Cross-Product Margin Stacks: Capital Efficiency in Multi-Asset Books

Multi-asset trading books should consume less margin than the sum of their components. The reason — risk offsets across correlated positions — is well understood in theory but inconsistently applied in practice. A quantitative framework for evaluating cross-product margin offsets, with worked scenarios across FX, metals, indices, and energy.
Part I — How Cross-Product Margin Works
1. The Risk Offset Logic
- Long AUD/USD and long XAU/USD are positively correlated. Holding both nets risk exposure lower than either standalone.
- Long EUR/USD and short USD/CHF behave like a single EUR-CHF view; margin should reflect the netted exposure.
- Properly modelled cross-product margin can reduce capital usage by 30–60% on diversified multi-asset books.
2. SPAN-Style Methodologies
- Most institutional venues use SPAN or VaR-based margin models that scenario-test combined positions.
- The model evaluates portfolio P&L across stress scenarios and sizes margin to the worst-case loss.
- Quality of offset depends on the scenario set and correlation assumptions baked into the model.
Takeaway: Margin offsets are model-dependent. Two counterparties with "cross-product margin" can deliver materially different capital efficiency.
Part II — Worked Margin Scenarios
3. Diversified G10 + Gold Book
Sample portfolio: long AUD/USD (USD 5M notional), long EUR/USD (USD 5M), short USD/JPY (USD 5M), long XAU/USD (USD 2M). Standalone margin: ~USD 850K. Properly netted margin: ~USD 380K (55% saving).
4. Energy + Metals Cross-Hedge
Sample portfolio: long WTI (USD 3M), long XAU/USD (USD 2M), short SPX500 futures (USD 5M). Standalone: ~USD 720K. Netted: ~USD 540K (25% saving). Lower offset because energy-equity correlation is regime-dependent.
Putting It All Together
Cross-product margin is real economic value, but only if the underlying model captures actual portfolio risk. Counterparties differ widely in modelling sophistication. Diligence should focus on the methodology — scenario set, correlation model, stress regime treatment — not on the marketing claim of "cross-asset margin support".
Sources
- CME Group SPAN methodology documentation, public version.
- BIS Basel III capital requirements for FX and commodities.
- Orbis Securities cross-asset margin scenario analysis, 2025–26.