Pivot points are not forecasts. They are a repeatable way to turn a finished period's high, low and close into reference prices for the next one. That makes them useful as a shared map; it does not make a touch of a line an edge. The standard calculation is documented by both Fidelity's indicator guide and IG's explanation of range trading.
The calculation is simple
A standard pivot point starts with the previous period's high (H), low (L) and close (C). The central pivot is (H + L + C) / 3; first support and resistance are then derived from that pivot and the same completed period. Fidelity publishes this family of formulas, as does IG.
P = (H + L + C) / 3
R1 = (2 × P) - L S1 = (2 × P) - H
R2 = P + (H - L) S2 = P - (H - L)Some chart packages also plot a third support and resistance level. The important point is less the count of lines than the input: once the selected period has closed, the levels are fixed by arithmetic. Fidelity explicitly describes three levels on each side in its implementation; IG documents the common five-point version. Fidelity and IG therefore agree on the core inputs while showing why a rule must name its exact variant.
Why the chart can disagree with itself
The phrase “previous day” hides a real implementation choice. Fidelity's guide assigns different pivot periods to different chart intervals and lists distinct daily start times for equities, forex and metals; IG notes that pivots can be applied on different timeframes. Fidelity's period specification and IG's implementation note are a useful reminder that platform, market and session definition can alter the inputs.
That is not a minor footnote. A daily pivot built from one session boundary is a different calculation from a daily pivot built from another. The formula has not failed; the test definition was incomplete. Before interpreting any level, record the instrument, chart timezone, session cut-off and formula variant. The same need for explicit definitions appears in Fidelity's guide and in IG's note that pivots can be applied on different timeframes.
How traders turn levels into a method
The usual discretionary reading is straightforward: treat P, S1 and R1 as places to observe whether price accepts, rejects or moves through a reference level. IG frames pivot points as potential areas where direction may change and places them alongside other range-analysis tools; CME's technical-analysis course cautions that patterns are indications rather than absolute rules and discusses confirmation. IG's pivot-point lesson and CME Group's course support that restrained framing.
The trouble starts when “rejection” means whatever makes the chart look tidy after the fact. A candle wick, a close back through the line, a failed breakout and a slow reversal are not the same event. Nor is a pivot level a reason to assume reversal rather than continuation. Technical-analysis education itself treats confirmation as part of the interpretation, not a property supplied by the line. CME Group makes that point directly; IG likewise describes pivots as one tool that can be combined with others.
What pivots do not prove
The calculation is objective. The story attached to it often is not. A level can coincide with a reversal, a pause, a breakout or nothing noticeable; the formula alone contains no rule for which outcome counts as success. That distinction follows from the mechanical formulas in Fidelity's reference and the conditional, confirmation-based treatment of technical signals in CME's course.
This is why screenshots are weak evidence. They select a market, a period, a level and a reaction after the result is known. A useful pivot method must survive the unphotogenic cases too: a level crossed without hesitation, a volatile session that touches several levels, and the costs of entering and exiting. For an explanation of why the execution assumptions belong in the evidence, see why a cTrader backtest changes across brokers.
How you'd actually test it
Start by freezing the specification before opening the results. Choose one instrument universe, one bar source, one timezone, one session cut-off and one pivot formula. Then define a signal that a program can recognise without interpretation: for example, whether a bar closes on one stated side of P after first touching it, not “price looked supported.” The need to state the period and formula comes directly from Fidelity's specification and IG's formula description.
Next, define the whole trade path before measuring it: entry timing, stop condition, exit condition, maximum holding time and what happens if more than one level is touched. Apply realistic spread, commission and slippage assumptions, then compare the rule with a deliberately simple baseline that uses the same holding and risk rules but ignores pivots. That comparison asks the only useful question: did the level rule add information, after costs, or merely add a story?
Finally, keep the design sample separate from the evaluation sample, and report the losing regimes as plainly as the attractive ones. The out-of-sample testing explainer and guide to verifying a cTrader backtest cover the discipline in more detail. realbacktesting is a trading-software studio for cTrader built around verifiable work: a method earns attention when its data, rules and costs can be checked, not when its lines look persuasive in hindsight.
Frequently asked
Are pivot points the same as support and resistance?
Pivot points are a formula-based way to create potential support and resistance references from a completed period's high, low and close. They are not the same thing as a proven future barrier; Fidelity classifies them as a support/resistance study, while IG presents them as possible change areas.
Do pivot points work on every timeframe?
The arithmetic can use different periods, but a test must state which period and session definition it uses. Fidelity maps pivot periods to chart intervals, and IG notes that pivots can be added on different timeframes. Fidelity and IG document those implementation choices.
Takeaway
Pivot points are easy to calculate. The hard part is proving that an unambiguous rule around them improves on doing nothing after real costs.