Gold sits in a 4 dollar range for forty minutes, then leaves it in two candles and never looks back for the rest of the session. That little range is the interesting part. Somebody had far more to sell than the buyers at that price could take, and when the buyers ran out the price had to travel until it found new ones. The box you draw around those forty minutes is a supply zone, and its only claim on your attention is that the seller may not have finished.
The zone is a record of an order that could not be filled
Large participants cannot click one button and be done. A pension fund hedging currency exposure or a bank working an order for a corporate client has to feed size into the book over time, and the book is thin compared with what they need. When they run out of counterparty at a price, the market gaps away from that price. The move is the evidence.
Two consequences follow, and everything else in this method comes from them. First, the origin of a sharp move is where the imbalance started, so that is where the unfilled remainder of the order is most likely resting. Second, an order worked over a morning is rarely worked once. Institutions that were buying at 1.0840 on Tuesday are often still buying at 1.0840 on Wednesday, because their mandate has not changed.
This is why zones and classic support and resistance point at the same places so often. They are two descriptions of the same fact: the price where someone with size cares.
Marking one so you do not redraw it later
The recipe is mechanical. Find a move that left an area quickly, in one direction, with little overlap between candles. Walk back to the last candle that closed against the direction of that move. The band from that candle's open to the far end of its wick is the zone. On a demand zone that is the last down candle before the rally, on a supply zone the last up candle before the fall.
Two mistakes make charts unusable. Drawing zones on every small pause turns the screen into wallpaper, and once there are eleven boxes on a chart, price is always inside one of them, so the method predicts nothing. The other is redrawing a zone after price has moved through it, which quietly converts an analysis method into a way of never being wrong on paper. Mark the box when you find it, screenshot it, and let the market vote.
Timeframe decides how much the zone is worth. A zone from the H4 chart holds orders from participants who trade in size. A zone from the M1 chart mostly holds retail stops and one algorithm's morning. If you are mixing them, be explicit about which chart you are trading, using the same higher to lower discipline covered in the guide to chart timeframes.
A zone gets weaker every time it is used
Each return to the zone consumes some of what was left. The first revisit is the one worth trading. By the third, the remaining orders have usually been filled, and what looks like an area of demand is really an area where price has learned to pass through unchallenged. That is the mechanism behind the old observation that levels break after they have been touched too often.
| What to look at | Weaker zone | Stronger zone |
|---|---|---|
| Departure from the area | Slow drift, heavy overlap between candles | Two or three wide candles with little overlap |
| Time spent building it | Hours of two way trade | A short base before the move |
| Prior visits | Tested two or more times | Untouched since it formed |
| Position in the trend | Against the higher timeframe direction | In line with the higher timeframe direction |
| Approach speed | Price arrives in a fast, one way run | Price drifts back slowly and loses momentum |
That last row is the one traders argue about most. A slow, corrective return means the market is not committed to the direction and a resting order can still turn it. A violent arrival means something has changed since the zone formed, and the participant who left orders there may well have cancelled them.
Entering without catching a falling knife
There are two entry styles and they suit different temperaments. A limit order at the edge of the zone gets the best possible price and a defined risk, and it will be filled on every single zone that fails. A confirmation entry waits for price to reach the zone and then produce a reaction on a lower timeframe, a rejection candle or a small structural turn, before committing. The confirmation entry costs part of the move and refuses a portion of the losers.
My own preference is the confirmation entry on anything counter to the higher timeframe direction and a limit only when the zone is aligned with it. The stop belongs on the far side of the zone plus a buffer for the spread, never inside the box, because a zone by definition is an area where price is allowed to wander before reacting. If that stop is too wide for your account, the position is too big; the sizing arithmetic sits in the risk management rules.
A zone that fails is information, not noise. Price cutting straight through a fresh demand area on the daily chart is one of the earliest signals that the trend has changed hands, and it often precedes the clean breakout lower that everyone trades three hours later.
Where the method stops working
Supply and demand analysis assumes the orders that made the move were placed for reasons that persist. Around a central bank decision or a monthly employment release, that assumption dies. Rates reprice, the mandate that produced the buying is withdrawn, and a beautiful untouched zone from last week gets ignored completely. Treat any zone that formed before a major data release as expired.
It also produces nothing tradeable in a market that is quietly ranging inside a wide band, because every candle sits near an old imbalance and the reactions are too small to pay for the spread. Leveraged trading carries a high risk of loss, and a method that fires constantly is usually a method being used outside the conditions it was built for. Fewer zones, marked on higher timeframes, judged against the reward on offer, is the version of this that survives contact with a real account.
"A zone is not a magic price. It is the spot where someone once had more to buy than the market could give them, and the only question I care about is whether they still do."
— Alex Onta, Executive Director, SINGUARD
Key Takeaways
- A zone marks the origin of an imbalance, so draw it from the last opposing candle before the move and never redraw it afterwards.
- Quality comes from how the market left the area, how long the base took to form, and whether the zone has been visited since.
- Limit entries pay the best price and take every loser; confirmation entries give up part of the move to filter failures.
- Zones formed before a rate decision or major data release should be treated as expired, and a failed zone is an early sign the trend has turned.
Frequently Asked Questions
What is the difference between a supply zone and resistance?
Resistance is a single price line where selling has appeared before. A supply zone is an area, usually the small consolidation from which a sharp fall began, and it is marked as a band rather than a line. In practice the two often overlap, and traders who use zones simply accept that the reaction can start anywhere inside the band instead of at one exact price.
How many times can a zone be traded before it stops working?
Every visit consumes some of the orders that made the zone, so reactions usually get weaker with each test. Many traders treat the first return as the highest quality one and stop using a zone after two touches. A zone that has been tested repeatedly is better read as a level that is about to give way than as an area that will hold again.
Should supply and demand zones be drawn from the wick or the body?
The common convention is to draw the zone from the open or close of the last candle before the impulse to the extreme of the wick, which produces a band wide enough to contain the reaction without being so wide that the stop becomes unusable. What matters more than the convention is using the same one every time, so that results across many trades are comparable.