promised i'd come back with data. here it is.
88,017 contracts. 421 symbols. 2018 to 2025. seven market regimes. every contract had a known outcome, expired worthless or got assigned. no simulations.
first the liquidity piece held up exactly as expected. thin OI means wide spreads and wide spreads quietly eat your premium before the trade even starts. learned that the hard way when commissions were $25-50 a contract. nothing new there.
the smart money angle is where i got it wrong.
hear me out…
my original theory was that high OI clustering at a strike meant institutions were positioned there. smart money telling you something. i wanted to know what they know.
the data says the opposite.
split all 88k contracts into four buckets by how crowded each strike was relative to other strikes in the same expiration (i call it "relative OI"). least crowded to most crowded.
the crowd is right about one thing, they find where the premium is. most crowded strikes yield nearly double the least crowded. but the trap is this… ur chasing twice the premium for maybe 4 fewer wins out of every 100 trades. the extra premium isn't free money. it's compensation for a strike the market has already crowded into.
2025 is the starkest example. most crowded strikes were yielding 46.8% annualized. least crowded were at 18.6%. 28-point gap. and the boring strikes still won more often.
my gut said high OI was the signal. the data said the edge goes the other way.
thing is, boring stocks naturally attract less crowded strikes. not bc anyone planned it that way, just bc institutions aren't piling into WFC or ED calls the same way they pile into NVDA. lower premium, less crowded, win more consistently.
been doing boring for 25 yrs for completely different reasons. turns out it was right for this one too. just didn't know why until now.
still think OI is the first thing to check before selling a call. just not for the reason i originally thought.
curious if you're chasing the crowded strikes or avoiding them.
OI follow up; ran 88k contracts. my gut was wrong. the data is more interesting
byu/sashazaliz inoptions
Posted by sashazaliz
2 Comments
Is your shift key broken or are you just trying to hide that you did not write this?
that extra premium at crowded strikes is basically the market marking up IV — dealer hedging gets harder when everyone’s positioned at the same strike. so you’re getting compensated for crowd risk, not harvesting edge. makes sense the win rate gap ends up small.
curious whether that held through 2020 — crowded strikes during covid had some weird spread/OI dynamics that might mess with the data.