Covered calls on shares you already own

Selling a call against stock you have held for years is a completely different trade from selling one against stock you bought this morning — and almost every covered-call calculator on the internet gets it wrong, because it divides the premium by your cost basis.

That single choice of denominator is how a perfectly sane trader ends up writing a strike 6% out of the money on a position with a 44% unrealized gain, then discovers in September that they have sold the whole thing and owe a decade of capital gains in one tax year. The premium was real. The return calculation was not.

Five real AAPL strikes, priced off the chain, with what each one earns decomposed into premium, capped upside and the delta you sold — and a fairly opinionated view of which column matters.

The denominator problem

Suppose you own 100 shares of AAPL. You bought them at $214.30. The stock is $308.91 today. You are up $94.61 a share — a 44.1% unrealized gain — and you would like to get paid for the fact that the stock has gone sideways for a month.

You sell the 18 Sep $325 call for $6.01. What did you just earn?

• $6.01 ÷ $214.30 = 2.80% — the cost-basis answer. Flattering and meaningless: your basis is a historical fact that has nothing to do with what the option is worth today.
• $6.01 ÷ $308.91 = 1.95% — the correct answer. The call is written against $30,891 of current share value. That is the capital the trade is actually using.
• $6.01 ÷ $16.09 of capped upside — the interesting answer, and the one nobody computes.

This is why our engine’s coveredCallYield returns two families of numbers: the period returns (premium yield, if-called return, both annualized against spot) and one deliberately un-annualized total return on basis. Annualizing a 44% gain that took four years over a 48-day option produces a number with a comma in it. We refuse to print it.

The position

Five strikes, priced

The two annualized columns move in opposite directions, and that is the entire strike-selection decision in one picture.

• Premium yield falls as you go up — 23.3% at the $315 strike, 9.4% at $335. Further out of the money, less credit.
• If-called return rises as you go up — 38.3% to 73.6% — because a higher strike means more share appreciation captured before the cap bites.
• The strike you pick is a statement about which of those two outcomes you would rather have, weighted by how likely each is.

Delta gives you the weighting. At the $325 strike, delta is 0.32 and the engine’s true risk-neutral probability of finishing above $325 is 30.0%. So roughly seven times in ten you keep the shares and the $601; three times in ten you sell at $325 having also banked the credit. (Delta and probability are close here but not equal — that gap has a cause, and it widens on high-IV names.)

The deep-in-the-money trap

Every few months someone discovers that the $290 call pays $26.20 — four times the $325 call — and posts about it as if they have found an edge.

The $290 call’s extra $20 of premium is your own money handed back to you — it is intrinsic value, which is to say share price you already had. Strip it out and the deep-ITM call pays $7.29 of actual time value against the $325 call’s $6.01, while surrendering every dollar of upside and making assignment near-certain. You would collect $14.81 a share less in total.

The last row is the cleanest way to see it: time value per unit of delta. You are selling delta — that is the risk you are taking on — and the $325 call pays nearly twice as much for each unit of it. Premium is not a number to maximize. Premium per unit of what you gave up is.

How much downside a covered call actually protects

“Covered calls give you downside protection” is technically true and practically almost useless. The protection equals the premium — nothing more.

A 2% cushion is a rounding error on a stock with a 28% implied volatility, whose one-standard-deviation move over these 48 days is ±$31.70. If your reason for selling calls is protection, you want a collar or a smaller position, not a call. If your reason is income on a position you intend to hold through the noise, the cushion is a pleasant side effect and the premium is the point.

Duration: why 20 days annualizes better

The short-dated call annualizes 1.57× better — the standard theta-decay argument, and it is genuinely true. Option value scales roughly with the square root of time, so 2.4× the days buys only about 1.55× the premium. Selling the front expiry repeatedly harvests the steep end of the curve.

It is also 2.4× the roll count: eighteen decisions a year at 20 days (365 ÷ 20) against 7.6 at 48, each one a chance to be assigned at a bad moment, each one crossing a bid-ask spread. On a penny-wide AAPL spread that friction is negligible; on a name with a $0.15 spread it eats a fifth of the advantage. And short-dated calls carry far more gamma, which means the strike that was comfortably out of the money on Monday is a problem on Thursday.

The strike below your basis

There is one rule in covered calls that is close to absolute: do not sell a strike below your cost basis unless the total premium collected on the position already covers the gap.

Here that would mean writing a call below $214.30, which on a $308.91 stock is a strange thing to do deliberately. But it happens constantly in the wheel, where an assignment leaves you holding shares at $110 with the stock at $95, and the only calls paying anything are struck at $100. Selling that call converts an unrealized loss into a realized one for a few dollars of credit. Track the adjusted basis across the whole cycle and you will know exactly where the line is.

What assignment actually costs

If AAPL finishes above $325, the shares go. Here is the full accounting:

A 54.5% total return on cost basis, which is a lovely number and also completely irrelevant to whether writing this call was a good idea — that gain was there before you sold anything. What the call earned is $6.01 plus $16.09 of captured appreciation on $308.91 of share value over 48 days: 7.15%, or 54.4% annualized.

A boring-stock comparison

Matched on delta rather than on strike distance, AAPL pays about a quarter more for the same probability of assignment — almost exactly the ratio of the two implied volatilities (28.0% vs 20.9%). That is not an inefficiency. It is the market charging more for a stock that moves more, and you are being paid precisely for the extra risk you are taking. There is no free premium anywhere on this table.

Do it in the builder

The comparison version — the same structure on KO — is worth opening in a second tab. Two ladders side by side is the fastest way to internalize what implied volatility is actually paying for.

If you already hold the shares, import your positions from a brokerage screenshot, or from a trade-history export in the one spreadsheet layout the file reader takes, and the call attaches to the real lot with the real basis, so the if-called number reflects your actual tax lot rather than an assumption in a blog post. More strikes and current chain context on the AAPL covered-call page.

Caveats

Next: how delta and IV actually determine a strike, or the full wheel cycle if you got here from an assignment.