Skip to content
All library documents

Testing Weekly Short Straddles Through the COVID Volatility Shock

Article QuantInsti blog

Summary

The project backtests a mechanical strategy of selling an at-the-money SPY straddle each week, using options with roughly 45–60 days to expiry and holding each position until expiration. It describes sourcing option prices, matching entry dates with expiries and strikes, and calculating returns and drawdowns. The test involved overlapping positions, with up to eight open at once and capital allocated across them. It reports that performance was poor during the dataset period, which included the COVID market crash and rapid rebound.

The article explains why short straddles can earn premium when the underlying stays near the entry strike, while losses grow when realized moves exceed what option prices imply. It proposes a different strategy for further investigation: short at-the-money options with long out-of-the-money wings sized for initial vega neutrality, plus regular delta hedging. Entry timing, wing deltas, hedging frequency, and exits are possible variables to optimize.

The reported backtest uses no delta hedging or loss management and covers a crisis period, so it does not establish how the strategy performs across regimes. The proposed tail-protected variation is an idea for future testing, not a demonstrated result.

Key ideas

  • The tested strategy sold weekly SPY at-the-money straddles with roughly 45–60 days to expiry and held them to expiration.
  • The backtest reports poor performance during a period that included the COVID crash and rebound.
  • Short straddles collect premium when prices stay near the strike but can suffer large losses when the underlying moves sharply.
  • The proposed alternative adds out-of-the-money options for initial vega neutrality and delta hedges at the open and close.
  • Results from this unhedged test do not establish performance across market regimes or validate the proposed tail-protection approach.

Tags

Cited by

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.