0DTE Strategy Research.

Methodology

How the data was collected, structured, and analysed — including the DST normalisation that materially changed the conclusions.

Data source

All trades recorded in this research are paper trades placed by an automated trading bot against live SPX 0DTE option chains. The bot recorded every entry and exit to a single CSV file with the following fields per trade: date, entry time, strike legs, credit collected, buying power requirement, profit target, status (closed or expired), exit time, exit P/L, and a free-text notes field.

The research period covered in this site runs from 2025-12-12 to 2026-05-07, comprising 91 trading days and 652 trades across 8 strategy variants. Five anomalous-credit trades from 2026-03-20 have been excluded as a data-quality measure (see the Caveats page).

What "0DTE" means

0DTE stands for "zero days to expiration" — options that expire on the same day they are traded. SPX (S&P 500 Index) options have daily expirations Monday through Friday, allowing the construction of multi-leg credit spreads that benefit from the rapid time decay (theta) in the final hours of the option's life.

The strategies analysed here are all credit-receiving structures — the trade collects premium at entry and aims either to (a) close at a profit when the spread loses value, or (b) hold to expiration and let the structure expire worthless for maximum profit. Losses occur when the underlying moves through one or both short strikes.

Strategy types

Iron Condor
A four-leg structure with a short call spread above the market and a short put spread below. Profits if SPX stays inside both shorts. Used in 20 Delta and 30 Delta variants (delta refers to the short strike's option delta at entry).
Iron Fly
A four-leg structure with the short call and short put at the same strike (typically at-the-money), with long protective wings further out. Higher credit, narrower profit zone, more sensitive to direction.
Dynamic 0DTE
A hybrid that selects between an Iron Condor and an Iron Fly based on the 30-minute price action since the open. Aims to adapt to whether the day looks trendy or rangebound.
Gap Filter 20D
A 20 Delta Iron Condor with an additional filter intended to skip entries on days that opened with a large gap from the prior close.

Definitions used in this site

Trade
One complete cycle from entry (multi-leg order filled) to exit (closed at debit, time-stop closed, or expired worthless / in-the-money).
Credit Collected
The total premium received at entry, expressed in dollars per spread (×100 for share equivalent).
Buying Power
The margin requirement assigned by the broker to the position. This figure is used as the denominator for any return-on-capital calculations on this site, but should be considered approximate — see the Caveats page for limitations.
Exit P/L
The realised profit or loss at exit, in dollars per spread, net of the original credit. Positive means a winning trade.
Win rate
Fraction of trades closing with Exit P/L > 0. Trades that close at exactly $0 are counted as losses for conservatism.
EXPIRED
A trade left to settle at the SPX closing print rather than closed manually before the bell.
CLOSED
A trade closed at a debit before expiration — either at a profit target, a time-based exit, or an early manual close.

The DST normalisation

The trading bot logs entry times in UK local time. The US market opens at 9:30 New York time (Eastern Time). Normally these are five hours apart (UK on GMT, US on EST → market open = 14:30 UK; UK on BST, US on EDT → market open = 14:30 UK).

However, US daylight saving begins on the second Sunday of March, while UK British Summer Time begins on the last Sunday of March. This creates an annual three-week window where the US has moved its clocks forward but the UK has not, so the offset is only four hours.

For the dataset analysed here, this window was 2026-03-08 to 2026-03-28. During this period, US market open at 9:30 ET corresponded to 13:30 UK, not 14:30 UK as in the rest of the year. The bot's strategy schedule recognised this and shifted UK clock times one hour earlier — but the raw CSV simply records the UK clock value.

Why this matters. If you bucket trades by UK clock time, a trade entered at "14:00 UK" on 2026-03-15 looks like a different bucket from a trade entered at "15:00 UK" on 2026-01-15. They are not — they are both at 10:00 ET = +30 minutes after open. Failing to normalise causes one strategy to appear to have two distinct entry slots when it has only one.

To correct this, every trade in this analysis has been re-bucketed by minutes after market open:

All time-bucket tables and charts on this site label entries as +15m, +30m, +1h00m etc. — relative to the open, not the wall clock.

What was excluded from analysis

What is & isn't on this site

This site contains: full per-strategy and per-time-bucket P/L summaries; monthly trajectories; day-of-week analysis for the top two strategies; and discussion of statistical limitations. It does not contain: the raw trade-by-trade CSV; specific strike levels or entry rules of the strategies; broker account details; or any personally identifiable information.

The raw CSV contains specific strike selection logic that is the intellectual property of the strategy author and is not published here. The aggregated statistics on this site are sufficient for evaluating the methodology and conclusions but not for replication.