Skip to main content

Renko & Heikin-Ashi

Updated Jun 23, 2026 at 8:47pm

Research Draft Medium 1,124 words

Two noise-smoothing alternative chart types built on the same premise: ordinary OHLC candles contain a lot of intrabar "chop" that obscures the trend, so transform the data to make trends visually obvious. Renko abandons the time axis entirely, plotting fixed-size "bricks" that print only when price travels a set amount. Heikin-Ashi ("average bar" in Japanese) keeps the time axis but redraws each candle from averaged OHLC values so that consecutive candles share color and lose their wicks during a clean trend. Both are excellent at one job — telling you whether a trend is intact — and dangerous at another: neither shows the real prices that actually traded, so you cannot read a true entry, stop, or fill off either one. That tension is the whole point of this node.

How it's calculated / formed

Heikin-Ashi — each candle is computed from the current real bar and the prior HA candle. The formulas are consistent across Wikipedia, Corporate Finance Institute, and Investopedia:

  • HA-Close = (Open + High + Low + Close) ÷ 4 — the average of the current real bar's four prices.
  • HA-Open = (previous HA-Open + previous HA-Close) ÷ 2 — the midpoint of the prior HA candle's body.
  • HA-High = max(real High, HA-Open, HA-Close).
  • HA-Low = min(real Low, HA-Open, HA-Close).

Because HA-Open depends on the previous HA candle, the series is recursive and must be seeded on the first bar (commonly HA-Open = (Open + Close) ÷ 2 of the first real bar, or simply the first bar's real open). This recursion is the source of both its smoothing and its lag.

Renko — a new brick is drawn only after price moves at least one box size beyond the prior brick's far edge; smaller moves are ignored. Bricks are uniform height and conventionally drawn at 45°, one per column. Two box-sizing methods (per StockCharts ChartSchool):

  • Fixed: a constant amount (e.g. $1, 10 points). Brick size never changes.
  • ATR: box size is set from the symbol's Average True Range (default 14 periods). It produces variable-looking bricks. Critically, StockCharts notes the ATR value — and therefore the brick size — is recomputed as new data arrives, so the entire historical chart can redraw when ATR shifts.

Renko has no time axis: a brick may take minutes or weeks to form. Plain Renko also ignores price gaps — it walks the brick count up or down as if price moved continuously through the gap.

How it's used in practice

Both are used the same way at a high level — trend identification and trend-following discipline, not precision timing:

  • Heikin-Ashi: traders read color and wicks. A run of green candles with no lower wicks signals a strong, healthy uptrend (mirror for downtrends with no upper wicks); the longer the wickless run, the stronger the trend is presumed to be. A small-bodied candle with wicks on both sides (a HA "doji") flags loss of momentum and a possible reversal. The intended discipline is "stay in the trade until the candles change character" — it discourages exiting on a single noisy bar.
  • Renko: traders treat a same-color brick sequence as trend continuation and a color flip (one brick the opposite direction, or two on stricter rules) as a trend-change signal. Because counter-trend noise smaller than the box is filtered out, Renko is popular for trailing trend trades and reducing overtrading. Diagonal trendlines and support/resistance read cleanly off the brick stair-steps.

For both, the operational swing mechanics — exact entry trigger, stop placement, target, and hold period — belong to the Swing Trading branch, not here. This node defines the chart types; it does not prescribe a trade plan.

Strengths & limitations

Strengths. Genuine, well-documented signal-to-noise improvement. Trends that look ragged on candles look clean here, which helps traders hold winners and avoid whipsaw exits. Renko's time-independence collapses long flat periods and stretches active ones, surfacing structure. Both can de-bias an over-reactive trader.

The critical honesty caveat — read this before trusting either chart. Both hide and distort the actual price:

  • Heikin-Ashi candles do not show real open or close prices. HA-Open is a midpoint of the prior synthetic candle and HA-Close is an average of four real prices — neither traded. You cannot read your fill, your stop, or the real day's close off an HA candle. The wickless "strong trend" look is partly an artifact of the averaging, not pure market reality, and HA also drops price gaps (the smoothing erases them). It is also laggy by construction — every candle carries forward the prior candle — so reversal signals arrive late.
  • Renko bricks lag and can repaint. A brick prints only after a full box of movement, so in a fast move the brick (and any signal off it) is late. Worse, because many platforms build Renko retrospectively from a chosen anchor, and because ATR-based box sizes recompute as new data arrives, the historical brick sequence can change or "backfill" after the fact. This makes naive Renko backtests unreliable (you're testing a repainted past that never happened live) and means the chart you see now is not guaranteed to be the chart you saw an hour ago.

Net: use both as a trend lens, never as a price tape. The #1 misuse is reading entries, stops, or targets directly off HA/Renko levels — always confirm true price on a standard candlestick/OHLC chart before acting. Both also share the general failure mode of trend tools: they shine in trending regimes and chop badly in ranges (Renko less so, since sub-box noise is filtered, but it then lags the turn).

System relevance

This is a chart-type definition node under Technical Analysis → Chart Types. It sits beside Point & Figure (the other classic time-independent, noise-filtering chart) and the standard Candlestick node — cross-link those rather than duplicating. Swing-specific entry/stop/target mechanics live in the Swing Trading branch. For any Delvantic analysis pipeline (e.g. the Augustus trade-setup agent), the operative caveat is hard: HA/Renko outputs are trend-context signals only; true entry, stop, and fill prices must come from real OHLC, never from HA candles or Renko bricks.

Sources