Point & Figure (P&F) is one of the oldest Western charting methods, plotting price as vertical columns of Xs (rising prices, demand) and Os (falling prices, supply) instead of a continuous price-vs-time line. Two parameters govern the whole chart: the box size (the price increment each X or O represents) and the reversal amount (how many boxes price must move against the current column to start a new, opposite column). Its defining property is that it is time-independent — only price moves large enough to fill a box ever print, so flat days, intraday chop, and minor fluctuations are filtered out entirely. The result is a deliberately de-noised view of supply/demand and clean, rule-based support/resistance. Its core tension: the same noise-filtering that gives clarity also depends on discretionary box and reversal settings, and the method's signals carry long usage but thin rigorous evidence of a standalone edge.

How it's formed

A new box prints only when price moves a full box size in the direction of the current column. The current column (say Xs) keeps extending up as long as price does not reverse by the reversal distance = box size × reversal amount. Once price falls by that full distance, the chart steps one column to the right and begins a column of Os (and vice versa). No move large enough means no new mark at all — which is why a single column can span one day or many weeks.

• Reversal amount: The 3-box reversal is the most popular method and StockCharts' ChartSchool default; A.W. Cohen popularized this "Three-Point Reversal Method" (with Xs and Os) and edited ChartCraft, the long-running P&F service. With box size 1 and reversal 3, a 3-point counter-move is needed to flip columns (1 × 3). 1-box reversals are more sensitive (more columns); higher reversals filter more.
• Box size / scaling: Set by traditional fixed increments, a percentage scheme, or ATR-based sizing. A larger box filters more movement and prints fewer reversals; a smaller box is noisier with more reversals. This choice is the chart's main discretionary lever.

How to read it

• Double-Top Breakout (basic buy signal): an X-column rises one box above the high of the prior X-column — demand overcomes the prior resistance.
• Double-Bottom Breakdown (basic sell signal): an O-column falls one box below the low of the prior O-column — supply overcomes the prior support.
• More complex named patterns (triple tops/bottoms, bullish/bearish catapults, signal-line variants) build on those primitives.
• Trend lines: P&F draws objective 45° Bullish Support Lines (up) and Bearish Resistance Lines (down) from chart geometry, giving a non-discretionary trend filter.
• Price objectives: P&F can project targets:

- Vertical count (bullish): column-of-Xs box count × box size × reversal amount, added to the column low. ChartSchool example: 10 Xs on a 1×3 chart → 10 × 1 × 3 = 30 boxes above the low. - Vertical count (bearish): StockCharts uses Dorsey's variant — box count × box size × 2/3 of the reversal amount (i.e. ×2 on a 3-box chart) — subtracted from the column high. - Horizontal count: projects off the width of a congestion/base — a wider base implies a larger target; preferred after large bases.

How it's used in practice

Traders use P&F to (1) strip noise and locate objective support/resistance and breakout/breakdown levels for entries and stops; (2) read trend via the 45° trend lines; and (3) set rough price targets via vertical/horizontal counts. A distinct institutional-leaning application is relative-strength P&F: Thomas Dorsey (Dorsey, Wright & Associates, later Nasdaq Dorsey Wright) built a following running P&F on a stock-vs-benchmark ratio to rank leaders/laggards for sector rotation and relative-strength screening. The style-specific entry/stop/hold mechanics belong to the Swing Trading branch — cross-link there rather than duplicating here.

Standing & evidence

P&F is genuinely old (P&F-style price tabulation traces to late-1800s Wall Street operators; "Hoyle" referenced it in 1898) and has had a continuous professional following — ChartCraft/Cohen, then Dorsey Wright's relative-strength franchise, gave it real institutional reach. But adoption is not the same as a measured edge: the breakout/objective signals are widely used yet have thin rigorous empirical support, far less studied than candlestick or classic bar-chart patterns. Crucially, every output is conditional on the box/reversal settings — the same series on different box sizes yields different signals and different targets, so any backtest is parameter-sensitive and prone to overfitting. The count-based targets are explicitly guidelines, not guarantees: StockCharts cautions they should be "taken with a grain of salt," as prices often miss them or reverse first. Treat P&F as a clean framing tool, not a proven predictive one.

Strengths & limitations

• Strengths: removes time/noise so support, resistance, and breakouts are visually unambiguous; rule-based 45° trend lines and double-top/bottom signals reduce interpretation; works across timeframes since it's price-driven.
• Limitations: box/reversal settings are discretionary and change the signals — the #1 misuse is treating one parameterization's output as objective truth or cherry-picking the box size that "looks right." It discards volume and time, so it misses time-based and volume context; in low-volatility drift or gappy/illiquid names the filtering can lag or whipsaw. Price objectives are soft estimates, not forecasts.

Sources

• StockCharts ChartSchool — Introduction to Point & Figure Charts (Xs/Os, box size, reversal distance, 3-box reversal as most popular, time-independence)
• StockCharts ChartSchool — P&F Scaling and Timeframes; P&F Bullish Breakouts / Bearish Breakdowns (double-top/bottom signals)
• StockCharts ChartSchool — P&F Price Objectives: Vertical Counts and Horizontal Counts (count formulas, Dorsey bearish 2/3 variant, "grain of salt" caveat)
• Wikipedia — Point and figure chart (history: 1898 "Hoyle", A.W. Cohen / ChartCraft Three-Point Reversal, Chartcraft IBM S/360 in the 1960s)
• Thomas J. Dorsey, Point and Figure Charting (Wiley) — relative-strength P&F application (publisher/retailer listings)