The previous day's high and low keep working because they are the two prices everyone can agree on without argument. They are unambiguous, every platform draws them the same way, and a large number of resting orders — targets, stops and breakout entries — get placed against them. Levels become real when enough people use the same one.
That sounds almost too simple to be the whole answer, so it is worth being precise about why it holds, where the effect comes from, and what it does not promise.
What makes these two levels different from every other level
Almost every line on a trading chart involves a choice. A trendline depends on which two lows you connect. A moving average depends on the period you type in. A supply zone depends on where you decide the zone starts and stops. Two competent traders looking at the same chart will draw them slightly differently, and that difference is enough to scatter their orders across a range of prices instead of concentrating them at one.
The previous day's high and low have no such choice in them. They are the maximum and minimum traded price over a defined session — a fact, reported identically by every data vendor, visible to anyone who scrolls left. Nothing about them is interpretive.
That matters for a mechanical reason rather than a mystical one. A price level only produces a visible reaction if enough executable orders are resting at or near it. Levels that require interpretation spread orders out. Levels that are objective concentrate them. The previous day's high and low are the most objective levels on an intraday chart, which makes them among the most crowded, which is why price so often does something noticeable when it arrives.
US market structure has the previous session built into it
This is not only a chart convention. The prior session is written into the rules that govern when trading stops.
US market-wide circuit breakers halt trading across every exchange when the S&P 500 falls far enough — and the yardstick is explicitly the day before. As the New York Stock Exchange states, the three thresholds are measured as a decrease against the prior day's closing price of the S&P 500 Index:
- Level 1 — a 7% decline. Halts trading market-wide for 15 minutes if triggered before 3:25 p.m. ET.
- Level 2 — a 13% decline. Same 15-minute halt, same cut-off.
- Level 3 — a 20% decline. Trading stops for the remainder of the day, whenever it is hit.
Read that again from a trader's point of view. On the most extreme days of the year, the number that decides whether the entire US equity market keeps trading is measured from yesterday's close. The previous session is not a habit technical traders picked up; it is the reference frame the market itself resets to every morning.
The same logic runs through the single-stock rules. The SEC's investor bulletin on limit up-limit down describes price bands of 5%, 10% or 20% depending on the security, computed from recent trading and doubled during the opening and closing periods — the two moments when the session boundary itself is under the most stress. Sessions are units the market takes seriously.
What is actually resting at yesterday's high
Four different groups of orders converge on that one price, and they arrive there for unrelated reasons:
- Breakout buy stops from traders waiting for the market to exceed the prior range.
- Protective stops from short sellers who are wrong if the previous high gives way.
- Take-profit sells from anyone long from lower, using the prior high as an obvious exit.
- Fade entries from traders selling the level on the assumption it holds.
Notice that these push in opposite directions. Buy stops and short covering push price up through the level; take-profits and fades push it back down. That opposition is precisely why the previous day's high produces a reaction rather than a predictable direction. Something happens there because the flow is dense; which way it resolves depends on which side is larger on the day, and you cannot know that in advance.
One honest caveat. Unlike round numbers, where a Federal Reserve Bank of New York study of a real order book documented the clustering directly, there is no comparable published order-book study for previous-session highs and lows. The reasoning above is a mechanism, not a measured result. If you want the version with numbers behind it, see round numbers as support and resistance — and where a round number and the prior day's high sit at the same price, you have two independent reasons for one level.
The four things price can do at the level
Naming these in advance is most of the work, because it stops you deciding what you are looking at after the fact.
| What happens | What it usually means | Reasonable response |
|---|---|---|
| Reaches the level and rejects | Opposing flow was heavier; the prior range still contains the market | Range logic — the opposite boundary becomes the reference |
| Breaks and closes beyond it | Breakout orders overwhelmed the fade; the range has resolved | Wait for the close, then for the retest to hold |
| Breaks, then closes back inside | A failed breakout; the orders beyond the level have been consumed | Often the strongest reversal signal the level gives |
| Never reaches it all session | Today is building a range inside yesterday's; conviction is low | Reduce size or stand down — this is a range day, not a trend day |
The third row deserves more weight than most traders give it. A break that fails has told you something a clean break has not: the resting orders on the far side have already been filled, and there is nobody left to push. That is the same machinery described from the other direction in liquidity grabs and stop hunts.
How to mark them, in order
- Decide your session definition once, and never change it. For US stocks, use the regular 9:30 a.m. to 4:00 p.m. ET session. For futures, most traders use the cash-session high and low rather than the full electronic range, because that is where the volume sat. A nearly-24-hour market has no natural midnight, so the definition has to be a decision — and an inconsistent one is worse than either choice.
- Draw three lines, not two. The high, the low, and the close. The close behaves as the session's centre of gravity, which is why it matters so much for whether and how a gap fills.
- Mark them before the open, with everything else. They belong on the chart during your pre-session markup, alongside prior swing points and round numbers — not discovered halfway through a move when you are already looking for a reason to be in.
- Write down what each level means for your plan. Not "watch the previous high" but "above the previous high and holding, I am looking for continuation toward the next reference; rejected there, I am looking for a rotation back toward the prior close."
- Keep the stop off the level. It is the most crowded price available. If the prior high is genuinely your invalidation, put the stop beyond it with room for a wick and size from that wider distance — the logic in how to set a stop loss that isn't a guess.
None of this is a strategy on its own. Previous-day levels are context: they tell you where the market currently considers itself expensive and cheap, which is the frame every other read sits inside. That framing role is why they appear in almost every approach covered in technical analysis basics, regardless of which indicators the trader otherwise uses.
Frequently Asked Questions
Why do the previous day's high and low act as support and resistance?
Because they are the only two levels on the chart that every trader draws identically. A trendline depends on which points you pick and a moving average depends on your settings, but yesterday's high is a single unarguable number that every platform reports the same way. That shared definition is what lets orders accumulate at one price, and accumulated orders are what produce a reaction.
Which matters more, the previous day's high and low or the previous day's close?
They do different jobs. The high and low are the boundaries of what the market was willing to pay, so they act as resistance and support. The close is the market's final agreed price and is used as an official reference point — US market-wide circuit breakers are measured against the prior day's closing price of the S&P 500. Mark all three and treat the close as the day's centre of gravity.
Do previous day levels work on futures and forex, which trade almost around the clock?
Yes, but you have to decide which session boundary you mean, because a 23-hour market has no natural midnight. Most futures traders use the regular-hours cash session high and low rather than the full electronic session, because that is where the volume is. The important thing is consistency: pick one definition and keep it, since a level only works when people are looking at the same price.
Where should the stop go on a previous day high breakout?
Beyond the level rather than at it, with room for the retest. The previous day's high is one of the most crowded prices on the chart, so a stop sitting exactly there is the easiest one in the market to reach. Decide the price that would prove the breakout wrong, place the stop past it, and size the position from that distance rather than shrinking the stop to fit a preferred size.
Bottom line
The previous day's high and low survive every change in fashion because they require no interpretation. Everyone gets the same two numbers, so everyone's orders land in the same two places, and dense order flow is what makes a level behave like a level. The market itself agrees: the rules that halt US trading on the worst days measure the damage from yesterday's close. Mark the high, the low and the close before the open, define in advance what each of the four outcomes would mean, keep your stop off the crowded price, and let the close — not the touch — tell you which outcome you got.