Time Spread Backtesting Concepts (Part 1)

I previously introduced my recent manual backtesting of time spreads that I hope to automate going forward in Python. Let’s back up a bit and discuss some general concepts pertaining to time spreads themselves.

A time spread is composed of a short option in one expiration and a long option at the same strike price in a more distant expiration. I proclaim the primary profit mechanism is differential time decay with short option decaying faster than the long.

Time spreads have delta risk, vega risk, and term-structure risk.

Delta risk refers to the positional tendency to maximize profit over time as the underlying market trades toward the strike price. The [roughly] triangular risk graph illustrates this. At expiration, profit is greatest at the strike price. Above and below, expiration profit wanes until it becomes zero or negative with the market trading sufficiently far beyond the strike.

For a directional time spread placed with current market significantly above or below the strike price, delta risk refers to loss accumulation as the underlying moves farther from the strike or as the underlying remains sufficiently far from the strike as short-option expiration approaches.

Time spreads placed ATM are sometimes referred to as nondirectional. The position can [eventually] profit regardless of whether the underlying moves up or down as long as it remains within a limited range. The ATM time spread still has delta risk because enough market movement away from the strike price either today or by expiration will drive PnL negative.

Vega risk—tendency of a position to lose money due to IV change—exists because the long option, which is farther from expiration, has more extrinsic value than the short. Time spreads are therefore vega positive: as IV increases, the long leg will increase more in value than the short leg will decrease. As IV decreases, the time spread loses value and the expiration curve, which always indicates PnL at short-option expiration, moves down. This is true assuming all else remains the same.

Term structure—the difference in IV between expiration months—often does not remain the same.

Term-structure risk for time spreads refers to an increase in horizontal skew. Horizontal skew is short-option IV minus long-option IV. Horizontal skew is generally negative, which makes positive horizontal skew advantageous for entry.* Should skew revert lower, the short option can lose more extrinsic value than the long, which biases the spread toward profitability. Once a time spread has already been opened, however, increased horizontal skew biases the spread toward loss because short-option extrinsic value increases more than long.

I will continue next time.

*—Always question why horizontal skew is positive as part of your trade assessment.