Zones, Breakouts and Retests

Why a level exists, and how to grade it

Levels are not drawn, they are remembered. Participants who bought at a price cluster orders around it: stops parked just below an old low, bids at a base that held, offers where a rally died, and hedges that reference a round number. That clustering is why price tends to respond at the same place twice, and it is also why the response degrades with time — positions get closed, books rotate, and the people who placed the original orders move on. A level is therefore a claim about orders, and claims can be graded. Five qualities do most of the grading. Repetition: a level that has turned price three times carries more evidence than one touched once. Recency: the most recent test matters, because the orders are more likely to still exist. Distance: a level that produced a long, sharp rejection is stronger than one price brushed past. Volume at the level: heavy participation at the turn is evidence of real size, while a level formed on thin trading is a description of a quiet week. Roundness: $50, $100 and index levels carry psychological clustering, which makes them weaker as a reference for something new and stronger as a place where orders sit. All five are observable before the trade, which is what makes level reading a discipline rather than a taste. The operative form is a zone rather than a price. A level built from three reversals inside a 4% band is a zone 4% wide, and a learner who demands an exact touch will miss entries while catching false precision. That has a practical consequence for stops as well: a stop placed at the exact level is inside the band where the level was built, so it tends to be hit by the same clustering it was meant to sit outside — which is why structural stops are placed a little beyond the zone, and why the sizing arithmetic from TR3 has to use that distance rather than the nominal one. The same level, two expressions — Breakout chase at $50.40, stop below the candle at $48.90: Risk $1.50 = 3.0% for a $2.40 run to the target · Retest entry at $48.60, stop below the level at $47.40: Risk $1.20 = 2.5%, and the same target pays 3.5R ← · Stop at the exact level ($48.20): Inside the zone that built the level — hit by clustering rather than by a break · Level as a zone, $47.90–$48.50: Order placement and stops reference the band, not the line ← Polarity is the useful half of the same idea: overhead resistance that is decisively broken usually becomes support, because the participants who sold there now defend the level they are watching, and the buyers who missed the break use it as an entry. That is why the retest is not just a second chance but the market’s audit of whether the break was real.

The retest is the audit, and the chase is the tax

A close above a level on confirmed volume is evidence, and evidence is not the same as a completed move. The retest is the return to the broken level from above, and its outcome is the cleanest available test of whether the break was repricing or a trap: if sellers cannot push price back through, the level holds as support and the thesis is intact; if they can, the break was a failed auction and the position should not exist. That is why the honest version of waiting costs you something — the trades that never look back — and buys you a defined stop on everything else. Then the arithmetic that makes the wait worth it. Chasing the breakout means the stop has to sit below a long, fast candle, so the entry-to-stop distance is large; buying the retest means the stop sits just beyond the level, so the same risk per share buys more shares for the same budget and the same target pays a larger multiple of risk. Nothing about the thesis changed between those two entries — only where the stop lives, and therefore only the geometry. A learner who computes the R multiple for both will find the difference is usually a factor rather than a percent, and that factor is the reason patient traders report better results on the same analysis. What the retest does not do is protect you from a range. Inside a range, every level is being tested constantly, so breaks and retests fire repeatedly and each one is a coin flip with costs attached — the regime filter from TR14 is what keeps the retest method from being applied where it does not work. And a retest that holds but then fails on the next session is a failed audit rather than a slow entry: the level was reclaimed, then given back, and the second break is the information that matters. • A break is judged on the close, not on an intraday excursion. • The retest is the market auditing the break: hold is confirmation, failure is a trap. • Waiting costs the moves that never look back and buys a defined stop on the rest. • Chasing puts the stop below a long candle; the retest puts it beyond a level — same thesis, better R. • Polarity: broken resistance tends to become support, which is why the retest is tradeable at all. • In a range, every level is tested constantly, so the method needs the regime filter. Do not average a stop further away to keep a chased position alive. A learner who buys $2.20 above the level and then moves the stop down because the first stop was hit has combined the worst of both: the chase entry and the retest’s risk, without the retest’s evidence. When a chased breakout fails, the information is that the break was not real.

Every chart has a hundred levels

A price chart is dense with places a line could be drawn — prior highs, prior lows, round numbers, the open, the close, moving averages, the previous day’s range. With enough candidates, some will appear to “work” purely by chance, and the eye finds them after the fact with impressive confidence. The problem is a multiple-comparisons one: the more levels you consider, the more likely at least one looks predictive in any sample. The repair is to define the level before you look at what happened, and to make the definition mechanical. “The high of the last twenty sessions” is a level; “a resistance area around $48” is a story told after the fact. Once the definition is fixed, a level can be scored the way any pattern is: how often price reacts at it, and whether that frequency is distinguishable from what a random level of the same age would produce. Most tested levels are not, and that is itself a useful finding. The trading consequence is that the *reason* a level might matter should come from something other than the fact that price turned there — a cluster of prior trades, an option strike with open interest, a round number where orders concentrate, a gap that left a region untouched. That is why order-flow and open-interest data make chart levels more than lines. A level with a mechanism behind it is a hypothesis; a level with only a memory is an artefact. If you drew the zone after the price stopped there, you have not found support — you have found a reason to have bought.

Why a strike price acts like a level

A level matters because participants hold positions around it, and the cleanest place to see that is the options market, where the levels are not drawn on a chart at all. Option strikes are fixed prices set by the exchange, and the open interest at each strike is published daily — which makes them the rare kind of level whose population is measurable rather than inferred. The mechanism is dealer hedging, and it is worth getting straight because it explains an observation that otherwise looks like superstition. A market maker who has sold calls is short those calls and therefore long the underlying, and it must sell stock as the price rises toward and through the strike to stay hedged. The same maker short puts must buy as the price falls. Hedging a short-option book therefore trades *against* the price move — buying weakness and selling strength — which damps movement and produces the pinning effect that leaves a stock hovering at a large round strike into expiry. When the maker has instead bought options from customers, the sign flips: the hedge chases the move and amplifies it. The difference between the two regimes is the sign of the market’s net gamma, and it changes what the same chart level does. Three readings follow, all from data that is freely available. The **open interest by strike for the nearest expiry** shows where the positions are concentrated. The **distance between the current price and the largest strikes** tells you whether a pin is plausible. And the **concentration of expiries** tells you when the effect decays: pinning pressure disappears after expiry, which is a plausible mechanical explanation for the tendency of a range to break shortly after a large expiry passes. The same logic reaches beyond single-name options. Leveraged and inverse funds must rebalance daily to maintain their stated exposure, which means buying into strength and selling into weakness in mechanical amounts; structured products with knock-in levels create flows at levels defined by a contract rather than by a chart; and index-option positioning concentrates flows at round index levels. Every one of these is a level with a participant attached to it, which is the standard this lesson uses to grade a line. Two honest qualifications keep it useful. The effect is real and modest, and it disappears when genuine information arrives — a level does not survive an earnings report. And it is a description of flows around a price, not a prediction: the strike tells you where the market has business, not which way it will go when that business is settled. Used as the strongest available answer to “which of these hundred lines matters”, however, it beats every line that was drawn because it looked clean. • Strike-level open interest is a level with a measurable population of positions attached. • Short option dealers hedge against the move and pin price; long-gamma books amplify it. • Pinning pressure ends at expiry, which is when ranges often break. • Leveraged funds and structured products create flows at contract levels, not chart levels. The practical lookup before grading any chart level: check the open interest at the nearest strikes around the price and the date of the next expiry. A round number with large open interest into a nearby expiry is a different object from a round number with none.

Sources

• Support, resistance and level formationSchwager, “Technical Analysis” (practitioner interviews)
• Breakout failure and retest studiesBulkowski, “Encyclopedia of Chart Patterns”
• Order clustering and stop placementMarket microstructure literature on liquidity at visible prices