Pivot Points High Low: Definition, Settings, and Why Every Pivot Is Known Late

The Pivot Points High Low indicator marks swing highs and swing lows on the chart: a bar whose high is above the highs of a set number of bars on each side, or whose low is below the lows on each side. It is how most charting tools find the turning points that traders use for trendlines, support and resistance and swing structure. It has one property that decides how it can be used: a pivot can only be confirmed after the bars to its right have closed, so the mark always appears late. This page covers the definition, how late the confirmation comes on real prices, and the ZigZag, which has the same problem with its last line.

How a pivot high and a pivot low are defined

LuxAlgo defines a swing high as a bar whose high stands above the highs of a set number of bars on either side, and a swing low as the mirror image. The number of bars checked on each side, called strength, lookback or left and right bars depending on the tool, is the only parameter: small values mark every wiggle, large values only the major turns. These swing points are also called pivot highs and lows, or fractals; a Williams fractal is the case with two bars on each side.

This is a different tool from the floor-trader pivot points, which compute support and resistance levels from the previous day's high, low and close. Both share the word pivot and nothing else.

The pivots on a real chart

The pivot chart below the introduction shows pivots with 5 bars on each side on daily Bitcoin, July 1 to October 28, 2026. Each mark sits on the bar that made the high or low, but it could only be drawn five bars later. The table lists all 16 with the day they became known and how far the close had already moved from the pivot by then.

By the time each pivot was confirmed, the close was already 1.36% to 10.87% away from it, in the direction of the new swing. The lowest low on the chart, 79,220.74 on October 24, is not marked: it is below the 5 lows before it and the 4 lows after it, but the fifth bar after it had not closed on October 28.

How late the confirmation comes

LuxAlgo puts it plainly: a swing high with strength 5 is only knowable five bars after it printed, and backtests that act on the pivot bar itself use information that was not available in real time. To measure the cost of waiting, we found every pivot on the full daily history of the S&P 500 ETF (January 1993 to October 2026) and Bitcoin (August 2017 to October 2026) and compared the close on the confirmation day with the pivot price.

The last two columns are medians; the pivots per year column counts highs and lows together. Wider settings find fewer, more important turns, and the price has moved further by the time they are confirmed: on Bitcoin, a 10/10 pivot low was confirmed with the close a median 11.39% above the low. Fewer bars on the right, as in 10/3, keeps the left-side filter but confirms sooner. No setting removes the trade-off; it only chooses where to sit on it.

The ZigZag and its moving last leg

The ZigZag connects swing points that are at least a set percentage apart; StockCharts gives 5% as the traditional setting. On a chart with highs and lows, a ZigZag high is replaced by any higher high, and a bar whose low is the set percentage below it starts a new ZigZag low; if a bar does both, the current trend wins. StockCharts warns that the ZigZag has no predictive power because it is drawn with hindsight, and that its last line is temporary until a new qualifying point appears.

MetaTrader 4 is different. An article on the MQL5 site, using the ZigZag included in the standard MetaTrader 4 terminal as its example, lists its parameters as ExtDepth, ExtDeviation and ExtBackstep, with default values of 12, 5 and 3. These are not the single percentage threshold used here, so a MetaTrader 4 ZigZag and a percentage ZigZag on the same chart will not mark the same turns; check which version a chart or a study uses before comparing them.

Here is that in practice, with a 10% ZigZag on daily Bitcoin. On August 18, 2026 the chart looked like this:

The last fixed point was the July 1 low at 57,800.19, and the last leg ran up to the July 21 high at 66,956.15, then down to the latest bar. Price had fallen at most 6.99% from that high (to 62,275 on August 1), short of the 10% that would have fixed it. A reader looking at this chart would see a swing high on July 21. Here is the same 10% ZigZag on October 28, from the same start date:

The July 21 high is gone. The next day, August 19, Bitcoin traded up to 70,000, above it, so the ZigZag high moved up, and kept moving with every new high until September 21 at 87,395.67. That point is still not fixed either: the lowest low since, 79,220.74 on October 24, is 9.35% below it. Earlier points were fixed with a delay: the May 6, 2026 high at 82,850 only became final on May 23, 17 bars later, when the low first reached 10% below it, with the close already 7.36% lower.

Does it work?

Pivots and the ZigZag describe structure; they do not forecast it, and both are only known after the fact. They are useful as inputs, for drawing trendlines, marking support and resistance or checking higher highs and higher lows, as long as every rule uses the confirmation date. Our bench builds its chart patterns from swing points dated on the day they are confirmed, and measures them against random entries (see our methodology); the Elliott wave and Fibonacci retracement pages use the same swings.

For other tools built on swings and averages, see the Williams fractal, the pivot points and the Donchian channel, or the full list of indicators.