0DTE Options: Who Actually Makes Money? Every Day Since 2022, Tested

0DTE stands for "zero days to expiration". It's an option that expires the same day you trade it. Buy one at 10 a.m. and by 4 p.m. it's either worth something or it's worth nothing.

They've taken over. In July 2026, same-day contracts were a record 66.2% of all S&P 500 (SPX) options traded, according to Cboe. In September 2026 about 3.4 million of them changed hands every day.

So who's making money? We took every trading day since May 2022, when SPX options began expiring every weekday. We priced that day's options from what the market itself was charging that morning, then checked what the S&P 500 actually did by the close.

What a 0DTE option is

An option gives you the right to buy (a call) or sell (a put) at a set price, the strike, until it expires. Most options have weeks or months left. A 0DTE option has hours.

That makes it cheap. An option's price is mostly paying for time, and a same-day option has almost none. So a small amount of money buys a lot of exposure to today's move. That's the appeal, and it's also the danger.

Same-day options aren't new. What's new is that the S&P 500 now has an expiry every weekday. Cboe added Tuesday and Thursday expiries in 2022, which filled the week. Since then, any trading day can be a 0DTE day.

Most of the trading is in two products. SPX is the full S&P 500 index option, about $775,000 of exposure per contract at today's prices. XSP, the mini version, is a tenth of that, about $77,500. Both settle in cash at the close, so there are no shares to deliver.

How we tested it

To know who wins, you need to know what the options cost. Cboe publishes an index called VIX1D. It's like the VIX, but it measures the move the market expects over just one day.

At the open, VIX1D is built almost entirely from that day's expiring options. So it tells you how big a move those options were priced for. Divide it by 15.9 (the square root of 252 trading days) and you get the day's expected move. A VIX1D of 16 means the options priced a typical move of about 1% between the open and the close.

Each morning we "bought" one at-the-money straddle on XSP: a call and a put at today's price. A straddle pays the size of the move, up or down, so it's the cleanest test of whether the market moved as much as the options said it would. Its price comes from the VIX1D reading. Its payout comes from the S&P 500's actual close. We did the same for a call and a put further out, the cheap "lottery tickets" many traders buy.

These are model prices, not real trades. Real buyers pay a bit more and real sellers get a bit less, because of the bid-ask spread and commissions. We'll come back to that, because it turns out to matter a lot.

What the options priced, and what happened

Over 1,111 trading days, the straddle priced an average move of 0.68% from open to close. The S&P 500's average move was 0.60%. The options charged for a bit more movement than the market delivered.

That gap held in every year we tested. It's the same thing that shows up in longer options, and it has a name: the volatility risk premium. People pay extra for protection and for the chance of a big win, and the sellers collect it.

Big days were rarer than the prices suggested. The index moved more than the full expected move on 26% of days. If the prices had been right, that would happen on about 32%. It moved more than double the expected move on 32 days out of 1,111, or 2.9%, against about 4.6%.

The sellers: small wins, and then one day

Selling that straddle every morning brought in $350 on an average day. The seller kept about 11% of it, roughly $39 a day. Over 1,111 days that came to $43,308, and the seller came out ahead in five of five years.

It's a steady climb with sudden drops. The worst came on 9 April 2025, the day the US paused most of its new tariffs. The S&P 500 jumped 9.9% between the open and the close. The straddle had brought in $1,626, and the seller paid out $4,916. That one day lost $3,290, about 84 average days of profit.

The second-worst day is the scarier one. On 10 October 2025 the options were cheap. VIX1D opened at 8.2, so the straddle brought in just $277. Then the index fell 2.8% by the close. The seller lost $1,602, about 6 times what they'd collected, on a morning that looked calm.

At its worst, the running total fell $7,435 from its high. And this is the gentle version. A seller with no protective options has no limit on the loss. A real account selling SPX straddles, ten times the size, would have lost $32,903 on that April day.

2025 was the weakest year, at $3,861. Four of the five worst days came that year. A seller's results depend less on the average day than on how many of those days turn up.

The buyers: the lottery tickets

Most retail 0DTE trading isn't selling straddles. It's buying cheap calls and puts that need a big move to pay off. So we tested that too: every morning, a call and a put each one expected move away from the open.

The calls paid off on 150 of 1,111 days, about 14% of the time. On the best one, 4 August 2026, a call paid 21 times its cost. Those days are why people buy them. But across every morning, each dollar spent on calls came back as 60 cents.

Puts did better, at 79 cents per dollar. They paid off on fewer days, 136, but the falls were bigger: a winning put paid 7.0 times its cost on average, against 4.5 for a winning call. Buying the straddle got back 89 cents. Nothing a buyer did came out ahead.

Then come the costs

Everything so far is before costs. The seller's $39 a day is what's left of a $350 premium. A commission of $0.65 a contract, common at US brokers, takes $1.30 of it to open two legs. The bid-ask spread takes more, and it gets much wider on the fast days that matter most.

Real data says the costs are the biggest part of the story. Researchers Heiner Beckmeyer, Nicole Branger and Leander Gayda studied Cboe trade records from January 2021 to February 2023. After SPX options began expiring every day, retail traders lost about $358,000 a day on them. More than 60% of that was trading costs.

So a buyer starts about 40 cents behind on a dollar of calls, then pays to get in. A seller's thin edge has to cover those same costs, then survive the April days. The steadiest earner is usually the market maker in the middle, who collects the spread from both sides.

Is 0DTE trading just gambling?

Not quite. A bet at a casino has a fixed house edge. Here the edge is small, it moves, and it can flip on a single day. Professionals use same-day options to hedge, to trade around news, or to sell premium with strict limits on the loss.

But for someone buying cheap calls on a hunch, the numbers look a lot like a lottery. Most tickets expire worthless, a few pay a lot, and on average you get back less than you paid. The big wins are real. They just don't add up to a profit.

If you trade them anyway

• Size every trade as a total loss. Most 0DTE options you buy will expire worthless, often within hours.
• Never sell one without protection. A straddle or naked option has no limit on the loss. Spreads cap it, and an iron condor caps both sides.
• Check the spread before you trade. On a cheap option, the gap between the bid and the ask can be a big slice of the price.
• Use limit orders. A market order on a fast-moving 0DTE option can fill far from the price you saw.
• Know your settlement time. SPX and XSP 0DTE options settle in cash on the 4 p.m. close.
• Keep a record. The wins are memorable and the slow losses aren't, so add up the whole year before you decide it works.

Outside the US

Same-day options aren't only American. In India, where index options are hugely popular, 59% of index options turnover was on expiry day in FY26, according to the market regulator SEBI. Its study found 87.7% of individual derivatives traders lost money that year.

Rules on who can trade options vary by country, and most brokers ask you to pass a knowledge test or meet experience requirements first. The maths in this article holds wherever you trade: an option priced for a bigger move than the market delivers loses its buyer money.

What this test can't tell you

• These are model prices built from VIX1D, not real fills. Real strikes come in steps, real prices skew toward puts, and every trade crosses a spread we didn't charge.
• We bought or sold at the open and held to the close. Most 0DTE traders get in and out during the day, which this doesn't test.
• The test covers 2022-2026, about four and a half years and one index. A different stretch, with a bigger crash, could look very different for sellers.
• VIX1D's opening value is an index reading, and it's annualised in Cboe's own business-time convention. We used the common rule of dividing by the square root of 252.
• Taxes aren't included. Prices were read on 2026-10-19.