Where do you actually find support and resistance?
Knowing what a level is doesn't tell you where to put one. Ask five traders to mark a chart and you'll get five different sets of lines. Here is a repeatable procedure that at least makes your lines checkable — by you, next month, when you've forgotten why you drew them.
The five usual sources
1. Prior swing highs and lows. The single most-used source, and now you already have them from Unit 2. A price that stopped a move once is the definition of a place where supply or demand appeared.
2. Consolidation zones. Areas where price spent many periods going sideways. A lot of shares changed hands there, so a lot of participants have a reference point in that band. Traders generally treat these as more meaningful than a single sharp spike, because more business was actually done.
3. Edges of a gap. When one period's range and the next period's range do not overlap — price opening beyond the previous session's high or low and never trading back across the space between — the empty band often becomes a watched region and its edges get marked.
4. All-time or multi-year highs and lows. Above an all-time high there is no overhead supply from previous buyers — nobody up there is waiting to break even, because nobody bought there. That structural fact makes the area behave differently from ordinary resistance.
5. Round numbers. Common enough to earn their own lesson next.
The procedure
- Zoom out first. Start on the highest timeframe you care about — weekly for most equity questions. Levels drawn on a weekly chart still matter on a daily chart; levels drawn on a 5-minute chart do not matter on a weekly one. Big to small, always.
- Mark only areas price reacted to more than once. One touch is a coincidence generator.
- Draw zones, not hairlines. More on this in Unit 4 — but start the habit now.
- Stop at five or six. A chart with twenty lines on it can "explain" any price. If everything is a level, nothing is.
A worked example
An illustrative stock over two years, on the weekly chart:
- Swing highs at $118, $121, $119 → these three cluster. Mark one zone: $117–122.
- A four-month sideways stretch between $92 and $97 → mark $92–97.
- A single spike low at $71 during one panic week → note it, but rank it lower: one touch, in unusual conditions.
Two clear zones and one weak one, from two years of data. That's a normal harvest. The zones are simply a record of where this stock has repeatedly done business — traders watch how price behaves when it re-enters them, and nothing about a zone recommends any action.
Rank them, don't just draw them
Not all levels are equal. Practitioners generally weight by: how many times price reacted there, how much time price spent there, how recently, and on what timeframe it was drawn. A weekly zone touched four times over two years is a different object from a daily line touched twice last month. Writing the rank next to the level keeps you honest when one of them is inconveniently in the way of a story you like.
In the data
One family of levels comes from arithmetic rather than from the chart: floor-trader pivots. They take one completed session, here Apple on 25 September 2026:
The pivot is (high + low + close) ÷ 3, the first resistance is 2 × pivot − low, and the first support is 2 × pivot − high. Every number comes from that one bar and nothing else, so a level built this way carries no record of how often price stalled there before, which is exactly the property the lesson's other sources are chosen for.
Try it now
- Work the procedure on the two schematics above: one zone from the three clustered turns, one from the edges of the gap. Write the touch count beside each and rank them against one another.
- Now start big, as the procedure says. Mark a maximum of five zones on the five years below, with a touch count and a recency beside each.
- Then check them against the last year of the same instrument. Are your zones still where price has been pausing? Observation only — you are testing your own marking, not forecasting anything.