Pivot Points and Fibonacci Retracements: How to Calculate Them

Pivot points and Fibonacci retracements are mechanical ways of turning past prices into horizontal levels on a chart. Pivot points are calculated from the previous period's high, low and close; Fibonacci levels from a swing high and a swing low that you choose. Neither uses any information beyond those few prices.

Their popularity is the main argument for them: because many traders and some automated systems watch the same levels, orders tend to cluster around them, which can make them places where price pauses or reacts. The evidence that they predict anything beyond that is mixed, and a level on its own is never a reason to expect a reversal. Where they are genuinely useful is in planning a trade. They give you objective places for an entry, a stop and a target, and therefore a reward:risk ratio and a position size you can calculate before you trade.

High, low and close are from the previous period: the previous day for daily pivots, the previous week for weekly pivots, and the previous month for monthly ones. Longer periods produce wider, slower-moving levels that swing traders often prefer.

Worked example: EUR/USD daily pivots

Yesterday's EUR/USD: high 1.1052, low 1.0968, close 1.1031. The range is 0.0084 (84 pips).

Traders who use pivots typically treat the market as firmer while price holds above P and softer below it, and watch R1/S1 and R2/S2 as the first areas where a move might pause.

Which high, low and close?

Forex trades around the clock, so "yesterday" depends on where the day is cut. The most common convention is 17:00 New York time, which is also the daily rollover. Brokers whose server clock is set on that basis (usually shown as GMT+2 in winter and GMT+3 in summer) produce five daily candles a week. Brokers on other server times produce a short Sunday candle and different highs, lows and closes, and therefore different pivots. When you compare your pivots with someone else's, check that you are using the same session cut-off.

Several variants use the same inputs with different weightings.

Fibonacci pivots

Camarilla

Camarilla levels are centred on the close rather than the pivot and sit much closer together. R3/S3 and R4/S4 are the ones most traders watch.

Woodie

Woodie's method gives the close twice the weight of the high and low.

DeMark

DeMark is the only method that needs the open, and it produces just one support and one resistance.

All five methods on the same day

Same EUR/USD inputs: high 1.1052, low 1.0968, close 1.1031, and an open of 1.0990.

The methods agree on the broad picture but differ by several pips at each level, which is one more reason not to treat any single level as exact.

The formulas work on any instrument; only the inputs change.

Gold daily pivots

Yesterday's gold: high $4,125.40, low $4,061.80, close $4,108.20, a range of $63.60.

Pivot-to-pivot trades often have poor reward:risk. Suppose you buy at S1 ($4,071.50) with a stop below S2 at $4,030.00 and a target at the pivot ($4,098.50). The risk is $41.50 and the reward $27.00, a ratio of 0.65, so the trade needs to win about 61% of the time just to break even before costs. Using calc.forex's $0.10 gold pip, the stop is 415 pips; risking $100 gives $100 ÷ (415 × $10) = 0.024 lots, rounded down to 0.02 lots for an actual risk of $83. Working through the numbers like this before entering is the main value of the levels.

Indices: which session?

Index CFDs often trade nearly around the clock, while the underlying cash index trades only during exchange hours: 09:30 to 16:00 New York time for the S&P 500, Nasdaq 100 and Dow, and 09:00 to 17:30 Frankfurt time for the DAX. Pivots calculated from the cash session's high, low and close differ from those calculated from a 24-hour CFD candle, sometimes by tens of points. Decide which you use and stay with it.

Weekly and monthly pivots

Weekly pivots use the previous week's high, low and close; monthly pivots the previous month's. The levels are wider and change less often, so they suit trades held for days or weeks, and many traders plot them alongside daily pivots to see where levels from different timeframes coincide.

Fibonacci retracements measure how far price has pulled back within a move, as a percentage of that move. The ratios come from the golden ratio, about 0.618: 38.2% is 0.618², 23.6% is 0.618³, and 78.6% is the square root of 0.618. The 50% level is not a Fibonacci ratio at all; it is included because traders watch the halfway point anyway.

Worked example: EUR/USD rally from 1.0800 to 1.1200

For a fall from 1.1200 to 1.0800 the same ratios are measured up from the low: 38.2% at 1.09528, 50% at 1.10000 and 61.8% at 1.10472.

Worked example: gold rally from $3,900 to $4,100

Reading the depth of a pullback

Traders commonly read shallow pullbacks, to the 23.6–38.2% area, as a sign of a strong move, and deep ones, beyond 61.8–78.6%, as a sign that the move may be failing. A pullback that goes all the way past the start of the swing (100%) invalidates it: the swing low in an up-move has been broken, and any levels drawn from it no longer describe the market. That is why stops for retracement entries are usually placed beyond the next level, or beyond the swing itself, rather than just past the entry level.

Extensions project beyond the swing and are commonly used as profit targets once price breaks past the prior high or low. calc.forex measures them from the start of the swing:

Some charting tools project extensions from a third point (the end of the pullback) instead, so levels from different tools may not match. Know which convention yours uses before comparing levels with anyone else's.

• Use clear, significant swings on the timeframe you trade. A level drawn from a minor wiggle on a five-minute chart carries little weight on a daily chart.
• Be consistent about wicks or closing prices. Most traders use the extreme high and low (the wicks); either works if you always do the same.
• Redraw only when a new swing high or low forms, not to make the levels fit what price has already done.
• Several timeframes and several tools produce a lot of lines. When levels from different swings or from pivots fall within a few pips of each other, traders call it confluence. With enough lines on a chart, though, something will always be close to the price.

A level becomes useful when it fixes the numbers of a trade before you enter. Continuing the EUR/USD example: after the rally from 1.0800 to 1.1200, price pulls back to the 61.8% retracement. The account is $10,000 in US dollars, risking 1% per trade.

Whether the trade is worth taking is a separate judgement. What the levels have done is turn it into numbers you can check: what you lose if you are wrong, what you make if you are right, and how often it needs to work. The same steps apply to a pivot-based plan, for example a short below R1 with a stop above R2 and a target at the pivot.

Stops placed exactly on a well-watched level tend to be hit by ordinary noise. A buffer based on volatility is more objective than a fixed number of pips. With a daily ATR of 0.0060 on EUR/USD (60 pips), a buffer of a quarter of the ATR adds 15 pips beyond the level, while a stop set at 1.5 × ATR from the entry is 0.0060 × 1.5 ÷ 0.0001 = 90 pips away. Either way, recalculate the position size from the final stop distance: the further the stop, the smaller the position.

Because it is easy to see what you want to see on a chart, a level-based idea is worth testing before you risk money on it. A simple manual test looks like this:

1. Write the rule down precisely: which level, which timeframe, what counts as a touch, where the entry, stop and target go, and when a setup is skipped.
2. Step through the chart in order without looking ahead, and record every occasion the rule triggers, not just the memorable ones.
3. Record each result in R, after costs. A long from the 61.8% level that is stopped out is −1R; one that reaches a target 3.38 times the risk is +3.38R, less spread and commission.
4. Collect enough trades. With 30 trades, a rule whose true win rate is 40% can show an observed win rate anywhere from about 22% to 58% purely by chance (a 95% range). With 100 trades the range narrows to about 30–50%.
5. Compare the results with the break-even point. Is the win rate comfortably above the break-even win rate for the reward:risk you actually achieved, and is the expectancy in R clearly above zero once costs are included?

The same discipline applies to any indicator or pattern. A test of past data can't prove that a rule will work in future, but it shows quickly when one doesn't, which is usually the more valuable answer.

• Different data, different levels. Pivot points depend on the session cut-off and on your broker's prices; Fibonacci levels depend on the swing you choose.
• Self-fulfilling, up to a point. Reactions at popular levels come from the orders placed there, and large participants know roughly where those orders sit. Spikes through a level that then reverse are common.
• Hindsight bias. On a historical chart the levels that held stand out and the ones that failed are easy to overlook. Test any rule on a complete set of past trades, including costs, before relying on it.
• News overrides levels. Around central-bank decisions and major data, levels offer little protection; spreads widen and stops can slip.

Used as planning tools, places to define risk and reward and sized with a consistent risk per trade, pivots and Fibonacci levels are simple and transparent. Used as forecasts, they invite overconfidence.