Support and resistance explained: a framework for levels that mean something

Support and resistance are not exact prices. They are areas where market participants previously made a visible decision: they accumulated, distributed, covered risk, or accepted a new value range. Treating every prior candle high as a level creates clutter; the goal is to find the areas where behaviour changed.

Start with structure, not indicators

Mark the highs and lows that caused a clear expansion, rejection or change in trend. Then move one timeframe higher and ask whether the same area remains visible. A level that exists only on a very short chart may still be tradeable, but it should not carry the same weight as a weekly or daily decision zone.

The practical filter is simple: if you cannot say what happened at a level — a session that opened and never traded back, a release that repriced the range, an obvious failure and reversal — it is probably a coincidence of two touches rather than a place anyone made a decision.

Build levels from auction behaviour

The most useful zones often originate where price moved rapidly away after a period of balance. That departure indicates an imbalance between willing buyers and sellers. On a return visit, the question is whether the original imbalance still exists. Slow overlap and repeated trade through the zone suggest acceptance; sharp rejection with follow-through suggests participants still defend it.

A zone should be bounded by the range of the decision rather than drawn as an artificially exact number. Width depends on volatility and timeframe: a one-minute chart may justify a narrow band, a daily chart needs considerably more. Writing the zone and the required confirmation in advance makes it far less tempting to reinterpret every wick after entry.

Reaction is different from acceptance

A wick through a level demonstrates that orders traded there. It does not prove the market accepted the new price. Acceptance is a process: price spends time beyond the level, trades volume there, and holds when it retests. Rejection is also a process: price probes, fails to sustain the move, and returns into the prior range.

• Use zones rather than a single pixel-perfect line.
• Give more weight to levels that align across timeframes.
• Note whether an event or a high-volume session created the level.
• Define in advance what price action would prove the level has failed.

The level types worth marking, ranked

Not all levels are the same kind of object, and they do not deserve equal weight. Some are records of a decision, some are reference prices that large numbers of participants watch, and some are pure coincidence.

Round numbers earn a mention because dismissing them is as wrong as worshipping them. There is nothing structural about 100,000 — but stops, take-profits and alerts cluster there because humans set them there, and clustered orders are the entire mechanism by which any level matters. The correct weight is small and non-zero.

The two categories at the top share a property worth naming: both record a moment when one side was unable to transact at the price it wanted. That is what makes them likely to matter again, and it is why a level with a story attached consistently outperforms a level found by scanning for touches.

Keep a level ledger

Four fields make a marked level reviewable rather than decorative: its origin, its timeframe, its current state and the decision it informs.

1. Origin: which session, event or expansion created it?
2. Timeframe: why does this zone matter on the chart being traded?
3. State: untouched, tested, accepted through, or rejected?
4. Decision: what would a successful or failed retest mean for the plan?

Worked through: a daily range high at $104.20, created by the session that followed a scheduled release, retested twice and rejected both times, is a level with an origin, a state and a decision attached. If price closes above $104.20 and then holds above it on a retest, the range thesis is finished and the next zone becomes relevant; if it wicks through and closes back inside, the range is intact and the failure itself becomes the reference point.

Keep the chart decision-ready

Every marked level should answer a question: where might I expect a reaction, where is my thesis invalidated, or where could a breakout be confirmed? If it does none of those, remove it. A small number of well-defined zones makes it easier to act consistently and to review the result honestly — and it makes the invalidation level obvious rather than arbitrary.

No level is permanent. The more times an area is tested, the more its resting liquidity and participant mix change. Reassess after major news, after a regime change in volatility, and after any sustained period of trade on the other side. What separates a level that failed from one that was never valid is covered in false breakout versus real breakout.