeting for FRM Part II and I'm confused about the relationship between marginal VaR, component VaR, and incremental VaR. How do portfolio managers use these decompositions to allocate risk capital across desks or asset classes?
Risk budgeting decomposes total portfolio risk into contributions from each position, enabling managers to allocate risk capital efficiently. The key tools are marginal VaR, component VaR, and incremental VaR.
Definitions:
| Measure | Definition | Use |
|---|---|---|
| Marginal VaR | Sensitivity of portfolio VaR to a small increase in position i: dVaR/dw_i | Optimizing position sizing |
| Component VaR | Position i's contribution to total VaR: w_i x Marginal_VaR_i | Decomposing portfolio risk |
| Incremental VaR | Change in VaR from adding/removing position i entirely | Add/drop decisions |
Critical Property:
Sum of all Component VaRs = Total Portfolio VaR
This is Euler's theorem in action — the components perfectly decompose total risk.
Marginal VaR Formula:
MVaR_i = z_alpha x (Sigma x w)_i / sigma_p
This is the contribution of asset i to portfolio volatility, scaled by the confidence z-score.
Worked Example — Silveridge Asset Management:
Silveridge has a $50M portfolio:
- US Equities: 50% ($25M), vol = 18%
- Intl Equities: 30% ($15M), vol = 22%
- Fixed Income: 20% ($10M), vol = 5%
Correlations: US/Intl = 0.65, US/FI = -0.15, Intl/FI = 0.05
Step 1: Portfolio Volatility
After the matrix calculation: sigma_p = 12.84%
Step 2: Marginal VaR (per unit)
MVaR_US = z x [w_US x sigma_US^2 + w_Intl x rho_{US,Intl} x sigma_US x sigma_Intl + w_FI x rho_{US,FI} x sigma_US x sigma_FI] / sigma_p
For 99% confidence (z = 2.326):
- MVaR_US = 2.326 x 0.1578 = 0.367
- MVaR_Intl = 2.326 x 0.1712 = 0.398
- MVaR_FI = 2.326 x (-0.0006) = -0.001
Step 3: Component VaR ($)
- CVaR_US = 0.50 x 0.367 x $50M = $9.175M
- CVaR_Intl = 0.30 x 0.398 x $50M = $5.970M
- CVaR_FI = 0.20 x (-0.001) x $50M = -$0.010M
Total VaR = $9.175M + $5.970M - $0.010M = $15.135M
Insight: Fixed income has negative component VaR — it's actually reducing total portfolio risk. This means adding more fixed income would decrease VaR.
Risk Budgeting in Practice:
Silveridge allocates risk budgets to each desk:
- US Equity desk: budget = $10M VaR
- Intl Equity desk: budget = $6M VaR
- Fixed Income desk: budget = $1M VaR (hedging benefit counted separately)
If the US desk's component VaR exceeds $10M, they must reduce positions. If below, they have capacity to take more risk.
Optimal Portfolio: In a risk-budgeting framework, the optimal portfolio equalizes marginal VaR per unit of expected return across all positions. If MVaR_i / E(R_i) is higher for one asset, you're taking too much risk per unit of return there.
For more risk management frameworks, explore our FRM Part II course.
Master Part II with our FRM Course
64 lessons · 120+ hours· Expert instruction
Related Questions
Why is DV01 so much smaller than dollar duration if both are supposed to measure rate risk?
When should I stop using modified duration and switch to effective duration?
How should I think about the relationship between Macaulay duration and modified duration instead of memorizing two separate definitions?
Why do hedge calculations often use dollar duration or DV01 instead of just modified duration?
When should I prefer historical simulation VaR over delta-normal VaR?
Join the Discussion
Ask questions and get expert answers.