← The Wheel

Questions people who sell puts ask us

Short answers in words. Open “the numbers” under any answer if you want the figures. Everything here is measured on real option quotes; nothing is a forecast.

Why do the returns look low?

Because we count them the hard way. The whole strike is the collateral, there is no margin, every sale is priced at the bid, and the record runs every year since 2016 including the bad ones. Counted like that, selling puts earns in the low-to-mid teens a year on the account, with a drawdown you can sit through. Someone making eighty percent a year is using margin and has not been through a year like 2022 yet.

The numbers

Per trade: about +0.4% of collateral in 11 days, 9–12% a year. On an account: $50K about 16%, $1M about 14% with the whole account as collateral and 20% per lot; worst drawdowns −13% to −17%. Out-of-sample years run higher than the full window.

Are you maximising risk-adjusted return, or just return?

We maximise what a dollar of collateral earns per day, at a price that actually fills. Nine trades in ten win. The worst trades cost about a sixth of the collateral, never more than the strike. Two losing years in eleven, and the buy-back rule beat holding to expiry in every one of those years, including the losing ones.

The numbers

Win rate 89–90%; worst 1% of trades −15% of collateral; 2018, 2022 and early 2026 negative; hold-to-expiry velocity 6% a year against 9% with the half-credit buy-back on the same trades.

We sell when spreads are wide, close when they are tight, and move capital fast. Why not that?

Three things the tape says. A wide spread is not extra premium, it is uncertainty about what you will get: priced at what the market actually pays, wide-spread names do not earn more per day than tight ones, and a fair-price order in a wide spread fills less than half the time. Closing when it is tight is exactly what our resting buy-back does, every session, without anyone watching. And rotating faster than half the credit does not earn more per year, because every extra round trip costs a spread.

The numbers

At the bid: spreads under 5% earn 11% a year, 10–15% spreads 7%. The model price is touched 73% of the time in tight names, 42–44% in wide ones; the bid 94–95%. A resting buy inside the ask is touched 30%, at the ask 95%. Buying back at 25% captured vs 50% captured: the same velocity, less credit per round trip.

How do I get alerted when it wants to sell?

The order is the alert. Every five minutes during the session, any validated name that is rich on a quote you could actually trade becomes a sell order at the bid, and you get an email that says so. A second email comes when it fills, with the buy-back level. There is no earlier warning, because the program acts the moment its conditions are met.

The numbers

Passes 09:00–15:55 ET. Discord and email to every beta member on placed, filled, closed, assigned, expired and stock-sold; nothing else emails.

Is this real money? What is the live track record?

Paper, since the eleventh of September. The fill record on the Prime Book tab is the whole live record and we will say that before you ask. The studies behind the rule are on real option quotes going back to 2016; the product is days old.

The numbers

Paper account at Alpaca, every order a real limit order at the bid; fills, non-fills and closes all recorded and shown.

What happens when I am assigned?

Holding to expiry gets you assigned on roughly a quarter of trades; buying back at half the credit makes it rare. Before the first order the card shows what the shares would cost you and what a bad open after assignment would cost. Prime Book sells assigned shares at the next open. On your own names the default is to keep them.

The numbers

25% of held-to-expiry trades assigned; the buy-back exits most trades in 11 days. Early assignment is detected the next session and handled the same way.

Why does it say my name is not validated?

Every name has to clear its own gates on this exact contract before we will show an order on it, and most do not. The card names the gate the name failed. A name that fails gets a reading and a history, never an order.

The numbers

168 of 837 tracked names are validated on the 30-delta, 30–45 day, half-credit rule. GLD fails gate V3; TSLA and NEM fail.

How much of my account should this be?

That is yours to decide, and the product never sizes your money. Every order shows its collateral in dollars so you can judge the fit. The honest range for a whole account, by size, is on the methodology page, once.

The numbers

v1 does not ask for account size. The four account sizes quoted are $50K, $100K, $250K and $1M; nothing in between is interpolated.

What does it not do that I already do?

It never assumes a fill at the mid. It never pays up to get out. It has no stops, because none survived testing. It uses no leverage. And it does not pick names by score, because under a capital cap a ranked list did no better than a random one. What it does is tell you whether today is a day to sell a given name, at a price that fills, and then rest the exit.

The numbers

Sold at the bid, bought back at a level; no price or time stop; ranked vs random under the cap: about the same.

Paper account. Decision support, not advice. Every figure behind these answers is a study on file with its date and its decision record; ask and we will show the file.