Swing trading guide
Short answer
A Judas Swing is an ICT term for an early false directional move followed by a reversal. This page tests one narrow version: the first break of a frozen reference range must close back inside on the same bar, before a fixed session deadline. A break alone is only a candidate; later price action cannot turn an unconfirmed candidate into a valid historical entry. The rules below are a research definition, not an official ICT formula or evidence of institutional intent.
The Inner Circle Trader published Understanding The ICT Judas Swing in December 2017. The lesson provides the terminology; the single-bar confirmation, windows, and exits below are explicit choices for testing, not a transcription of the original method.
Use Europe/London for every time below. An IANA timezone follows daylight saving changes. Changing the TradingView chart clock only changes displayed times; Pine calculations need an explicit timezone. Do not mix these hours with a New York-time variant.
Freeze range, wait for first excursion, confirm or reject, then record the outcome. A high-side rejection tests a short; a low-side rejection tests a long. Neither direction is known to succeed when the confirmation arrives.
Price can break a range and continue. The rule therefore needs an invalidation level, a time exit, stated costs, and a record of every eligible day, including days with no trade.
Use one GBPUSD feed, five-minute standard candles, and Europe/London time. Record its minimum tick. This is an intraday session test despite the word swing in its name.
The feed and interval determine the observed extremes. Changing either changes the experiment.
Collect bars opening from 00:00 through 05:55. At 06:00, freeze their highest high H, lowest low L, and midpoint M = (H + L) / 2 for that London date. Skip dates with incomplete reference data.
Later prices must never widen the range used to judge an earlier breakout. The 06:00 to 07:00 gap is outside both research windows.
From 07:00, take the first bar whose high reaches H plus one tick or whose low reaches L minus one tick. Require that same bar to close strictly between L and H. If it closes outside, on a boundary, or crosses both sides, reject the day. Do not search for a prettier later reversal.
This strict same-bar variant removes discretionary waiting and prevents selecting the winning sweep after seeing the session.
Require the confirmation close and next scheduled five-minute bar open to be before 10:00 London time. A high-side excursion tests a short and a low-side excursion a long at that next open, with costs. Allow one attempt per date; reject a missing next bar.
A bar opening at 09:55 closes at 10:00, too late under this rule. Never backdate an order to its earlier extreme.
For a short, set the stop one tick above the excursion high; for a long, one tick below its low. Target M. If still open, request a market exit when a bar closes at 11:00 and record the next available fill. Record gaps across a stop or target as execution exceptions, without inventing a fill at the planned level.
Confirmation does not guarantee continuation. An adverse move after confirmation is a losing setup, not a reason to delete the date.
Record missing data, no excursion, rejected excursion, late confirmation, filled trade, and execution exception separately. Freeze all rules before evaluating later dates and include spread, commission, and slippage.
The denominator includes quiet and failed sessions. A collection of reversal screenshots cannot establish a win rate.
| Risk | What to check |
|---|---|
| Hindsight labeling | Log the range, sweep timestamp, confirmation close, eligible entry, and rejected setups without looking at later bars. |
| Session drift | Keep Europe/London fixed for this test and inspect both UK daylight saving transitions. A New York-time variant needs its own declared windows. |
| Intrabar ambiguity | Do not assume whether a target or stop was touched first inside one bar. Use lower-timeframe data or record the assumption. |
| Thin or volatile conditions | Report spread and slippage by session. Exclude a date only through a rule written before reviewing its outcome. |
| Overfitting | Choose the windows and buffers on one sample, freeze them, and publish results from a later holdout sample separately. |
These templates define a research process. They are not trade calls or evidence that a setup will make money.
Rule to test
Take H = 1.27000, L = 1.26800, and a tick of 0.00001 for all three cases. At 07:10, the first excursion reaches 1.27030 and the bar closes at 07:15 at 1.26980. Its low stays above L. The close is inside the frozen range and before the deadline. With an illustrative next-open short fill of 1.26970, the stop is 1.27031 and the target is 1.26900.
Planned distance to the stop is 0.00061, or 6.1 pips; distance to the midpoint is 0.00070, or 7 pips. Gross reward/risk is about 1.15 before costs. This is a valid signal fixture, not an assertion that the target is reached.
Rule to test
At 07:10, the first excursion reaches 1.27030 but closes at 07:15 at 1.27020, above H. The day is rejected without an entry. A later rise to 1.27100 does not change that decision; neither would a later fall below H.
Record one rejected excursion and zero trades. Selling at 1.27030 would assume the reversal before confirmation. Waiting for a later close inside would be a different strategy and needs a separate predeclared rule.
Rule to test
No earlier excursion occurs. The 09:55 bar reaches 1.27030 and closes inside at 1.26980 at 10:00. The next scheduled bar also opens at 10:00. Both fail the strict before-10:00 deadline, so there is no entry.
Record a late confirmation even if price reaches 1.26900 at 10:10. The outcome cannot move the deadline. Check the close timestamp, not only the 09:55 label printed beneath the candle.
Illustrative prices only: H = 1.27000, L = 1.26800, M = 1.26900, and a feed tick of 0.00001. One GBPUSD pip is 0.00010 in these examples. These are arithmetic fixtures, not historical trades or suggested settings.
Record the London date, feed, range boundaries, first excursion open and close times, confirmation result, scheduled entry, actual simulated fill, stop, midpoint, and rejection reason. Preserve the record even if price later reverses.
A high and low on the same candle do not reveal their sequence. Reject a candle crossing both reference boundaries; inspect lower-timeframe data when an open trade touches both stop and target in one bar.
Use Trading Journal plan and review notes to compare the confirmation time with your actual entry. An entry before the required close is a process violation even if it profits. Journaling does not prove that the setup has an edge.
Give Pine Script AI Coding Agent the frozen-range rules and ask for a state machine with collecting, armed, confirmed, rejected, and finished states. Request plots of H, L, the first excursion, and the confirmation timestamp. Review generated code and test its date reset, timezone, and order timing in TradingView. Pineify helps create and revise the script; it does not supply a certified Judas Swing signal, guarantee fills, or demonstrate profitability.
Build an auditable Pine Script testSources checked 2026-10-17
This page is an information tool, not investment advice. Judas Swing and ICT labels do not prove manipulation, predict a reversal, or promise returns. The GBPUSD rules and invented prices are research fixtures, not trade recommendations or observed results. Simulated results can omit real fills, spread changes, slippage, and market disruptions, and they do not guarantee future performance.
