Support and Resistance: Zones, Breakouts, Fakeouts

Support is a price area where falling prices have stopped before, because buyers were willing to step in there. Resistance is the mirror image: an area where rising prices stalled because sellers showed up. Neither is a law of physics. They exist because traders remember prices: the people who bought at a level and watched it fall want out at breakeven, and the people who missed a bounce want in if it comes back. That memory is fuzzy, which is the first thing to build into how you draw levels.

Draw zones, not lines

Price chart with a lower support zone and an upper resistance zone. After a breakout, prior resistance acts as new support (polarity).

Look at any level on a real chart and the reactions do not line up to the cent. One high stops at 50.27, the next at 50.55, the next at 50.00. Drawing a single line forces you to pick one, and then every other touch looks like a "break" or a "miss". Draw the band from the lowest reaction to the highest instead. A zone that is wide relative to the stock's usual daily range is a sign that the level is vague; a narrow one that keeps getting tested is worth paying attention to.

A few things make a zone matter more:

• More touches, spread over time. Three reactions over two months beat three in one week.
• Clear reactions. A level that caused a sharp reversal says more than one price drifted past.
• Round numbers and prior extremes. People anchor on 50, 100, last year's high.
• Recency. A level from last quarter is fresher in traders' minds than one from five years ago.

A ceiling tested four times: NVIDIA around 50

NVIDIA (split-adjusted prices) spent the second half of 2023 unable to hold above 50. It peaked at 50.27 on August 24, then again at 50.55 on November 20, 50.43 on December 18 and 50.00 on December 29. Each time it pulled back, once as far as 45.01 in early December. That is a resistance zone of 50.00 to 50.55, with four touches over four months.

On January 8, 2024 it opened at 49.51 and closed at 52.25, 3.4% above the top of the zone, on 1.76 times its 20-day average volume. The next day volume rose again, to 2.03 times average. The lowest price over the following 20 sessions was 51.69, still 2.3% above the zone. Twenty sessions after the breakout, on February 6, the close was 68.22, 30.6% above the breakout close.

Waiting for the retest, and when it never comes

The textbook follow-up to a breakout is the retest: price comes back to the old resistance, finds buyers there, and turns up again. Old resistance becoming new support is called polarity, and a successful retest is one of the cleaner entries in chart reading, because your stop can sit just below the zone.

NVIDIA never gave that entry. The closest it came was 51.69, and anyone waiting for 50.55 missed the whole move. That is the trade-off of pullback entries: better price and tighter risk when they come, nothing at all when the move is strong. A common compromise is to take a small position on the confirmed breakout and add only if a retest holds.

Two breakouts that failed

Alphabet (GOOGL) offers two useful failures. In mid-2022 it topped out between 118.58 and 119.68 three times: June 6, June 27 and July 8. On August 15 it closed at 122.08, 2.0% above the zone. But volume that day was only 0.61 times its 20-day average; the sessions leading into it were quiet too, between 0.57 and 0.90 times average. Four sessions later it closed at 117.21, back below the zone, and on August 22 it was at 114.24.

In early 2025 Alphabet built a similar ceiling between 201.00 and 202.29, touching it on December 17, January 7 and January 21. On February 4 it closed at 206.38, again 2.0% above the zone, and this time on 1.73 times average volume. It looked like the real thing. But Alphabet reported earnings after the close that day. On February 5 it opened at 191.07, below the entire zone, and closed at 191.33, 7.3% below the breakout close. By March 4 it had traded as low as 165.80.

Breakout or fakeout: checking the evidence

No single row separates the winner from the losers. The August 2022 failure is explained by volume; the February 2025 failure had good volume and was undone by news that no chart could show. What the table does suggest is a routine: require a close beyond the zone, not an intraday poke; look at volume; check the calendar for earnings; and decide in advance that a close back inside the zone means the breakout failed. That last rule would have limited the damage in both Alphabet cases.

Support works the same way, upside down

Everything above applies to support with the signs flipped. A support zone with several clean bounces is where buyers have shown up before. A close below it on heavy volume is a breakdown; a close below it on thin volume that recovers within a few sessions is often a shakeout. Candles at the zone help: a hammer at support tells you buyers defended it during that session.

Marking levels on your own chart

1. Start with a daily chart of about six months. Mark the obvious turning points where price reversed sharply.
2. Group turning points within roughly one average daily range of each other into a single zone.
3. Keep the two or three zones nearest the current price. More than that and every candle is "at a level".
4. For each zone, write what a break would look like (a daily close beyond it) and what a failure would look like (a close back inside).
5. Check volume and the earnings date before trusting any break.

Zones also set up your risk: the distance from your entry to the far side of the zone is a natural stop, and the risk management guide shows how to turn that distance into a position size.