Module 3
SPAN & SPAN 2 — the futures standard
SPAN is the margin standard for futures and listed options, used by CME and most clearing houses. Classic SPAN builds a risk array per contract: 16 scenarios combining price shocks at fractions of the scan range (±⅓, ±⅔, ±1) with vol shocks up and down, plus two extreme-move scenarios (about 2× the range) covered at a fraction — typically 35% — to capture short-option tail risk. The worst array value across scenarios is the scanning risk.
Then come the charges for what scanning misses:
- Intra/inter-month spread charge — the scan assumes parallel curve moves, so calendar spreads look riskless. They aren't. Charge for it.
- Inter-commodity spread credit — offsets between correlated products (crude vs heating oil; increasingly crypto vs equity index).
- Delivery / spot month charge — heightened risk approaching expiry and physical delivery.
- Short option minimum — a floor for deep-OTM short options that scan to near zero but carry unbounded tail risk. Every risk manager has a war story about this one.
- Net option value — long option premium is an asset, subtracted at the end.
SPAN 2replaces the array with a proper VaR framework — filtered historical simulation, an expected-shortfall-style base requirement, and explicit liquidity and concentration add-ons as first-class components rather than bolted-on charges. The migration is underway because arrays don't handle multi-asset portfolios or liquidity cost well. If the VaR/ES machinery is unfamiliar, the VaR module covers it.
🎛 Risk array explorer
Short options, hedge with futures, and watch the 16-scenario risk array
Risk array — charged loss per scenario
Orange bars are the two extreme-move scenarios (2× the scan range) charged at 35% — they exist to catch the short-option tail that the ±1 scan misses.
Scanning risk
$334
Short option minimum
$400
Requirement
$400
Classic SPAN: price shocks at ±⅓, ±⅔ and ±1 of the scan range crossed with vol up/down, plus two extreme scenarios, worst charged value wins. Now hedge the short calls with one long future — scanning risk collapses, and the short option minimum takes over as the binding floor. That floor is there because deep-OTM shorts scan to nearly zero but carry unbounded tail risk. Educational tool — not investment advice.
Things to try
- • Short 2 calls, no futures: the worst scenario is price up, vol up — and the extreme +2 scenario (orange) is closing in on it.
- • Add 1 long future as a delta hedge: scanning risk drops sharply, and the short option minimum becomes the binding number. That's the floor doing its job.
- • Crank the scan range to 20% (a crypto future, not a bond future) and watch the array explode — same book, different market.
Test yourself
Why does SPAN charge calendar spreads at all, when a long June / short September future has almost no outright price exposure? (Hint: the scan shocks the whole curve in parallel — what risk does that assumption erase?)