Case study
A year of the wheel: smaller drawdowns, smaller gains. What was it selling?
The wheel sells a put, takes the shares if it is assigned, sells calls on them until they are called away, and starts again. It is often sold as income. Here is one path through it in Options Meerkat, on five large caps and then one stock, Coca-Cola (KO): what a year of it did, why, and what a backtest cannot promise.
The data: backtests over the year to the close of Friday, October 23, 2026, on each option's daily closing prices, and that close's chains for today's put. The answers the dashboard plays with each step were recorded then, from the same data; the dashboard around them is live. Backtests are part of Pro.
Recorded answers and backtests from the close of October 23, 2026, played on the live dashboard.
-
Step 1 of 5 · Five stocks, one year
Backtest it against buy and hold
“Backtest the wheel over the past year on META, AAPL, JPM, KO and NVDA against buy and hold: return, max drawdown, and how many times shares were assigned and called away.”
Each run sells a 30-delta put about 30 days out, holds it to expiry, and once assigned sells 30-delta calls the same way, from October 23, 2025 to October 23, 2026.
Stock Wheel Buy & hold Wheel max DD B&H max DD Assigned / called META −4.0% −12.5% 18.2% 30.1% 2 / 2 AAPL +15.9% +34.0% 5.9% 13.8% 2 / 2 JPM +9.8% +12.6% 7.4% 15.5% 3 / 2 KO +13.6% +32.6% 2.5% 8.5% 2 / 1 NVDA +14.6% +15.8% 5.1% 20.2% 0 / 0 The wheel trailed buy and hold on all five, and lost less on the one that fell. Its worst drawdown was between a quarter and three-fifths of the stock's. Smoother and lower is what the wheel is built to be: the gap is widest on KO and AAPL, the two that rose most. We follow KO. (The wheel, backtesting options)
-
Step 2 of 5 · One stock's year
Walk through the trades
“Backtest the wheel on KO from Oct 23, 2025 to Oct 23, 2026 and walk me through its trades: when shares were assigned, when they were called away, and how far KO ran past the call's strike.”
Date What happened Strike KO Dec 26, 2025 Put assigned 71 69.87 Jan 30, 2026 Called away 72 74.81 Oct 2, 2026 Put assigned 86 85.65 Oct 23, 2026 Call still open 90 92.71 KO rose from $69.94 to $92.71 over the year, and the wheel owned it for about two months of it. The rest of the time it held cash and sold puts below the price: seven of its nine puts expired, each keeping a credit of 63 cents to $1.26 a share. Its net premium for the year was $601 on $10,000.
The two stretches with shares show what the calls cost. In January the 72 call was sold for 46 cents and closed with KO at $74.81, $2.81 a share given up past the strike. On October 23 the 90 call, sold for 95 cents, was $2.71 in the money. In a stock that keeps rising, the wheel sells the rise for a small credit, and is then out of the stock, selling puts under it, while it goes on rising.
-
Step 3 of 5 · The one setting that matters
Change the delta
“Backtest the wheel on KO over the past year at 15 delta and at 40 delta, against the 30-delta run: return, max drawdown, and assignments.”
Delta Return Max DD Premium Assigned / called 15 +4.0% 0.7% $410 0 / 0 30 +13.6% 2.5% $601 2 / 1 40 +14.0% 3.0% $937 3 / 2 The delta is roughly the chance the option ends in the money, so it decides how often the wheel holds shares. At 15 delta all eleven puts expired: little premium, no assignments, and the return of a cash account. At 40 delta it took half again the 30-delta premium and was assigned more often, and ended about where 30 delta did, with a deeper drawdown. Not one of the three came near the stock's +32.6%. (The Greeks)
-
Step 4 of 5 · From the backtest to today
Find today's put, and what is in its way
“If I started the wheel on KO today, which put would it sell, what does it pay, and is anything in the way before it expires?”
- The put
- Nov 27 90 33 days, 30 delta
- Bid / ask
- 1.09 / 1.35 $122 at the mid
- Cash set aside
- $9,000 1.4% in 33 days
- Breakeven
- $88.78 KO at $92.84
- Reports
- Tue, Oct 27 before the open
The backtest's rule picks a put that holds KO's report, two sessions away; its last four reports moved it about 3.5% on average, against a strike 3.1% below the price. The Fed decides the next day. The rule does not look at the calendar; you can. And it picked the Nov 27 expiry, whose 90 put has 92 contracts of open interest and a 26-cent spread, when the Nov 20 one beside it has 5,511 and a 9-cent spread, with a bid 8 cents lower: a backtest's price is a daily close, not the fill you would get.
-
Step 5 of 5 · The edge of the data
Name what a backtest cannot promise
“What can't a year's wheel backtest on KO tell me about running it from here?”
One more run says it plainly. The same five backtests, started four days later, on October 27, 2025:
Stock From Oct 23 From Oct 27 META −4.0% +6.4% AAPL +15.9% +7.8% JPM +9.8% +4.0% KO +13.6% +8.5% NVDA +14.6% +30.6% - One path. A year of the wheel is about a dozen trades, and which dozen depends on the day it starts. A single run is one draw, not the strategy's return.
- The fills. Prices are each contract's daily close, a few of them modeled where a contract did not trade; assignment comes only at expiry, and dividends are left out.
- The year ahead. KO's year was a steady climb. A falling year, or a gap through the strike on a report, is a different wheel.
- Your situation. The cash it ties up, what else your account holds, and how much of it you can lose.
The data narrowed “the wheel is income” to “on these five stocks, last year, it gave up most of the rise for a smoother ride, and how much depended on the day it began.” Whether that trade suits you is a decision the data does not make for you.
-
Your turn
Ask your own questions
Backtests run on any symbol of the S&P 500 or the Nasdaq-100, with Pro. Open the dashboard, sign in, press @, and ask Meerkat:
Sign in and ask Open the dashboard
More case studies: a high IV percentile · deep put blocks · an earnings report · dealer gamma
Click a question to copy it. Nothing on this page is investment advice; past results do not promise future ones, and options can lose their whole value quickly.