Adjust box size and reversal to filter more or less market noise · Hover symbols for price + date details
A Point & Figure Chart encodes price action as a sequence of columns, each containing either X symbols (rising price, demand dominant) or O symbols (falling price, supply dominant). The x-axis is not a time scale — it is a column index. Time only appears as month markers on the first symbol of any new month, and these are informational annotations, not axis positions. The y-axis is the price scale.
Two parameters control the chart's resolution. The box size sets the minimum price move required to add a new symbol — it filters out moves smaller than this threshold. The reversal amount (expressed as a multiple of box size) sets the minimum price move in the opposite direction required to start a new column. Together they act as a two-stage noise filter: box size filters micro-fluctuations, reversal amount filters minor retracements.
Most financial charts plot price against time, which means quiet consolidation periods (low volatility, few price changes) take up the same horizontal space as high-activity periods. This time-scaling distorts the visual weight of price action: a three-week consolidation looks as wide as a three-week rally.
The P&F chart eliminates this distortion by only advancing to a new column when a reversal occurs. Each column represents a price movement event, not a time period. Quiet markets produce no new columns; active markets produce many. This makes the chart compact for consolidation and expressive for trend — the visual width is proportional to market activity, not elapsed time.
Box size and reversal amount are the two parameters the analyst controls, and they determine what the chart considers "signal" versus "noise." A smaller box size (e.g. $1) makes the chart more sensitive: more symbols per column, more columns, more reversals detected. The chart becomes wider and reveals more structure — useful for short-term trading.
A larger box size (e.g. $5) filters out minor fluctuations, producing fewer, taller columns that represent only major price movements. This is better for detecting long-term trends and major support/resistance levels. The reversal multiplier (typically 3×) means a new column only starts when the price reverses by three full box sizes — filtering out brief counter-trend moves.
This implementation exposes both parameters as sliders. Dragging box size from $1 to $6 on the same data shows how the same price series can produce a busy, detailed chart or a clean, structural one — the "right" setting depends entirely on the trading horizon and tolerance for noise.
A Candlestick Chart encodes the same OHLC data but plots every session on the time axis regardless of whether meaningful price movement occurred. During consolidation phases, candlesticks produce a series of small, near-identical bodies that convey no directional information but consume significant chart space. The P&F chart collapses all quiet periods to nothing and expands only at decisive supply-demand shifts.
A Line Chart on closing prices discards open, high, and low — and still suffers the time-scale distortion problem. It cannot identify support and resistance levels as cleanly as P&F because those levels emerge from the structural alignment of column tops and bottoms, which the time-independent column layout makes visually apparent.
FT Visual Vocabulary — Change Over Time (specialist variant, time-independent). The P&F chart belongs to the Change Over Time category but deliberately rejects the time axis in favour of a price-movement axis — this is the defining design decision of the form. Abela quadrant: Relationship (supply vs demand balance) nested inside Change. Tufte principle applied: maximum noise reduction through parameterised filtering — the box size and reversal amount are explicit data-ink controls that the analyst sets to match the signal they are trying to see. The one design decision worth knowing: support and resistance lines are derived algorithmically from prior column extremes (X column tops become resistance; O column bottoms become support), not drawn manually — they emerge from the chart's structure rather than being imposed on it.