Judas Swing in ES and NQ: Spotting the Sweep and Reversal

The usual way traders lose to a Judas Swing is by treating every quick poke through a high or low as one. A break of the overnight low gets bought, price keeps falling, and the stop goes. The ICT (Inner Circle Trader) concept describes a specific version of the move: an early-session push through an obvious level, a fast reversal, and a drive the other way. The definition is narrow on purpose, and the narrowness is what keeps you out of most bad trades.

The ICT definition

In the ICT framework the Judas Swing is a false move early in a session, taken during a window called a kill zone. The conventional windows, in Eastern time, are 7 p.m. to midnight (Asian), 2 a.m. to 5 a.m. (London), and 7 a.m. to 10 a.m. (New York). Other ICT teachers draw them differently, and nothing about these windows is proven to matter. Treat them as a filter you choose and test, not as a law of the market.

The explanation that goes with it is that large participants need counterparties, so price is pushed through obvious stop clusters to find them. That is an interpretation of why price behaves this way. I can't verify anyone's intent from a chart, and it doesn't need to be true for the pattern to be worth studying. What you can verify is whether price swept a level and reversed.

The three conditions

1. A visible pool of resting orders. Prior day high or low, overnight extremes, an obvious equal high or equal low, or a round number. The more obvious the level, the better the setup definition, but also the more traders are watching it.
2. The time window. The sweep happens in the session you've decided to trade, such as the New York morning.
3. Displacement. After the sweep, a large, decisive candle back through the level, usually leaving a fair value gap behind it. Without it, you only have a poke.

Identifying it in real time

1. Set a daily bias before the session from higher-timeframe structure. In a bullish bias, the Judas move you want is the downward sweep.
2. Mark liquidity: prior day and week extremes, overnight high and low, equal highs and lows.
3. In your window, watch for price to run toward the pool on the side opposite your bias.
4. Wait for the sweep (a break of the level, often only a few ticks) and the displacement back through it. No displacement, no trade.

Entry, stop, target

Three common entry methods, in order of patience:

• Fair value gap retest: enter on a pullback into the gap the displacement left, on a 1 or 3 minute chart.
• Structure shift: on a 5 minute chart, enter after the first higher low following the sweep.
• Optimal trade entry: use a Fibonacci retracement from the sweep low to the displacement high and look for entries between 61.8% and 79%.

The stop goes beyond the sweep extreme, because a return there means the idea is wrong. Targets are the next obvious pool: prior day high, equal highs, or a prior session extreme.

A hypothetical worked example

These numbers are invented. Bias is bullish. ES sweeps the overnight low, printing 5,012.00, then a displacement candle runs to 5,030.00. The retracement range is 18 points.

Risk is 5.5 points times $50 = $275 per ES contract, or $27.50 per MES at $5 per point. Reward over risk is about 4.4 to 1 before commissions and slippage. A limit entry may not fill, and if price runs without retracing you simply don't trade. The stop can fill worse than 5,011.50 in a fast market.

Why it fails

• No bias. Without a direction filter, every sweep looks like a reversal and half of them continue.
• Entering before displacement. Fading the first push is the same as catching a falling price. The sweep can extend.
• An unremarkable level. A random swing point without equal highs or lows or a session extreme isn't an obvious stop pool.
• Hindsight. On a finished chart every reversal looks like a Judas Swing. In real time you have to decide before the reversal.
• Moving the bias mid-session to justify a trade you already want.

How to test it

Replay the New York morning for each session and log only the setups that meet all three conditions as you would have seen them live: date, bias, level swept, displacement yes or no, entry, stop, result in R. Record the ones that failed too. Thirty logged examples show whether you can apply the rules consistently, but they are not a statistically proven edge, and a simulated fill doesn't validate queue position or slippage. Run the R results through the edge calculator to see how wide the uncertainty still is.

Relation to fixed reference levels

KPL levels are fixed reference prices anchored to the regular-session opening print, with four levels in the standard ES/NQ map. They don't identify liquidity pools. If a KPL happens to sit near a level that gets swept, it's one more reference you can note, not evidence that the sweep will reverse. See the KPL trading guide for how those levels are built.