Selling a 0DTE credit spread after SPX moves: +37% a trade for 45 days, negative over three years

Backtested on real SPX same-day option prices (45 to 51 sessions, July to October 2026), then checked on SPY one-minute data over three to eight years. Updated 8 Oct 2026.

Short answer: no, and this is a lesson in short samples. Selling a same-day credit spread in the direction of the morning's first move made +37% of the risk a trade over 45 days of real option prices. It passed my statistical checks, barely. Then I checked the same idea over three years of SPY, and it broke even or lost.

Selling against the move lost money every year. Selling far out of the money never lost in the short sample, but it collected so little that over the long run the occasional big loss more than wiped out the profits.

+37%a trade with the move, 45 days of real prices

-1% to -3%a trade with the move, three years of SPY

-14% to -18%a trade against the move, every year

~50%of first moves continued: a coin flip

What is a credit spread after a move?

A credit spread sells one option and buys a further one as protection, collecting a premium up front. A bull put spread profits if the index stays above the sold put; a bear call spread profits if it stays below the sold call. With same-day (0DTE) options, the premium decays to nothing by the close if the index stays on the right side.

The idea here is to wait for the morning's first meaningful move and then sell a spread: either with the move, betting it won't fully reverse, or against it, betting it has gone far enough. Both are popular, and they can't both be right.

The exact rules I tested

1. With the move

Trigger: the first time SPX moves 8 points from the open. Up: sell a bull put spread. Down: sell a bear call spread.
Strikes: the sold option from 5 points in the money to 20 points out of the money from the trigger price, with wings 5, 10 or 20 points wide.
Management: take profit at 25%, 50% or 75% of the credit, or hold to expiry. Stop at 0.5×, 1× or 2× the credit, or no stop. 288 combinations.
Fills: sold at the bid, bought at the ask, on real quotes.
Trend filter (also tested): only take the trade when the 15-minute 18 and 50-period moving averages agree with the move.

2. Against the move

Trigger: SPX moves 0.3 times its 14-day average daily range from the open (about 20 points) before 11:30. Up: sell a bear call spread. Down: sell a bull put spread. Same strikes, widths and management grid. Also tested at 0.2, 0.4 and 0.5.
A bigger version: after a move of a full average daily range, sell the spread half a range further out.

3. Far out of the money

After the 8-point move: sell a with-the-move spread with the sold option 0.75 or 1 average daily range away from the trigger.

The long-run check

Same-day option prices only exist for my short sample. For the long run I used SPY one-minute data, scaled every distance to SPY's price, and assumed the same credits as a share of the spread width that the real options paid. Held to expiry, since managed exits can't be replayed without option prices.

Results

With the move, 45 days: 198 of 288 combinations made money. The best (sell at the money, 5 wide, take profit at 75%, stop at 2× the credit) made +37% of the risk a trade and won 88% of 32 trades. It was positive in August and in September to October, and the chance of finding a result this good by luck across all 288 was 4.7%.
With the move, three years of SPY: held to expiry, −1% to −3% a trade overall (one in-the-money version made +2.6%). 2023 to 2024: −5% to −9%. 2025: around 0 to +9%. 2026 before the option sample: −2% to −6%. My 45-day window was the best stretch of all.
The trend filter lifted the 45-day result from 37% to 50%. Over three years of SPY, trades where the trend agreed did worse than trades where it didn't, in every year.
Against the move: only 52 of 256 combinations made money in the 45 days, against 93 for the with-the-move side at the same trigger. Over three years of SPY: −14% to −18% a trade in every combination and every year. The bigger version needed a credit of about 14% of the spread width to break even; the market was paying about 1%.
Far out of the money: no losses in the 45 days, but credits of only $10 to $44 at a full range away, and no credit at all on 7 of 32 days. On SPY, price closed beyond that distance on 4.5% of trigger days. At those credits, every combination lost money over the long run, in every year.
Did the first move continue? On three years of SPY and QQQ, the first move of about 0.1% closed further along about half the time. On SPY the up moves did a little better, which is the market's normal upward drift.

Why it failed

The first move doesn't predict the rest of the day. It continues about half the time. A spread that bets on either outcome is betting on a coin flip, and paying the spread to do it.
A short sample can pass every test and still be a regime. August to October 2026 happened to be a stretch where morning moves kept going. The result was real in that window and passed a strict check. It didn't survive a longer one.
Selling against a big move means selling into a trend day. After a large move, the index closed back inside its trigger level only half the time. Big-move days tend to keep going, and the options on the side you'd sell are cheap after the move.
Far out-of-the-money spreads hide their risk. A strategy that never loses for two months can still lose money if the one loss in twenty is many times the usual win.

Did I test it wrong?

Quite possibly. I tested the version of 0DTE credit spreads after a move described above, with settings I thought were fair, and it didn't work. That doesn't prove every version fails. I may have picked the wrong settings, the wrong market or timeframe, or missed a filter that experienced traders take for granted. Many strategies also depend on judgement that a set of fixed rules can't capture.

If you trade 0DTE credit spreads after a move and it works for you, I'd genuinely like to hear how. Tell me the chart and timeframe, what counts as a setup, any filters, where you get in, where your stop and target go, and when you get out. I'll test your version on the same data, the same way, and publish the result here, whichever way it comes out.

Send me your version

What would change my mind

A 0DTE credit spread rule that holds up over several years with real option prices, including a volatile stretch, not just the calm months where most of my option data comes from. Or a way to tell, at the time of the trigger, which days will keep going and which will reverse.

Questions

What is a 0DTE credit spread?

Selling one option and buying a further one, both expiring the same day. You collect a premium and keep it if the index stays on the right side of the option you sold. The bought option caps the loss.

Should you sell a credit spread with the move or against it?

Neither worked over the long run. With the move looked excellent for two months, then broke even or lost over three years. Against the move lost 14% to 18% of the risk a trade, every year.

Are far out-of-the-money 0DTE spreads safe?

They win almost every time, which is the trap. In my 45-day sample they never lost, but they collected so little that over three years of SPY the rare loss (about one day in twenty) more than wiped out the profits.

Does the first move of the day continue?

About half the time. On three years of SPY and QQQ, the first sizeable move of the morning continued no more often than a coin flip.

How it was tested

Data: real SPX same-day option quotes recorded every few seconds, July to October 2026 (45 to 51 sessions). Where a strike was outside the recorded range, it was priced from the option smile. SPY one-minute data from 2018 to October 2026, checked against a second source.
Engine check: a bug that valued a briefly missing quote as worthless was found and fixed before these results.
Backtest, not live trading: these results come from historical prices with trading costs included. Backtests usually flatter a strategy, so a failed backtest is a strong "no".

This page records my own testing. It isn't financial advice.

Disclosures · Privacy · [email protected]