The Sweep Before the Move: Reading Equal Highs, Killzones and the COT Report Together

Every Smart Money Concepts chart on the internet has the same two words written above a cluster of wicks: equal highs. The idea is right. Resting buy-stops above a pair of matching highs are the fuel that lets a larger player fill a sell order without moving the market against himself. Price runs up, takes the stops, and reverses. You sell the reversal.

The problem is not the idea. The problem is that equal highs form constantly, and most of them are not swept — or they are swept and price keeps going. If you sell every sweep you will be right often enough to keep believing in it and wrong often enough to never make money from it.

What separates the sweeps that reverse from the sweeps that run is not on the chart. It is in two things most SMC traders never look at together: when the sweep happens and who is already positioned on the other side of it. This post is the three-part check we run on the terminal before a sweep becomes a trade, worked through on gold as it stands this week.

Part one: the level has to be a real pool

Two highs within a fraction of an hourly ATR of each other are a pool. One high is a swing. The difference matters because a pool has stops stacked above it from everyone who shorted the first high and everyone who shorted the second — a swing only has one set. The terminal's rule is mechanical: two H1 swing highs within 0.2 ATR of each other, and we add 0.4 ATR of room above the higher one, because a sweep does not stop at the exact tick. It overshoots. The stop for any short goes beyond that room, never on the highs themselves.

On gold right now the hourly chart has that exact shape. Two highs printed at 4,217.6 and 4,217.8, with price sitting just under them at 4,214. With a 12-point hourly ATR the area of interest is 4,217.6 to 4,222.7 and the stop sits at 4,224.5. Above that there is a second supply at 4,294–4,301, a third at 4,318–4,331, and the 4-hour supply at 4,326–4,336.

That ladder is the first lesson. The 4-hour supply is a perfectly good zone. It is also more than nine hourly ATRs away. If your execution chart says "sell from 4,330" while price is at 4,214, you are not executing a bias, you are waiting for a different week. The area that matters is the nearest one price can actually reach in the session you are trading — and here that is the equal highs, three points away.

A level is only an execution level if price can reach it inside the window you are trading. Everything further away is a map.

Part two: the sweep has to happen in a killzone

Liquidity is taken when there is enough volume to take it. That volume arrives on a schedule. The London killzone (02:00–05:00 New York time) and the New York killzone (07:00–10:00) are when the institutional order flow that actually moves gold is active. Asia (20:00–00:00) builds the range that London and New York then raid.

A sweep of equal highs at 11:00 New York time, after the morning session has finished, is far more likely to be a slow drift that keeps going than a deliberate raid that reverses. The same sweep at 08:15 New York, into the first hour of the US session, with the London high already set, is the textbook pattern: Asia builds the highs, London pushes into them, New York takes them and reverses.

Our rule is simple and deliberately strict. The H1 trigger — the market-structure shift after the sweep — must print inside a killzone. If the sweep happens outside one, we wait. If the shift comes outside one, we skip the trade. This single filter removes more losing sweep trades than any refinement of the zone itself.

Friday afternoons deserve a special mention because gold loves to run stops into the weekly close. A sweep at 14:00 New York on a Friday is almost never the beginning of a new leg. It is the end of the old one. The next real window is Asia on Sunday evening.

Part three: the positioning has to allow the reversal

This is the part that makes the method ours rather than another retelling of ICT. A sweep reverses when the side being squeezed is crowded. If speculators are heavily long and price runs the highs, there is nobody left to buy the breakout — the late longs who chase it are the fuel for the reversal. If speculators are flat or short, the same sweep can simply be the start of a trend, because there are still buyers who have not bought.

The CFTC's Commitments of Traders report tells you which world you are in, once a week. The numbers for gold from the latest report:

Read that as a trader rather than an economist. Speculators are heavily net long but they cut seven thousand contracts last week while open interest fell — that is long liquidation, not fresh buying. The commercials, who are usually right at turns, are covering shorts into the dip. The 3-year COT index at 60 is "mild long": not an extreme, not a crowd that has to unwind.

What does that mean for the equal highs at 4,218? It means a sweep there is a neutral event on positioning. There is no crowded long side begging to be squeezed, so a sweep-and-reverse short is a short against the trend with ordinary odds, not a high-conviction trade. If the index were above 85 — speculators at a three-year extreme, the way they were in the spring — the same sweep would be one of the best setups the market offers, because every late long above the highs becomes a seller on the way down.

Our scale, so you can apply it yourself:

• COT index above 80 or below 20 in the direction being swept: the reversal has fuel. Full size.
• Index between 60 and 80: ordinary. Take the trade only with the killzone and the structure shift both in place, and take profit at the first liquidity below.
• Index between 40 and 60: no edge from positioning. The sweep is noise unless the higher timeframes are already in your direction.
• Hedgers against the crowd (commercials at one extreme, speculators at the other): the strongest signal this report gives. It does not time the turn, but it tells you which side of the sweep to be on.

Putting the three together on this week's gold

Here is the whole read as it sits on the Market Trend tab right now.

• Bias: the daily and 4-hour structure is still bullish — higher highs since the 4,091 low on the 7th. The terminal's composite bias for gold is bullish. Any short is a counter-trend trade.
• Area: equal highs at 4,217.6–4,222.7, three points above price. Stop beyond the sweep room at 4,224.5. First liquidity below at 4,197.6 (the previous hourly low), then the hourly demand at 4,166–4,169.
• Mode: reversal. H1 is still making higher highs, so there is no trade until the sweep prints and an hourly structure shift follows — a close below the last higher low. Shorting into the highs because they are "equal" is exactly the mistake this post is about.
• Timing: the next windows are London 02:00–05:00 and New York 07:00–10:00 New York time. A sweep outside them is ignored.
• Positioning: COT index 60, speculators trimming, commercials covering. No crowd to squeeze. Ordinary odds; take profit at the first pool, do not hold for the demand zone.

Notice how different that is from the usual "equal highs → sell" picture. Two of the three checks are lukewarm. The honest read is: a small short is available if the sweep and the shift both arrive inside a killzone; otherwise the better trade is still the long from 4,166–4,169 that the bias has been pointing at all week. That is what a bias-correct execution looks like. The chart tells you where. The clock and the report tell you whether.

A note on what the tools cannot do

The COT report is published on Friday for the previous Tuesday. It will not tell you what happened on Wednesday. It tells you the posture the market carried into the week, which is what matters for whether a sweep reverses — posture changes slowly, sweeps happen in minutes. Treat it as the weather, not the forecast.

And none of this is a signal. Nobody can tell you a sweep will reverse. The point of the three-part check is to only pay for the sweeps that have a reason to reverse, and to sit out the rest. Over a hundred setups that is the entire edge.

Everything on SMC Terminal is analysis and education, not a recommendation to buy or sell. The numbers in this post are from the terminal's COT Positioning and Market Trend tabs as of the date of publication; verify anything decision-critical at cftc.gov. Manage your risk.