Heikin Ashi vs Renko
Heikin ashi averages each candle so a trend looks cleaner while the time axis stays intact. Renko discards time altogether and prints a brick only when price moves a set distance, so both charts show values that were never actually traded.
Two ways of making a chart calmer to look at. They achieve it differently, they lose different information, and they share the one property that matters most: neither shows prices that traded.
What each one is
Heikin ashi averages each candle against the one before it. The result is a smoother sequence of bodies with fewer contradicting bars, plotted on the ordinary time axis. Heikin ashi covers the calculation.
Renko removes time completely. A brick prints only when price has moved a chosen distance, so a quiet hour produces nothing and a fast one produces several. Renko charts covers it.
Both are derived from the same underlying bars. Neither adds information; both remove some, and what they remove is the part that makes a chart hard to read.
Where they differ
Whether time survives. Heikin ashi keeps every interval, so a quiet period is visible as quiet. Renko deletes it, and a flat afternoon simply does not appear on the chart at all.
What sets the smoothing. Heikin ashi has no parameter — the averaging is fixed. Renko has a brick size, and that single number changes the chart more than any indicator setting changes an indicator.
How late each turns. Heikin ashi lags because it averages. Renko lags because a reversal needs enough movement to print an opposing brick, which is a distance rather than a delay.
What each hides. The averaged version hides indecision inside the bar. The step version hides duration entirely, so you cannot tell whether a move took a minute or a week.
Where they agree
Neither shows a traded price. The values on both charts are constructed, so an order placed at a level read off either one is placed at a number nobody transacted at.
Both flatter hindsight. Scroll back through either and the trends look obvious and the entries look easy, because the bars that would have shaken you out have been smoothed or skipped.
Both need the real chart for execution. Entry, stop and target come from the unsmoothed series, and on this site’s shared series the ninetieth percentile bar range is 1.101.
And both cost a round trip per trade — about 2% of the median bar range of 0.493 here — which the smoothing does nothing to reduce.
Which one to use
Use heikin ashi when you want smoothing but need the clock. Anything scheduled — a session open, a data release, a close — requires a time axis, and the step chart cannot give you one.
Use renko when time genuinely does not matter to your method. If your rules are about distance travelled rather than about when, deleting the axis removes a distraction rather than information.
Use renko when a quiet market is drowning your chart in bars that contain nothing. Compressing them to a single brick is a legitimate use of it.
And in either case take the entry from the real chart. Reading a level off a constructed price is the one error that turns both of these from a display choice into a losing habit.
Why the brick size decides everything
Because it sets how much movement counts as an event. A small brick produces a busy chart and frequent signals; a large one produces a calm chart and few. The market did not change.
And because it is usually chosen by appearance. People pick the size that makes the past look cleanest, which is choosing a parameter by fitting it to the chart you already have.
Using either one alongside the real chart
Keep the unsmoothed series open beside it. The smoothed chart is a reading aid; the other one is where your orders fill, and having both visible removes the temptation to place a level off the wrong picture.
Mark levels on the real chart and read direction on the smoothed one. That division of labour uses each for what it is good at, and it is the only arrangement in which the constructed prices cannot reach your order ticket.
Check the trade count both ways before trusting a rule. The same rules produce a different number of entries on a smoothed series than on the raw one, and if you have not counted both you do not know which number your expectation is based on.
And keep one chart type across a whole record. Switching after a losing run produces a new set of signals and destroys the ability to judge the old set, which is how a display preference quietly turns into a method with no evidence behind it.
The original data
Of the 24,971 unique videos in research/search-study-corpus.jsonl, 1 compares the two directly in the
title, at 1,228 views. Separately, heikin ashi appears in 52 titles at a median of 33,639 across 45
channels, and renko in 72 at a median of 5,347 across 41. The counts come from site/rank_compare.py
and site/corpus_count.py.
52 videos on the averaged version at 33,639 against 72 on the step version at 5,347. Fewer videos and more than six times the audience per one — the chart that keeps its time axis is the one people actually watch, despite the other having more coverage.
The answer to the question on that chart is that the level you are looking at does not exist. Switch to the unsmoothed series before placing anything — the trend may well be real, and the price in front of you is not.
When it fails
The failure is backtesting on either chart, and the results are not achievable. A smoothed series produces a run of clean trades in review because the bars that would have stopped you out were averaged or skipped. The equity curve looks excellent. Run the same rules on the unsmoothed data and the trade count changes, the stops are hit, and the entries fill at different prices — because the chart the test ran on displays values that were never available to trade.
The second failure is placing an order at a constructed price. It never traded.
A third is choosing a brick size by how the past looks. That is fitting.
A fourth is using renko for anything time-sensitive. There is no clock.
A fifth is reading a smoothed chart as fewer reversals. They are hidden, not absent.
And a sixth is switching chart type after a losing run. The record becomes unreadable.
Related
Heikin ashi covers the averaged candles. Renko charts covers the fixed-distance bricks. And candlesticks covers the unsmoothed chart your orders actually fill on.
Both of these are honest about what they do and both get used dishonestly. A smoothed chart in review looks like a method that would have been easy to follow. The reason it looks that way is that the noise you would actually have had to sit through has been removed from the picture.
— Michael Whitman
This page is educational, not financial advice. Test every idea on your own charts before risking money.