Pivot points: levels calculated before the open
StockCharts' ChartSchool, on its Pivot Points page, defines them as levels that use the prior period's high, low and close to estimate future support and resistance levels. It recalls where they come from: floor traders, who at the beginning of the day calculated a pivot point from the previous day's prices, then two supports and two resistances around it, and used them throughout the session. TradingView's help page on the Pivot Points Standard indicator gives a more cautious definition: levels at which price might meet support or resistance.
These levels have two distinctive features. They are known before the session starts, since they depend on past prices only: that is why StockCharts files them among leading indicators, unlike a moving average, which is recalculated on every candle. And they are fixed: once set, the same page writes, they do not change and remain in play throughout the day.
The word 'pivot' is used for other things too. A pivot can simply mean a swing high or low on a chart, which is an observation and not a calculation. Jesse Livermore's 'pivotal point' is a key price whose break confirms a move: the article on Jesse Livermore presents it. The 'episodic pivot' is a breakout that follows major news, described in the article on the gap and go strategy. None of these notions is calculated with the formulas in this lesson. Pages consulted in October 2026.
How pivot points are calculated: the classic formula
Three prices are enough. The high, the low and the close of the reference period, the previous day for a chart of a few minutes. The letters are the usual ones: P for the pivot, R for resistances, S for supports.
The pivot. P = (high + low + close) ÷ 3. StockCharts, TradingView and thinkorswim's page on the PivotPoints study write the same average.
The first level. R1 = 2 × P − low, and S1 = 2 × P − high. In other words, R1 is as far above the pivot as the previous day's low is below it, and S1 as far below as the previous day's high is above: that is how thinkorswim describes them. The gap between R1 and S1 therefore always equals the previous day's range, high − low.
The second level. R2 = P + (high − low), and S2 = P − (high − low): the pivot, plus or minus the previous day's range. So far, the three sources give exactly the same formulas, under the names 'Standard' at StockCharts and 'Traditional' at TradingView.
The third level, where the sources part ways. StockCharts' list of formulas stops at R2 and S2. TradingView, in its 'Traditional' type, writes R3 = 2 × P + (high − 2 × low) and S3 = 2 × P − (2 × high − low). Thinkorswim places the third level one range away from the second, that is R3 = P + 2 × (high − low) and S3 = P − 2 × (high − low), which TradingView also offers, but under another name, the 'Classic' type. Two formulas for the same R3 label: the example below puts a number on the gap.
Pivot points worked by hand: seven levels for one day
The prices are invented. The reference day has a high at 104.00, a low at 98.00 and a close at 102.50. Its range is 104.00 − 98.00 = 6.00. It is the framed candle on the left of the diagram.
The pivot. P = (104.00 + 98.00 + 102.50) ÷ 3 = 304.50 ÷ 3 = 101.50.
The first level. R1 = 2 × 101.50 − 98.00 = 105.00. S1 = 2 × 101.50 − 104.00 = 99.00. The gap between the two is indeed 6.00, the previous day's range. They are not at the same distance from the pivot: 3.50 above for R1, 2.50 below for S1, because the previous day's close was in the upper part of its candle.
The second level. R2 = 101.50 + 6.00 = 107.50. S2 = 101.50 − 6.00 = 95.50.
The third level. With TradingView's 'Traditional' formula: R3 = 2 × 101.50 + (104.00 − 2 × 98.00) = 203.00 − 92.00 = 111.00, and S3 = 2 × 101.50 − (2 × 104.00 − 98.00) = 203.00 − 110.00 = 93.00. With thinkorswim's formula, which is TradingView's 'Classic' type: R3 = 101.50 + 2 × 6.00 = 113.50, and S3 = 101.50 − 2 × 6.00 = 89.50.
What the example shows. The first five levels are the same everywhere: 95.50, 99.00, 101.50, 105.00 and 107.50. The third depends on the software: 111.00 or 113.50 for R3, 93.00 or 89.50 for S3, a gap of 2.50 at the top and 3.50 at the bottom, for a day that only travelled 6.00. Before comparing an R3 read elsewhere with the one on your screen, check which type is selected.
Fibonacci, Camarilla, Woodie, DeMark: other levels, same prices
StockCharts writes that there are at least five different versions of pivot points and details three of them. TradingView offers six in a single indicator: Traditional, Fibonacci, Woodie, Classic, DM and Camarilla. All start from the same day. Here is what they give on the day of the example, with the formulas written by those two sources.
Fibonacci. The pivot does not change, 101.50. The levels sit 38.2%, 61.8% and 100% of the previous day's range away from it: R1 = 101.50 + 0.382 × 6.00 = 103.79, R2 = 101.50 + 0.618 × 6.00 = 105.21, R3 = 101.50 + 6.00 = 107.50, and on the other side S1 = 99.21, S2 = 97.79, S3 = 95.50. StockCharts and TradingView write the same formulas. The Fibonacci R3 is therefore the traditional R2 under another name, and the Fibonacci R2, at 105.21, lands 0.21 from the traditional R1.
Camarilla. According to TradingView's formula, the levels no longer start from the pivot but from the previous day's close, to which a fraction of the range multiplied by 1.1 is added or subtracted: one twelfth, one sixth, one quarter, then one half. On the example: R1 = 102.50 + 1.1 × 6.00 ÷ 12 = 103.05, R2 = 103.60, R3 = 104.15, R4 = 105.80, and S1 = 101.95, S2 = 101.40, S3 = 100.85, S4 = 99.20. The levels are much tighter: the Camarilla R3, at 104.15, is below the traditional R1.
Woodie. The pivot is no longer calculated with the previous day's close but with the current day's open, counted twice: P = (high + low + 2 × open) ÷ 4, according to TradingView. With an open at 102.40, P = (104.00 + 98.00 + 2 × 102.40) ÷ 4 = 101.70, R1 = 105.40 and S1 = 99.40. These levels are therefore not known before the open.
DeMark ('DM' at TradingView). The formula depends on the direction of the reference candle, and there is only one support and one resistance: StockCharts spells it out. The previous day opened at 99.50 and closed higher, so X = 2 × 104.00 + 98.00 + 102.50 = 408.50, P = X ÷ 4 = 102.13 (102.125 rounded), R1 = X ÷ 2 − 98.00 = 106.25 and S1 = X ÷ 2 − 104.00 = 100.25.
The upshot fits in one sentence: for the same day, the 'first resistance' is 103.05, 103.79, 105.00, 105.40 or 106.25 depending on the variant. Neither source names one variant as better than another, and nothing on their pages makes it possible to do so.
Which reference day? The timeframe and the close that is used
The formulas are simple. What changes the levels most is the three prices fed into them, and each platform has its own rule.
The period depends on the chart's timeframe. At StockCharts, 1, 5, 10 and 15-minute charts use the prior day; 30, 60 and 120-minute charts the prior calendar week; daily charts the prior month; weekly and monthly charts the prior year. At TradingView, the 'Auto' setting takes the day up to and including 15 minutes, the week above 15 minutes and below one day, the month from one day and above. The two rules agree intraday, not beyond: on a weekly chart, one uses the past year and the other the month. At thinkorswim, the study counts its weeks and months back from the last day, so its weekly levels are different each day; the platform has another study, Persons Pivots, which takes the last complete week.
The daily candle is not always the one you think. TradingView explains it in the description of its 'Use Daily-based Values' option. On stocks, extended hours data is usually not included in the daily candle: pivots calculated on intraday data that includes it are, according to that page, very different. On futures, the daily close usually represents the settlement price, which is an average value and not the price of the last trade, and the daily candle uses the electronic trading session, while an intraday chart may display another one. NinjaTrader's documentation on its Pivots indicator provides three ways of obtaining those three prices: calculating them from intraday data, reading them from daily bars, or entering them by hand.
What a different close changes. Take the day of the example again with the same high and the same low, but a close recorded at 100.40 instead of 102.50. The pivot moves from 101.50 to 100.80, R1 from 105.00 to 103.60, S1 from 99.00 to 97.60, R2 from 107.50 to 106.80. The pivot and the second levels shift by 0.70, the first and third levels by 1.40. Two screens, the same market, two grids.
On a market that trades around the clock, the question is sharper still: forex and cryptocurrencies have no official close each evening, and the 'day' there starts and ends at the time chosen by whoever supplies the data. The pages read here do not give that time: it has to be checked on your platform, for your instrument. As long as two platforms do not cut the day at the same moment, they have neither the same high, nor the same low, nor the same close.
MetaTrader 5 does not plot pivot points out of the box: its help lists 38 built-in indicators, split into four groups, and MetaTrader 5's help page on technical indicators counts none by that name. There they come from a custom indicator, whose formula and reference day are those its author chose.
Reading pivot points during the session
The reading described by StockCharts. The pivot sets the tone: a price above it is seen as positive, with the first resistance as a target, and a break above R1 points to R2. The reading is symmetrical below the pivot. Supports and resistances are used like ordinary levels: you watch for a bounce or a failure there, and the page recommends having it confirmed by a candlestick pattern or by another indicator. It adds that a move beyond the second level shows strength, but also an overbought condition that could lead to a pullback. These are reading conventions, not rules.
The session in the diagram, candle by candle. It opens at 102.40, above the pivot. Candle 2 dips exactly to 101.50, candle 3 prints 101.45, that is 0.05 below the pivot, and closes at 102.20: the level held, to within 0.05. Price then rises to 104.90 on candle 9, 0.10 from R1, without touching it, then falls back to 103.20. On candle 15, it goes through R1 without pausing and closes at 105.60.
The high of the session comes on candle 19, at 106.80: 0.70 below R2, on none of the seven levels. Price then drops back below R1 on candle 22, returns to it on candle 23 with a high at 105.05 and a close at 104.90, then finishes at 104.20. In one session, the same grid therefore produced a nearly exact bounce, a top close to a level without touching it, a break with no reaction and a reversal where it drew nothing.
Precision is judged in volatility. Has a price that stops 0.10 from a level respected it? On the example, 0.10 is less than 2% of the previous day's range, which is 6.00: the question is settled with a measure of the usual range, not by eye. That is the subject of the lesson on the ATR. Setting that tolerance before looking at the chart avoids choosing it afterwards, level by level.
The more variants, the closer a level always is. Between the low and the high of this session, that is 5.35 points, the six variants calculated above set 13 distinct levels, and the widest gap between two neighbours is 0.92. With the close at 100.40 from the previous section, R2 would be 106.80, exactly the high of the session: the same candles give a level hit to the cent on one screen and a top in the middle of nowhere on the other. A coincidence between a price and a line therefore proves little as long as the variant, the reference day and the tolerance have not been set in advance.
None of these readings is an order to buy or sell. They are to be tested on your market, your timeframe and your variant, with the method in the article on how to backtest a trading strategy. Tradoshi does not record pivot points at the time of your trades: they are read on your chart. The journal is for the next step: you tag every trade taken on this reading, for example 'bounce off the pivot', and the dashboard filters by tag to show what it produced for you.
What the research says about support and resistance
The study closest to the subject is Carol Osler's Support for Resistance: Technical Analysis and Intraday Exchange Rates, published in July 2000 in the Federal Reserve Bank of New York's Economic Policy Review. It covers the currency market: the German mark, the yen and the pound against the dollar, from January 1996 to March 1998, with quotes sampled every minute from 9:00 to 16:00 New York time. It tests the support and resistance levels that six firms, banks and information providers, sent to their customers each day.
Its measure: price 'hits' a level when it comes within 0.01% of it, and the trend is said to be interrupted if, fifteen minutes later, it is back on the right side of the level. Result: price bounces 60.8% of the time on average off the published levels, against 56.2% off arbitrarily chosen levels. The difference is statistically significant for most firms and currencies, and it persists five business days after the levels were published.
What this study does not say matters just as much. Those levels do not come out of a formula: they come from analysts' judgement, more than 70% of them end in a zero, and the firms agree little with each other, on roughly 30% of levels. The study therefore does not test pivot points, and its text does not mention them. Nor does it measure whether a profit can be made from them: the author writes that this would be a subject for future research. And a bounce 60.8% of the time against 56.2% by chance leaves nearly four cases in ten where price does not bounce.
Another study by the same author, the Federal Reserve Bank of New York's Staff Report No. 125, offers an explanation: the stop-loss and take-profit orders of a large currency dealing bank cluster strongly at round numbers, which are often used as support and resistance levels. It does not cover calculated levels either.
As for how often a calculated pivot level holds, we have read no published study that measures it. The percentages circulating on commercial sites could not be traced to a publication, and this lesson does not repeat them.
The limits of pivot points
They see one period only. The seven levels come out of three prices from a single day, a single week or a single month. A very quiet previous day gives tight levels, a very busy one gives distant levels, and nothing in the calculation takes into account what happened before, or an announcement to come.
They depend on the variant. On the example, the first resistance ranges from 103.05 to 106.25 depending on the formula, and the 'classic' third level itself has two definitions. Stacking several variants on one screen multiplies the lines and, with them, the coincidences.
They depend on the platform. Day, week or month depending on the timeframe, extended hours included or not, settlement price or last trade, cut-off time on markets that trade around the clock: a difference of 2.10 on the close moved R1 by 1.40 in the example. The most distant levels may also not appear at all: StockCharts points out that the second or third sometimes exceeds the price scale of the chart.
A level is not a signal. The two reference sources word it carefully: levels at which price might meet support or resistance, to be confirmed with other aspects of technical analysis. A pivot support is handled like any other level: the article on break and retest describes what happens when it gives way, and the one on supply and demand zones another, discretionary way of setting levels.
They say nothing about the trend or the exit. Whether the move is speeding up or running out of steam is the question of the lesson on momentum and the Rate of Change, and that of a stop that trails price is covered in the lesson on the Parabolic SAR. The page on technical analysis recalls why a screen covered in lines agrees with any scenario.
Frequently asked questions
What is a pivot point in trading?
It is a price level calculated in advance: the average of the high, the low and the close of the previous period. Other formulas set supports (S1, S2, S3) and resistances (R1, R2, R3) around it. These levels stay fixed for the whole of the following period.
How do you calculate pivot points?
P = (high + low + close) ÷ 3, then R1 = 2 × P − low, S1 = 2 × P − high, R2 = P + (high − low) and S2 = P − (high − low). With a high at 104, a low at 98 and a close at 102.50, the pivot is 101.50, R1 105, S1 99, R2 107.50 and S2 95.50.
Why do my pivot points differ from one platform to another?
For two reasons. The selected variant is not the same (traditional, Fibonacci, Camarilla, Woodie, DeMark), or the three starting prices differ: extended hours included or not, settlement price or last trade on futures, the time at which the day is cut on a market that trades around the clock.
What is the difference between classic and Fibonacci pivot points?
The pivot is the same. Fibonacci levels sit 38.2%, 61.8% and 100% of the previous period's range away from it, whereas the classic formula uses the high, the low and the full range. The Fibonacci R3 equals the classic R2.
Are pivot points reliable?
We have read no published study that measures how often a calculated pivot level holds. Carol Osler's study (Federal Reserve Bank of New York, 2000) covers support and resistance levels published by six firms in the currency market, from 1996 to 1998: price bounces there 60.8% of the time, against 56.2% off arbitrary levels. Those levels did not come out of a formula.
Key takeaways
- A pivot point is the average of the high, the low and the close of the previous period; supports S1 to S3 and resistances R1 to R3 are derived from it and stay fixed all session.
- The first five levels have the same formula at StockCharts, TradingView and thinkorswim; the third level has two definitions depending on the software.
- Fibonacci, Camarilla, Woodie and DeMark give other levels for the same prices: on the example, the first resistance ranges from 103.05 to 106.25.
- The levels depend on the high, the low and the close the platform uses: timeframe, extended hours, settlement price, cut-off time.
- No study we read measures how well a calculated pivot level holds; Carol Osler's covers levels published by analysts, not a formula.
Going further
These blog articles dig into this lesson's ideas, one subject per article.