Retest of a Level: Why Price Returns to a Broken Boundary

A retest is a return of price to a level it has just broken, followed by an attempt to bounce off it from the other side. The point is not the pattern on the chart itself but the test it represents: a level that acted as resistance should now begin to act as support (and vice versa). If the bounce holds, the breakout is confirmed and the old boundary has switched roles. If price moves back through the level and settles inside the old range, there was no breakout. It was a false breakout, and the retest turned out to be not a confirmation but a return home.

What gets retested: a zone, not a line

A level is not a pixel-thin line but a zone where orders used to accumulate: previous highs, the edge of a consolidation, a round-number price, the price of a large placement or buyback. That is why a retest almost never comes back to exactly the same point. It comes back to an area, and the tolerance for an "undershoot" or an "overshoot" has to be set in advance. Otherwise every trade will either be missed or opened on a random tick.

The practical conclusion: mark the level using candle bodies and the boundaries where price actually spent time, not the extremes of isolated spikes. A spike is the trace of someone's stop order being taken out, not a market consensus on price. The basic mechanics of the move beyond the boundary itself are covered in a separate piece on breaking a level, and a retest cannot be read without that concept.

Where the return comes from

A return to the level is not chart mysticism but a consequence of the order structure. A breakout has participants who entered on momentum and are left without any cushion. There are those who held positions inside the range and are waiting for a chance to exit at breakeven, and their orders sit right at the boundary. And there are those who missed the breakout and want to get in at a better price. All these groups place orders in the same area, and price returns there not because "that is how it is done" but because that is where there is someone to trade with.

The bounce happens when, in the area of the level, those willing to continue the move outnumber those wanting to push price back. Hence a simple corollary: a retest that unfolds slowly and in a narrow range reads differently from a retest that slices straight through the level with a wide candle.

How to tell a confirmed retest from a failed breakout

The distinction rests on how price behaves inside the zone, not on the fact that it touched it.

Step 1. Mark the boundary using closes, not wicks. Step 2. Wait for price to return to the zone and watch what it does there: whether it contracts and slows down, or passes through the zone without pausing. Step 3. Define the invalidation price in advance, the level beyond which the idea stops being valid. For a retest, that is price settling on the far side of the zone, not a single spike.

What a retest does not account for

A level is a memory of past trades. It loses its force when the market receives new information: financial results, a dividend decision, a corporate event, a change in the structure of the security. After such events, price can move through the old zone with no regard for it at all, simply because a different set of expectations is now being traded.

So before taking a trade on a retest, check whether it falls on a date that moves the price by itself. Upcoming corporate dates are collected in the events calendar, payouts in the dividend calendar, and earnings releases in the issuer reports section. A retest scheduled for a record date or a results release is not a technical setup but a bet on the content of the news.

A separate case is a change in the formal trading conditions of a security. When the listing level changes, so does the set of participants who have access to the security; the old levels should then be treated as outdated until a new trading history builds up.

What to do with it

A retest is a tool for defining an entry point and a point of failure, not an argument in favour of a direction. The direction has to come from a separate reason: the trend structure, the news flow, the valuation of the security. The instrument's liquidity matters too: in a thinly traded security the level zone blurs, and a retest turns into noise. You can check liquidity and the current market structure in the Russian market stocks section, and the background events in the news. Terminology on the subject is collected in the glossary.