Support and resistance:
which levels matter and which are drawn noise
Anyone can draw a line on a chart, and once you have drawn six, price will bounce off one of them and you will believe in lines.
The useful question is not where the levels are. It is why any level would work at all, and that answer immediately tells you which five are worth marking and which forty are decoration.
Why levels work at all
There is nothing mystical about a horizontal line. Levels work for three mechanical reasons, and every reason is about where other people have orders.
Memory of transacted price
If a lot of volume traded at 20,150 this morning, a lot of people have positions with that as their entry. When price returns, those people act: some take profit, some get out at break even, some add. That concentration of decisions at a specific price is what produces a reaction.
Stop orders cluster at obvious places
Everyone who bought the low puts a stop just under it. Those stops are resting sell orders sitting in a known location. This makes obvious lows a target rather than a fortress, which is the opposite of what beginners assume.
Everyone is looking at the same chart
The previous day high is visible to every participant. A level being widely watched is a self reinforcing reason for it to matter, and it is why round numbers and session extremes work better than a line you drew from a fractal nobody else can see.
The consequence
A level is useful in proportion to how many people can see it and how many orders sit there. That single criterion eliminates most of the lines on a typical beginner chart, because they are levels only that trader can see.
The five that matter intraday
Ranked by how reliably they produce a reaction. Marking these five takes two minutes and replaces most of what people draw.
Previous day high and low
The most watched levels on any intraday chart. Every participant can see them, they represent the extremes of a full session, and they frequently act as targets before they act as barriers.
Overnight high and low
The extremes of the session before the cash open. Particularly relevant in the first hour, because the overnight range often gets swept in the first fifteen minutes.
The opening range
The high and low of the first fifteen or thirty minutes after 9:30. High participation makes these meaningful for the rest of the session. The opening guide covers how to use them.
Today high and low, as they form
Dynamic, and they update. The current session extremes are where stops accumulate through the day, which makes them the most common sweep targets in the afternoon.
Round numbers
Whole hundreds on an index. They matter because humans place orders at them, not because of anything structural. That is a sufficient reason.
| Level | Why it matters | When it is most active |
|---|---|---|
| Previous day high and low | Watched by everyone, full session extremes | All day, especially as targets |
| Overnight high and low | Pre open extremes, thin liquidity above and below | First 30 minutes |
| Opening range | Highest participation period of the day | 10:00 onwards |
| Session high and low | Where stops accumulate as the day runs | Afternoon |
| Round numbers | People place orders there | Anytime, weakest of the five |
If a level is not on this list, ask who else can see it. Usually the answer is nobody.
How to draw them in two minutes
Before the session, not during. Drawing levels while a trade is open produces levels that justify the trade.
The routine
- Open a daily or hourly chart. Mark the previous day high and low.
- Mark the overnight high and low from the futures session.
- Note the nearest round numbers above and below current price.
- After 9:45 or 10:00, add the opening range high and low.
That is four to six lines. If your chart has more than eight, you have built a grid where everything is near a level, which means nothing is.
Use the same chart every day
Switching timeframes to find a level that supports your idea is one of the most reliable ways to lose money, and it is almost invisible from the inside. Fix the timeframe, draw the same five levels, every session.
Do not delete levels that failed
A level price sliced through is still information: it tells you the session is trending and that levels are not holding today. Erasing it removes the evidence that would have stopped you fading the next one.
Mark them, then leave them
Levels are drawn before you have a position. Once you are in a trade, your judgement about where the line should go is no longer independent.
Lines are zones, not prices
This is the correction that saves the most money, and it is a small change in how you think.
Price does not respect a number
The previous day high is not 20,150.00 exactly. It is an area of a few points where the transition happened. Treating it as an exact price means placing a stop one tick beyond it, which is the single most commonly hit location on the chart.
How wide is the zone
Proportional to current volatility. If the last ten candles average twelve points, the zone is several points wide. At midday with three point candles it is much tighter. This is the same principle as reading candle size relative to recent candles.
What this changes practically
Your stop goes beyond the zone rather than beyond the line. That means a wider stop, which means a smaller position for the same dollar risk. Traders who keep size constant and tighten the stop to the exact line are choosing to be stopped out by noise.
The signal you are looking for
Not a touch of the price. A reaction at the zone: a rejection wick, a failure to close beyond it, a candle that closes back inside. The reaction is the evidence, the line is only where you looked for it.
Why support becomes resistance
The most quoted idea about levels, and worth understanding mechanically rather than as a rule.
What actually happens
Price breaks below a level that had been holding. Everyone who bought at that level is now losing. When price comes back up to it, a large number of those people are looking to exit at break even rather than take a loss. Those exits are selling pressure arriving at exactly that price.
Why it is not automatic
The flip requires that people were actually positioned at the level. If price barely touched it before breaking, there is nobody trapped and there is no flip. A level that held three times and then broke produces a strong flip. A level touched once does not.
How to judge it
Count the reactions before the break. More tests means more trapped participants means a stronger retest. This is also why a level that has been tested four times is more likely to break: each test consumes the orders defending it.
The apparent contradiction
Many tests make a break more likely, and also make the flip stronger once it happens. Both are true and they are the same mechanism from different sides: orders get used up, then the people who placed them become the fuel for the move the other way.
Break, sweep or failure
Price going beyond a level is three completely different events that look identical for the first few seconds. Telling them apart is most of what level trading is.
The sweep
Price pokes beyond the level, triggers the stops resting there, and returns inside within the same candle or the next one. The wick is long, the close is back inside. This is not a break. It is the level being used as a source of liquidity, and it frequently precedes a move in the opposite direction.
The genuine break
Price closes beyond the level and the next candle also closes beyond it, often with an increase in range. Participants have accepted the new price. Fading this because the level was supposed to hold is a common and expensive mistake.
The failed break
Price closes beyond, appears accepted, then the following candle closes back inside. Now the traders who entered on the break are trapped, and their exits provide fuel. This is the highest quality of the three setups, because you know specifically who is wrong and what they must do about it.
How to tell them apart without guessing
Wait for the candle to close, and then wait for one more. That costs you a few points of entry and eliminates a large fraction of false signals. The liquidity guide covers why sweeps happen so consistently at obvious levels.
| Event | What the candle does | What it means |
|---|---|---|
| Sweep | Long wick beyond, closes back inside | Stops collected. Often reverses |
| Genuine break | Closes beyond, next candle holds | Price accepted. Do not fade it |
| Failed break | Closes beyond, next closes back inside | Traders trapped. The best of the three |
All three look the same for the first thirty seconds. The close is what separates them.
How to actually trade a level
A level is a place to look, not a signal. Four steps turn it into a decision.
One: decide the context before price arrives
Is the session trending or ranging? In a range, levels hold more often and fading them works. In a trend, levels get sliced and fading them is how you lose repeatedly to the same move. Decide this before you have a position.
Two: wait for a reaction, not a touch
The level is where you start paying attention. The trade needs evidence: a rejection wick, a failure to close beyond, a failed break. Entering on the touch is entering on the assumption that the level will work.
Three: put the stop beyond the zone
Not one tick past the line, because that is where everyone puts it and it is the most efficiently hunted price on the chart. Beyond the zone, sized to current volatility, with position size reduced accordingly.
Four: know where you are going
The natural target of a trade from one level is the next level. That gives you a reward to risk number before you enter, and if it is below one, the trade is not worth taking regardless of how good the level looks. The risk guide covers why that ratio decides everything.
The five mistakes
These account for most of the money lost trading levels.
- Too many lines. Fifteen levels on a chart means price is always near one, which means the levels are not filtering anything. Five is enough.
- Stop one tick beyond the line. The most predictable stop location in existence. Put it beyond the zone and reduce size instead.
- Entering on the touch. A touch is not evidence. The reaction is the evidence, and it takes a candle to form.
- Fading levels in a trend. In a trending session levels get broken all day. Fading each one because it worked in a range yesterday is the most consistent way to lose a trending day.
- Redrawing levels after entering. Once you have a position, your judgement about where the line belongs is compromised. Draw before, not during.
Four of the five are solved by preparing before the session and waiting for a candle to close. That is not sophisticated, and it is most of the difference.
Frequently asked questions
What is support and resistance in trading?
Price areas where a reaction is more likely because orders concentrate there. They work for three mechanical reasons: a lot of volume transacted at that price, stop orders cluster just beyond obvious highs and lows, and the level is visible to every participant.
Which support and resistance levels are the most reliable?
The previous day high and low, the overnight high and low, the opening range, the current session extremes and round numbers. All five share one property: every participant can see them, which is what makes a level matter.
Where should I put my stop when trading a level?
Beyond the zone rather than one tick past the exact line, and size the position down to keep the dollar risk constant. One tick past the line is the most predictable stop location on the chart, which is exactly why sweeps target it.
What is the difference between a breakout and a liquidity sweep?
A sweep pokes beyond the level, collects the stops resting there and closes back inside, usually leaving a long wick. A genuine break closes beyond and the next candle holds. They look identical for the first seconds, which is why waiting for the close matters.
Why does support become resistance after it breaks?
Because the people who bought at that level are now losing, and when price returns they exit at break even. Those exits are selling pressure arriving at exactly that price. The effect is strong only if people were actually positioned there, so a level tested several times flips harder than one touched once.
How many levels should I have on my chart?
Four to six. Beyond eight, price is always close to a line, so the levels stop filtering anything and start justifying whatever you already wanted to do.
Sources
- Osler, Support for Resistance: Technical Analysis and Intraday Exchange Rates (Federal Reserve Bank of New York).
- Osler, Currency Orders and Exchange Rate Dynamics: how order clustering produces reactions at round numbers.
- Admati and Pfleiderer, A Theory of Intraday Patterns (1988): where volume and liquidity concentrate.
- CME Group, Micro E-mini Nasdaq-100 contract specs: tick size for zone width.
- Nasdaq, market activity data: intraday volume distribution.
A level is where you look, not what you trade
Three minute rounds with sweeps, failed breaks and real costs. Learn the difference where being wrong costs you a round.
Play a tournament