Support, resistance, and why levels work at all

Draw a horizontal line on a chart where price turned around twice. Price comes back to it and turns around again. Why?

The honest answer is not "because it is a support level." That is a description, not a cause. The cause is about people with positions and memories, and once you see it, you can tell the difference between a level that matters and a line you happened to draw.

The mechanism#

Three groups of traders create pressure at a price where something significant happened before.

Buyers in profit who add. Buyers in loss who sell to escape. Watchers who missed it and swore they would not twice.

All three act at roughly the same price. That concentration of intent is what a level is. Not geometry — clustered memory.

What makes a level strong#

Not all levels are equal. Four things make one worth respecting:

Number of touches. A price rejected three times has more history than one rejected once. But be careful — see the caveat below.

Volume at the level. Heavy volume means many positions were opened there. Many positions means many people with a memory of that price.

Recency. A level from three weeks ago is more alive than one from two years ago. Participants change.

Reaction size. A level that produced a violent reversal matters more than one that produced a pause. The size of the reaction tells you how much pressure was sitting there.

Round numbers do real work#

24,000 on Nifty is not technically special. There is no structural reason it should matter.

It matters anyway, because humans place orders at round numbers. Stops sit at 24,000, not 23,987. Targets are set at 24,500. Option strikes cluster there, which concentrates hedging activity around them.

A round number becomes a level through nothing more than collective habit — and that is entirely sufficient. The crowd's behaviour is the mechanism, not a proxy for one.

Broken support becomes resistance#

This is the flip, and it follows directly from the mechanism.

Price breaks below a support level. The people who bought at that level are now trapped in a loss. When price rallies back to it, they sell to get out flat — and that selling turns the old floor into a new ceiling.

The level did not change. The positions around it did. Understanding this means you can predict the flip rather than memorising it as a rule.

Levels are zones, not lines#

Beginners draw a level at exactly 24,017.35 and are confused when price trades to 24,024 before reversing.

Memory is not precise. Nobody remembers 24,017.35; they remember "around 24,000." So a level is a zone, usually a few tenths of a percent wide, not a line.

This matters practically. A stop placed one point beyond a level will be taken out by ordinary noise around it. Place stops beyond the zone, and accept that the zone has width.

When levels stop working#

Levels are contextual, and there are conditions where they fail systematically:

In a strong trend, levels in the trend's direction get cut through. The pressure from the trend outweighs the pressure from memory.

On news, prior levels are largely irrelevant. The information that created them has been replaced.

After a long time, the participants have turned over. A level from years ago is a historical fact, not an active force.

And there is a genuine tension in the touch-count rule. More touches means more history, but each test also consumes the orders sitting there. A level tested five times has had most of its pending supply absorbed — which is why heavily tested levels often break. Both effects are real, and which dominates depends on volume and context.

Check yourself

0 of 4 answered

1. 1.Price breaks below a well-established support level, then rallies back to it and stalls. Why?

   The level has magnetic propertiesTraders who bought at that level are now trapped in a loss and sell to exit flatAlgorithms are programmed to sell thereIt is a coincidence with no mechanism behind it

2. 2.Which level is most likely to produce a reaction?

   A line drawn through an obscure intraday wick that only you noticedAn obvious prior swing high at a round number, on heavy volume, from last weekA level from four years agoA level derived from a complex indicator calculation

3. 3.You place a stop one point below an obvious support level. What is the likely problem?

   The stop is too far from entryLevels are zones, and everyone else's stops are clustered in the same place, making it a targetStops cannot be placed below supportIt will not be filled at all

4. 4.A support level has been tested five times. What does this tell you?

   It is extremely strong and unlikely to breakIt is guaranteed to break on the sixth testIt has more history, but each test also consumes the orders there — heavily tested levels often breakThe number of tests is irrelevant

What to take away

• Levels work through clustered memory and trapped positions, not geometry.
• The best levels are the obvious ones, because more people can see them.
• Round numbers matter purely through collective habit — and that is enough.
• Broken support becomes resistance because the positions flipped.
• Levels are zones with width. Place stops beyond the zone.
• Obvious levels attract stop clusters and are often probed before reversing.
• Repeated tests both strengthen and consume a level.