Consolidation describes bounded price behaviour on a chosen interval. Range trading hypothesises that a boundary approach may reverse; the description alone does not justify an order. First identify the observations, costs and conditions for rejecting the idea.

What trading a consolidation actually means

The same episode can look sideways on H1 and be a pause in a daily trend. Record instrument, source, price side, interval and observation window. Every finite set of Highs and Lows makes a rectangle, including a trend; examine the path, not only width. Its past appearance cannot tell whether the next bar stays inside.

A range and a breakout rule select different conditions; a defined switch or combined plan is another version to test. The mean-reversion approach explains the hypothesis, while systematic range trading separates qualification and execution. The technical analysis guides provide broader context; none establishes profitability from a style label.

Range width does not locate the closes

One extreme wick can widen a rectangle without dispersing closes. Check what your measure describes.

Scroll sideways to see the full diagram

A long wick, unchanged closesCandles have Low 99, Open 100 and closes 100 or 101. The third has High 120 instead of 102. Extreme range rises from 3 to 21; closing-price spread stays 1.ONE WICK, UNCHANGED CLOSES99102120Without long wick: High −Low =3.After wick: High −Low =21.Close spread:101 −100 =1 in both.

Fictitious model on one shared scale: all Open 100 and Low 99; closes 100,101,100,101,100. Initially every High 102; then only the third High becomes 120. This comparison predicts no continuation of the range. TradingView — Repainting confirms that OHLC are separate candle components, and High, Low and Close may change while a bar develops.

Five fictitious closes are 100,101,100,101,100. Only the third High rises from 102 to 120.

max(High) −min(Low) increased sevenfold. max(Close) −min(Close) did not change. Both are valid but answer different questions: extreme candle span versus closing-price spread in this sample. Do not switch measures after viewing the chart merely to obtain a convenient range. Record the measure and retain the long wick unless a predefined rule says otherwise.

Is this really a range, or just a pause in a trend?

For a reproducible range description, save bounds when they become known, define contact tolerance and rearming, and check whether one extreme wick dominates the width. A small set of bars or a third bounce is not a validated safety gate.

  1. Directional-strength filter. ADX measures trend strength, not which side has the upper hand. Below 25 is a commonly used weak-trend filter, but the threshold is not a guarantee of a range; a rising reading can also reflect a downtrend. Define and test the chosen threshold with the price structure.
  2. Same width, different paths. Our fictitious five H1 Bid closes A are 1.0808,1.0818,1.0802,1.0817,1.0812; B are 1.0801,1.0805,1.0810,1.0815,1.0819. Separately assume both have highest High 1.0820 and lowest Low 1.0800: 20-pip width. A changes direction and ends +4 pips; B’s successive closes all rise and end +18. The closes alone did not supply those extrema or a trade.
  3. Distinct contact episodes. For our own lower bound 1.0800, count a Close within 2 pips above it as a contact, but rearm a new episode only after a Close more than 5 pips above. Distances 1,2,0,7,1,8,2 give five contact bars but three episodes. This numbering does not establish statistical independence. The support/resistance guide helps define bounds.

A quiet calendar, holiday month, pair name or Asian session label does not establish a stable range or predictable costs. Compare actual matched data before claiming that one instrument or season gives more opportunities; no such frequency study is supplied here.

Where to open the position — at the edge, but with a margin

Distinguish entry placement from stop placement. Buying at support does not by itself cause a loss on a five-pip dip; the result depends on where the protective stop sits and how it fills. A confirmation entry inside the range gives up some distance to the target and changes entry-to-stop risk. Define both prices and size the position from their actual distance.

A boundary touch or closed candle is an observation, not proof that identifiable buyers stepped in. Define any pattern or RSI condition and when it becomes available. A low/high oscillator reading does not automatically increase bounce probability. Later confirmations can alter price, selected cases and costs.

Stop loss and target — a worked example

Hypothetical scenario — EUR/USD H4, range 1.0800–1.0900, long position
Entry (15 pips above support, after a confirming candle)1.0815
Stop loss (15 pips below support, i.e. outside the range)1.0785 — 30 pips of risk
Target (5–10 pips ahead of resistance, here 10)1.0890 — 75 pips of potential
Reward to risk75 against 30, or 2.5 to 1
Invalidationa chosen closed-bar retirement condition must be defined separately; it is not a guaranteed lasting breakout

The separate H4 table assumes actual Ask entry 1.0815, planned Bid stop 1.0785 and target Bid 1.0890. Its 30/75-pip distances and gross ratio 2.5 are arithmetic, not recommendations. For 10,000 EUR with hypothetical USD 2 round-trip commission and no other fees, exact-fill net outcomes are −USD 32 or +USD 73: net ratio 73/32≈2.28. Ordinary stops can execute worse. No universal 10–20-pip buffer or 1% size rule makes loss small or bounded.

A separate cost screen assumes Bid moves 1.0807→1.0814, 7 pips. For 10,000 EUR, Ask entry 1.0808 with 1-pip spread gives USD 6 gross / USD 5 net after USD 1 commission; Ask 1.0811 with 4-pip spread gives USD 3 / USD 2. Financing and conversion are zero assumptions, and spread is already in the fill sides. The same Bid move does not mean the same net result, or prove the exit will occur.

Support and resistance are candidate areas inferred from past prices. A range plan needs an explicit exit if those boundaries fail.

When to stop immediately

Define when to retire new-entry qualification rather than infer an irreversible end from ADX or a candle shape. A declared Close outside the bounds can disable a model even if price later returns. Save the state change and any restart rule in advance.

Existing-position handling is separate from finding a new range. Keep the predeclared exit and actual order confirmations; an active stop does not wait for the bar to close. Do not expand old bounds after seeing a break merely to preserve the original interpretation.

The most common mistakes

  1. Undefined interval. H4 can be sideways while D1 rises. Those descriptions are compatible, not proof that one plan must trade against the other.
  2. Assumed fixed protection. No five- or 20-pip distance determines whether a stop will trigger or how it fills. Calculate monetary distance and test the selected rule.
  3. Holding a losing position after a breakout. "Price will come back" is sometimes true, but it is no longer the trade you opened.
  4. Whole width mistaken for reward. Measure from actual entry to exit and add fees once. Stop distance describes loss, not another fee to subtract from the entire range.
  5. Increasing risk after repeated tests. Repeated touches do not establish a predictable countdown to a breakout. Reassess the price structure and keep position size within the predefined risk budget rather than assuming either that a level must hold or that it must fail next.

What to do tomorrow

  1. Save comparable input data. Specify provider, interval, quote side and window. Examine sequential closes and extremes rather than select a pair from an unmeasured quietness ranking.
  2. Define boundaries and contact episodes. Save tolerance, rearming and determination time. No three-touch or 100-pip minimum follows from the concept.
  3. Record a complete model before evaluation. Separate entry condition, actual sides, quantity, costs and exits. Indicator agreement does not prove independent confirmation.
  4. Separate retirement and execution. Define how new-entry qualification ends, how an open trade exits and what would permit a new model; a few sessions of waiting has no universal efficacy.
  5. Check event and product conditions. Record release times, provider terms and observed spreads. No scheduled report or narrow quote guarantees a calm period or realised profit.