Strike selection from a return distribution and implied volatility

Strike choice as a regime-aware map

Option premium analysis compares collected premium with the chance a chosen strike is reached, so a write sits inside a broader volatility and portfolio regime rather than as a standalone bet.

The archive workflow starts with a return distribution, reads that distribution as an expiration map, and only then asks whether the premium looks cheap.

Editorial reading: treat strike choice as a probability problem first. Convert the return map into a percentile strike, then stress-test that map with implied volatility and a walk-forward check before the premium looks cheap.

A four-week return map

Rolling four-week percentage changes in the S&P 100 from October 1976 through December 1995 formed a bell-shaped distribution with a mean of 0.7% and a standard deviation of 3.9%.

Under that sample, most four-week observations fell inside one standard deviation of the mean, a band from -3.2% to 4.6%. Standard deviation here is a historical spread of returns around their average, used to mark one-sigma and multi-sigma bands that frame out-of-the-money strikes.

A normal distribution is a bell-shaped model of percentage returns used only as a first map of how often a market has historically stayed inside a given range.

A percentile strike from the same map

The empirical 90th percentile of those monthly-scale S&P 100 changes was 5.1%, so 90% of observations were 5.1% or smaller.

If the index sits at 550, a writer seeking about a 90% historical chance that a short call expires unexercised can locate that strike from the same return map rather than from an arbitrary offset. That placement is a percentile strike: a strike set at an empirical or modeled return cutoff so a target share of historical moves would have left the option unexercised.

Sorting raw returns to read a percentile is presented as the purest historical-probability method, while a spreadsheet percentile function is the less cumbersome alternative.

Fat tails and short puts

Equity returns are described as only roughly normal and as having much fatter tails than a bell curve predicts, which makes short out-of-the-money puts especially difficult to evaluate from the normal model alone.

Fat tails are more frequent extreme moves than a bell curve predicts. Far out-of-the-money short puts are harder to size from a normal model alone for that reason.

Implied volatility after the historical map

Implied volatility is the market's priced expectation of movement, used to judge whether a historically distant strike is still cheap after the current regime is taken into account.

Editorial reading: a strike that looks distant on the historical return map is not automatically cheap. Check that distance against implied volatility before treating the collected premium as cheap.

Walk-forward check of the strike rule

Walk-forward analysis re-estimates volatility on earlier data, applies that estimate to the next unused period, then rolls the window forward so a strike rule is tested as one procedure.

The specified check estimates a standard deviation on an earlier window, applies it to the next unused year, then recalculates and marches forward. The archive illustrates that sequence with IBM data from 1982 through 1998.

If historical standard deviation proves too unstable to be useful, the text says to revise the rule or switch to a more stable underlier. Index options tend to be more stable than single-stock options.

Planned buybacks and gap risk

Naked short calls or puts can produce large, unhedged losses, so the text requires a planned buyback and mentions credit spreads as one way to cap gap risk when an exit at the intended price is unavailable.

A credit spread sells one option and buys a farther out-of-the-money option on the same side to cap gap risk that a naked write cannot exit at a planned price.