Calendar Spread Backtest: 48 Cycles, Term-Structure Timing
Long ATM calendars — sell the front month, buy the back. Win rate by IV rank, the contango-vs-backwardation regime split, and why the long calendar is one of the few short-premium-adjacent trades that actually likes rising IV.
Simulated data for display. Illustrative narrative — not a verified live backtest. Build real backtests on the strategy builder.
All strategy backtests →The Test Setup
Underlying: SPY and a few large-cap names with liquid monthly chains.
Structure: long ATM calendar — sell the front-month option, buy the back-month at the same strike (net debit). Net long vega, net positive theta.
Entry: front month ~30 DTE, back month ~60 DTE, strike at the money. Two entry filters compared: IV rank, and the sign of the term-structure slope (front IV vs back IV).
Exit: close at front-month expiration, or earlier at +25% / -25% of debit.
Period: January 2021 – December 2024, 48 monthly cycles.
Return by Entry IV Rank
| IV Rank at Entry | Win Rate | Avg Return on Debit | Note |
|---|---|---|---|
| Below 30 ★ | 58% | +14% | Front IV has room to rise |
| 30–50 | 51% | +6% | Middling edge |
| Above 50 | 42% | -9% | IV more likely to fall, draining the long leg |
The calendar is long vega, so it wants to be entered when IV is low and has room to expand — the mirror image of a premium-selling strategy. Entered into already-high IV, the more likely path is contraction, which hurts the back leg you own. Check IV rank on the IV calculator before entry.
Term Structure: Contango vs Backwardation
| Term-Structure Slope at Entry | Win Rate | Avg Return on Debit |
|---|---|---|
| Contango (back IV > front IV) ★ | 61% | +18% |
| Backwardation (front IV > back IV) | 39% | -15% |
A positively-sloped (contango) curve means you sell the cheaper near-month vega and own the richer far-month vega — the structure pays you. Backwardation (common right before earnings or in a panic) inverts that: the long leg you own bleeds IV faster than the short leg you sold as the curve normalizes. Combining the two filters — low IV rank and contango — produced the best bucket in the illustrative test.
Anatomy of the Worst Trade
A SPY ATM calendar opened for a $1.85 debit during a quiet stretch. Three weeks later a sharp macro headline pushed SPY ~6% away from the strike before front-month expiration. The tent is widest at the strike and worthless far from it, so the position lost roughly -$0.40 against the $1.85 debit, near the -25% stop, despite IV actually rising.
The lesson: a calendar is direction-neutral only inside a band around the strike. A large move — not a volatility move — is what kills it. Sizing small and respecting the stop kept this from being worse.
Five Takeaways
- Enter when IV rank is low. The long calendar is long vega — it wants room for IV to expand, the opposite of selling premium.
- Demand a contango term structure. Back-month IV above front-month was the difference between +18% and -15% in the illustrative test.
- Pin risk is the goal, big moves are the enemy. The tent only pays near the strike. A 6% move kills it regardless of vega.
- Avoid backwardation and pre-earnings entries. Inverted curves drain the long leg as they normalize.
- Use a hard stop. -25% of debit caps the damage from the directional moves that occasionally blow through the tent.
Price the calendar before you open it
Model front/back-month pricing and the theta edge, then check the term structure on the IV tools.
Related Reading
Backtest narrative is illustrative — built from typical calendar spread mechanics and historical volatility regimes, not from live broker fills. Past performance, simulated or real, does not predict future results. See methodology.