Three Weeks Tight: Reading Clustered Weekly Closes

A growth stock clears a base at 54.80, runs for six weeks, and then the weekly chart goes quiet. Three Friday closes land at 61.20, 61.65 and 61.05. On the daily chart the same three weeks look busy, with sessions that swing more than a point from high to low. A trader who already owns the stock, or who is studying it for a possible continuation, has to decide what that quiet weekly picture is worth. The three weeks tight pattern is the name for that cluster of closes, and most of the work in reading it lies in checking what surrounds it.

Every price, volume figure and percentage in this lesson belongs to hypothetical charts built so the arithmetic can be checked by hand. None of it describes a real stock, and none of the levels is a suggested entry or exit.

What the three weeks tight pattern measures

The pattern comes from the growth-stock chart work associated with William O’Neil and his study of past market leaders. Its subject is narrow. It looks at three consecutive weekly closing prices and asks whether they finished close together. That’s the whole observation.

Daily ranges are a separate question. A stock can trade across a wide band on Tuesday and Wednesday and still close Friday within a few cents of the prior Friday. The pattern only cares about where each week settled, so a narrow daily range on every session is a stricter test than the one the pattern actually sets. When I check a candidate, I write the three closes down on their own before looking at anything else. In the opening example that gives 61.20, 61.65 and 61.05, and those three numbers are the pattern. The bars around them are context.

The advance has to come first

A tight cluster means something only after a stock has already shown it can move. The conventional setting is an established uptrend, often a few weeks after the stock cleared a sound base such as a cup-with-handle. In the opening example the stock broke out at 54.80 and closed at 59.10 the week before the cluster began, so the first tight close at 61.20 sits 11.7% above the breakout level.

The same three closes inside a long sideways range, or in a stock trading below a declining 40-week moving average, describe a stock that has gone nowhere. They match the arithmetic and miss the setting. A cluster that forms after a sharp decline is also a different animal, because the stock hasn’t yet proven that buyers will lift it.

Successive closes or the whole cluster

The usual formulation in growth-stock literature puts each weekly close within roughly 1% of the prior week’s close. Some descriptions allow a little more room, around 1.5%, for stocks that normally move further. Treat those figures as conventions passed down through chart books rather than measured thresholds. There’s no study I can point to showing that 1% works better than 1.2%.

The comparison is between successive closes. Three closes give two comparisons:

• Week 2 against week 1: 61.65 / 61.20 = +0.74%
• Week 3 against week 2: 61.05 / 61.65 = -0.97%

Both pass a 1% test. The whole cluster also fits inside 1%, since the closes run from 61.05 to 61.65, a spread of 0.60, or 0.98% of the lowest close.

The two tests can disagree, and that’s where readers get confused. Take closes of 60.00, 60.55 and 61.10. Each step is under 1% (+0.92% and then +0.91%), so the successive-close test passes. The full spread is 1.10, or 1.83% of 60.00, which fails a whole-cluster test at 1%. A steady drift in one direction can pass the successive version while the stock is still walking higher. I’d call that a slow advance, and I’d want the article or screener I’m reading to say which version it uses. Some charting tools also compare the first of the three closes with the week before it, which makes the pattern effectively four closes long. You’ll want to check the definition before trusting a scan result.

Reading completed weekly bars

A weekly close is the last trade of the final session of the week, normally Friday. Until that session ends, the weekly bar isn’t finished. If the stock is sitting at 61.30 on Wednesday afternoon, that number tells you where the week is so far and nothing about where it will close. A pattern counted from an unfinished bar can disappear by Friday.

Holiday-shortened weeks still count as weeks, though a four-session week gives the price less time to wander, which can make a close look tighter than it would otherwise be.

The other measurement that matters is the highest price inside the structure, kept separate from the closes. In the opening example the three weekly highs were 62.10, 62.40 and 62.25. The top of the structure is 62.40, set in week 2, which is 1.22% above the highest close of 61.65. Traders who mark a reference level for the pattern conventionally take it from that intraweek high, slightly above the cluster, and the closing prices sit underneath it. If you draw that line at 61.65 instead, you’ve marked a level the stock already traded through during the pattern itself.

Two clusters with the same closes

Here are two hypothetical stocks whose weekly closes look almost identical. The difference sits in everything the closes leave out.

Stock A, the constructive pause:

• Weekly closes: 61.20, 61.65, 61.05 (steps of +0.74% and -0.97%)
• Weekly ranges: 60.30 to 62.10, 60.85 to 62.40, 60.40 to 62.25
• Range as a share of the close: 2.94%, 2.51%, 3.03%
• Weekly volume: 1.9 million, 1.4 million, 1.2 million, against a 10-week average of 2.1 million
• Setting: breakout at 54.80 six weeks earlier, no prior highs above the stock

Stock B, the superficially similar cluster:

• Weekly closes: 61.20, 61.75, 61.15 (steps of +0.90% and -0.97%)
• Weekly ranges: 57.90 to 64.80, 58.20 to 65.30, 57.40 to 63.90
• Range as a share of the close: 11.27%, 11.50%, 10.63%
• Weekly volume: 2.0 million, 2.6 million, 3.4 million, against the same 2.1 million average
• Setting: breakout at 48.90, so the first tight close sits 25.2% above it, and the week before the cluster closed at 66.40

On closes alone, B passes the same test A passes. Its whole cluster spans 0.60, or 0.98% of 61.15. The first thing I notice in B is the 7.10 range in week 2, from 58.20 to 65.30, which is more than eleven times the spread of its three closes. That’s a stock swinging hard within each week and happening to settle in the same place on Friday.

The rest of B’s chart leans the wrong way. The cluster started with a 7.8% drop from the 66.40 close, so the tight closes sit under a recent high instead of near the top of the advance. Weekly highs fell from 65.30 to 63.90 while weekly lows fell from 58.20 to 57.40. Volume rose through the three weeks and finished at 3.4 million, well above average, which is heavy trading for a stock making no net progress. And the highest price in B’s structure, 65.30, is 5.75% above its highest close of 61.75, so any reference level drawn from that high sits far from where the stock has been settling.

A screen built only on closing prices would return both stocks. Only one of them shows the quiet, contained trading the pattern’s supposed to describe.

What the tightness might reflect

The usual explanation is a temporary balance between supply and demand. After an advance, some holders take profits and some new buyers step in, and for three weeks those forces offset each other closely enough that the price settles near the same level each Friday. Quiet volume, as in Stock A, fits that reading because fewer shares are changing hands while the balance holds.

That explanation is an interpretation. Price and volume records can’t show who is trading. A cluster of tight closes is often described as institutions quietly holding or adding to positions, and the chart can’t confirm that. The same picture could come from a stock that’s simply lost the attention of most participants. Any claim about accumulation needs evidence from outside the chart, and even then it’s partial.

Context that decides the read

Traders who study the pattern tend to use it in two ways. A holder of an existing leader may read a tight cluster as evidence that the advance is resting in an orderly way. A trader studying a possible continuation may note the high of the structure as a reference level and watch whether the stock later trades above it on expanding volume. Some traders also discuss it as a point where an existing position could be added to under a plan like the one in pyramiding into a position. None of those uses is a signal by itself, and the lesson here makes no recommendation about entering or adding.

Five surrounding conditions do most of the work:

• Trend. Is the stock above a rising 10-week and 40-week moving average, with the cluster near the top of the move?
• Volume. Is weekly volume drying up during the cluster, as in Stock A, or expanding while price stalls, as in Stock B? The broader method is covered in reading volume.
• Relative strength. Is the stock holding its gains against the market during the pause, or is its relative strength line slipping?
• Nearby resistance. Does a prior high sit just above the cluster, the way 66.40 sits above Stock B?
• Distance from the earlier breakout. Stock A’s three closes sit 11.4% to 12.5% above its 54.80 breakout. A cluster 25% or more above the last proper base carries the risks discussed in breakout pivot price extension, because a pullback to the old base would be a long fall.

A tight cluster that fails two or three of these checks hasn’t passed much beyond the arithmetic.

How it differs from a flat base and a volatility contraction

The three weeks tight pattern is short by design. A flat base is conventionally described as lasting at least five weeks, with a correction from high to low of up to roughly 15%. It measures the full depth of a sideways structure, intraweek lows included. A three-week cluster with closes inside 1% can sit inside a flat base, or appear on its own during an advance, but the two describe different spans of time and different measurements.

A volatility contraction sequence is different again. It looks for a series of pullbacks that shrink one after another, for example a 21% decline, then 11%, then 5%, each measured from swing high to swing low. Its subject is the shrinking depth of successive swings over several weeks or months. Three weeks tight looks only at whether the last three weekly closes landed together, whatever the earlier swings did. A chart can show both, and a final tight cluster sometimes forms at the end of a contraction sequence, but one doesn’t imply the other.

Where tight weekly closes stop telling you anything

The limitations are worth as much attention as the definition, and they’re the part most often skipped.

Weekly observations arrive slowly. It takes three full weeks to confirm the pattern, and a lot can change in that time, and the chart won’t flag it until the next Friday close. A reader working from weekly closes is always looking at information that’s days old.

Quiet trading can reflect limited interest. Shrinking volume during a cluster fits the balance story, and it also fits a stock that traders have stopped watching. The chart doesn’t separate the two.

Apparent tightness can fail abruptly. A fourth week can gap below the cluster’s low, and a stock like Stock A could trade under 60.30 within a session. When the pattern breaks, it’s often quick.

A recognizable pattern also stays silent on company quality. It describes how a stock has traded for three weeks, while earnings, competitive position and valuation sit outside it, and so does any dependable future return. Reading it well means checking the advance before it, the intraweek swings inside it and the volume beneath it, and then accepting that the next weekly close can still undo the picture.

Learn the pattern. Ride the trend. Keep the gains.

Educational content only. Not investment advice. Trading involves risk. You are responsible for your decisions.