Support and Resistance: Finding the Levels That Matter

What would have to be true for a line on a chart to matter? Only one thing: orders sitting at that price that nobody has filled yet. Everything else follows from one question. Each time you draw a line, ask whether that claim is still plausible.

Three mechanisms put orders at a price

Memory of fills comes first. Traders who bought at $42 and watched a rally to $48 will consider $42 again, and traders who wanted in at $42 and missed will leave a bid there. Both behaviours concentrate demand at a remembered number.

Breakeven selling is the mirror. Traders who bought at $48 and rode the stock down to $42 spend the entire recovery waiting to get out at what they paid, and that supply is what makes an old high hard to clear on the first attempt. Resistance is often thickest exactly where the last rally died.

Then there is the arbitrary one. Limit orders, stop orders and option strikes all cluster at whole dollars and at multiples of ten, for the plain reason that people pick round numbers when they are asked to name a price. More resting size sits at 50.00 than at 49.63. No fact about the company is involved.

How wide is the zone

Price does not turn at a number. Take four reactions in the same area. The lows are 41.78, 42.10, 41.92 and 42.31. A line through any one of those is false precision. A stop at 41.75 will be removed by ordinary noise.

Mark the zone 41.78 to 42.31. Its width is 42.31 - 41.78 = 0.53, which is 0.53 / 42.00 = 1.3% of price. That percentage is the number you carry forward, because it is the minimum room your stop has to allow if you intend to trade the level at all. A stock whose zones are 1.3% wide and a stock whose zones are 5% wide are not the same trade with a different ticker, and sizing them the same way is how an account bleeds without any single decision looking wrong.

Every other technical tool on this site has a formula. This one does not. There is no calculation that produces 41.78 to 42.31 from the data, which means the discipline has to come from somewhere else, and in practice that means drawing the zone before you have a position and writing the numbers down.

Rank the levels, and keep very few

Twelve lines on a chart guarantee a line near any price you look at. That makes all twelve worthless. Rank instead, on three things you can check: how many separate turns occurred in the area, how far apart in time those turns were, and how much volume traded on the reactions.

Evidence for the zone	Stronger version	Weaker version
Number of turns	Four distinct reversals	One wick
Spacing in time	Turns months apart	Four touches in ten sessions
Participation on the turns	Reversal on 3.0 times average volume	Sideways drift on 0.7 times average

A zone that produced a sharp reversal on three times normal volume is a stronger claim than a zone where price drifted quietly for two days. The volume analysis guide covers how to measure that participation properly.

Use daily bars to find levels even if you trade shorter timeframes. The levels that matter are the ones many people can see. Most people are looking at the daily chart.

Trend lines are the weakest version of this

A trend line connects rising lows or falling highs. Two points draw it. Any two points define a line. The third touch is the first evidence it means anything. It is the most subjective object in technical analysis: shift your choice of second low by one bar and the angle changes, and so does everything you concluded from it.

Moving averages do a similar job with no drawing involved. There is no room to flatter yourself, which is a real advantage. The 50-day and 200-day are watched widely enough that orders collect around them. They also update on their own. The moving averages guide has the calculation. It also has the conditions where those dynamic levels stop working.

Role reversal, and the trade it defines

When resistance breaks, it often becomes support. Traders who sold at the level watch it break and some buy the pullback, and traders who missed the breakout get a second chance at a better price.

Carry the zone forward. The stock spent months rejecting 41.78 to 42.31. It breaks out to 44.50, pulls back to 42.40, and holds.

Item	Value	Arithmetic
Entry	42.80	Taken once the retest is holding
Invalidation	41.60	Below the zone, allowing the full 0.53 of width
Risk per share	1.20	42.80 - 41.60
Risk budget	$250	1% of a $25,000 account
Position size	208 shares	250 / 1.20 = 208
Cost of being wrong	$249.60	208 x 1.20

Nothing in that table forecasts anything. The level supplied a price at which the idea is wrong, the arithmetic turned that price into a share count, and that transformation is the entire practical value of support and resistance, because it works exactly as well when the level fails.

False breaks are a mechanic

Stop orders cluster just beyond obvious levels. That is where everybody was told to put them. Pushing price through the level triggers those stops, and the resulting market orders create a burst of trading that resembles demand for a few minutes or a few hours. Then the stops run out. Nothing is left behind the move, and price falls back into the range.

Two filters reduce how often this catches you, and both cost you something. Requiring a daily close beyond the zone makes the move survive to the closing auction, and an intraday poke no longer counts, at the cost of a worse entry. Requiring participation means comparing the day’s volume with the 50-day average. 5.8 million / 2.0 million = 2.9 times normal is a different event from 0.8 times normal. Neither filter is free. Neither is reliable. I would sooner say so than pretend some setting removes the problem.

Where levels stop being levels

Scheduled news overrides them completely. A stock that opens 12% below the previous close has skipped every zone on the chart, and there was no price inside that gap at which a stop could have been filled. Levels are a claim about resting orders. A gap is the market reopening somewhere those orders never were.

Levels also decay. A zone built from trading two years ago has been picked over. Prefer zones from the last six to twelve months. Prefer ones tested recently enough that the orders are plausibly still there.

The failure mode that costs the most money is subtler. A trader who buys at support, watches price break it, and buys more because the level is still on the chart has converted a risk tool into a rationalisation. The level was only ever a hypothesis. Price trading through it is the test coming back negative.

One chart also cannot tell you whether the whole market is sliding. Support in a single name has a harder job when the index is breaking down, and the market breadth tool shows how many stocks are above their moving averages and making new highs, which is the background those individual levels hold or fail against.

The stock screener filters for names near 52-week highs and above their moving averages, which is a quick way to find stocks with little overhead supply left on the chart. Read the candlestick chart guide for the bar-level evidence that a zone is holding, then the chart patterns guide for the structures levels build into, and check yourself against the technical analysis basics quiz.