Candlesticks vs Heikin-Ashi: Traded Prices vs Averages (SPY Data)

Michael Whitman · Trading since 2017, full-time since 2020 · YouTube creator, 137K subscribers · Published 2026-09-09 · Updated 2026-11-02 · Reviewed 2026-11-01

Candlesticks draw the open, high, low and close that actually traded; heikin-ashi averages them into a smoother bar whose open, high and low are often prices that never traded. On SPY, that held on weekly, daily, hourly and 5-minute charts: the heikin-ashi open sat outside the real bar’s range on about half of all bars.

What each one is

A candlestick draws the period’s open, high, low and close. The body runs from open to close and the wicks reach the high and the low, so every edge is a price that traded. Candlesticks covers the anatomy and how color is decided.

A heikin-ashi bar is built from those same prices with two averages. Its close is the mean of the open, high, low and close. Its open is the midpoint of the previous heikin-ashi bar’s open and close, which carries yesterday’s smoothed body into today. Its high and low are whichever is more extreme: the real high or low, or the bar’s own calculated open and close. Heikin-Ashi walks through all four formulas.

So the two differ in what a number on the chart means. On a candlestick, a price is an event. On heikin-ashi, the close is an average of events, and the open is an average of averages.

Where they differ

Whether the edges traded. The heikin-ashi close always lands inside the day’s range, because it is an average of four prices from that range. The open does not have that property. Measured on SPY, it sat above the day’s high or below its low on 55.93% of sessions, and on each of those days the bar’s high or low was that calculated open rather than a traded extreme.

How often the color changes. A real candle is colored by its own open and close, which flip often. A heikin-ashi bar compares a smoothed close with a smoothed open, so the color holds for longer runs. That is the visible appeal of the chart, and it comes from the formula rather than from the market.

Whether a gap can appear. The heikin-ashi open is the midpoint of the previous bar’s body, which is always inside the previous bar’s range, so a heikin-ashi bar can never open beyond the bar before it. SPY printed real gap days throughout the sample; the heikin-ashi version of the same chart shows none.

When each one turns. A candlestick shows a reversal day as soon as it closes. A heikin-ashi bar changes color only after the averages cross, which takes the turn day plus whatever the smoothing still remembers.

Reading one day both ways

Take SPY on 2 Oct 2026. The real candle opened at $770.58, traded between $767.15 and $772.65, and closed at $769.64. The close was below the open, so the candle is a down candle.

The heikin-ashi bar for the same day says the opposite. The previous heikin-ashi bar’s open and close were $765.80 and $763.20 (rounded), so today’s heikin-ashi open is their midpoint, $764.50. The heikin-ashi close averages the four real prices: ($770.58 + $772.65 + $767.15 + $769.64) / 4 = $770.005. The close is above the open, so the bar is an up bar.

Its low is not a price SPY touched that day. The heikin-ashi low is the lowest of the real low ($767.15), the calculated open ($764.50) and the calculated close, which makes it $764.50, about $2.65 below anything that traded on 2 Oct. A stop placed under that bar would be measured from a level the session never reached.

Five sessions, worked through both ways

One day hides how the averaging builds up, so here are the five sessions to 2 Oct. Each heikin-ashi open below is the midpoint of the row above it: on 29 Sep, ($768.75 + $766.80) / 2 = $767.78. That open sat above the real high of $766.98, so the heikin-ashi high that day, $767.78, is a price SPY never reached.

Notice the 30 Sep row. The real candle closed $3.82 under its open, a clear down day, yet the heikin-ashi close of $765.17 still came out above the averaged close of the day before. The bar stayed red only because its open, $766.43, was carried in from the bar before it.

Across a longer stretch the same pattern repeats. From 14 Sep to 30 Oct 2026, 35 sessions, SPY’s real candles changed color 17 times and the heikin-ashi bars 12 times. The heikin-ashi open sat outside the day’s real range on 20 of the 35 days, close to the long-run share in the data section below.

Read the two panels against each other. The lower one holds its color through stretches where the upper one alternates, and each of its bars opens at the midpoint of the body before it, so none can open beyond the bar to its left. Every number in the table, and every bar in both panels, is in the September-October file.

Which one to use: Candlesticks or Heikin-Ashi?

Use candlesticks when you are placing an order. An entry, a stop and a target are prices, and only the real candle reports prices that traded. On more than half of SPY sessions, the heikin-ashi high or low was not one of them.

Use heikin-ashi when the problem is reacting to every single bar. If one red candle inside a rising stretch keeps pushing you out of a position you planned to hold, a chart whose color changed half as often is a legitimate way to see the stretch as a whole, provided you accept that it shows turns late.

Why the smoothing costs time

Each heikin-ashi open carries half of the previous bar’s body forward. That memory is what holds a color for several bars, and it is also why a turn shows up late: a bar has to overcome the averaged past before its color can change.

Measured at each heikin-ashi color change on SPY, the turn to up came a median of one session after the lowest low of the down run before it, with the close on the turn day a median 1.75% above that low. Turns to down came a median of one session after the run’s highest high, 1.39% below it. Those are distances paid before the color confirms anything.

Neither number is a forecast: both are measured in hindsight, against an extreme known only once the run ended.

The original data

The sample is SPY’s daily bars from 2 Jan 1996 to 2 Oct 2026, 7,739 sessions (Yahoo Finance, split-adjusted). Opens start in 1996 because SPY’s earliest opens often just repeat the prior close (57 times in 1993-94). Real candles: 4,068 up, 3,619 down and 52 with the close exactly equal to the open. Heikin-ashi bars: 4,505 up and 3,234 down, none flat.

The heikin-ashi open landed outside the real day’s range on 4,328 of 7,738 sessions (55.93%): above the high on 1,617 (20.90%) and below the low on 2,711 (35.03%). The heikin-ashi high or low differed from the real extreme on exactly those sessions, because the close never leaves the range. The median gap between the heikin-ashi open and the real open was 0.57% of the price; for the close, the median gap was 0.25%.

Real gaps: 728 sessions (9.41%) traded entirely above the previous day’s high (466) or below its low (262). Heikin-ashi gaps: 0, as the construction requires. SPY also opened beyond the prior day’s range on 2,811 sessions (36.33%), every one of them absorbed into a heikin-ashi open drawn from the bar before.

What this does and does not show. The table measures how the two chart types draw the same market. It says nothing about whether either color helps anyone trade; no rule was tested here. Every bar, both ways, is in the SPY candles and heikin-ashi file.

Does a shorter or longer timeframe change it?

Fresh SPY downloads on four timeframes give the same answer. The same formulas were run on Yahoo Finance weekly, daily, 60-minute and 5-minute bars, each series seeded on its first bar, which is then left out of the counts. Intraday heikin-ashi bars carry on across the overnight break, as a continuous chart draws them.

On every timeframe, between 46.70% and 55.93% of heikin-ashi opens were not a price inside the bar they sit on. Switching to a 5-minute chart or a weekly one does not turn the smoothed open into a traded price. The color also changed on roughly a quarter of bars everywhere (22.79% to 25.80%), against about half for real candles (48.56% to 51.57%). The daily row runs 16 sessions past the table above and gives the same 55.93%. The 5-minute sample is twelve weeks of one fund, small next to thirty years of daily bars.

Checked against Nasdaq’s SPY history, which rounds to the cent: on the 2,512 daily bars from 26 Oct 2016 to 26 Oct 2026 (the 20 Apr 2026 row, one price with no volume, left out), the heikin-ashi open sat outside the range on 57.91% of Nasdaq bars and 57.95% of Yahoo bars. Weeks built from Nasdaq’s daily bars, 521 from 31 Oct 2016, gave 55.00% against 54.81% for Yahoo’s own weekly bars. Every count is in the timeframe file.

When it fails

The costly failure is taking a stop or entry from a heikin-ashi extreme. On 2 Oct 2026 the bar’s low sat $2.65 under the real low. A stop placed a few cents beneath that heikin-ashi low is wider than the trader measured from the real chart, the position size worked out from it is off by the same amount, and none of it shows until the order fills at a real price somewhere else.

A second failure is reading a long color run as strength. Runs of five or more were 31.06% of heikin-ashi runs and 5.76% of candle runs on the same days. The difference is the averaging, not a stronger market.

A third is hunting candlestick patterns on smoothed bars. An engulfing bar or a doji needs a real open and close.

A fourth is backtesting on heikin-ashi values. Fills at prices that never traded produce results no account could have had.

Questions people ask

Is heikin-ashi better than candlesticks?

For reading prices, no: only candlesticks show the open, high, low and close that traded. Heikin-ashi is easier to read for direction, because its color changed on about a quarter of SPY bars against about half for real candles, but that calm comes from averaging, and turns show up late.

Why do heikin-ashi candles show prices that never traded?

The heikin-ashi open is the midpoint of the previous heikin-ashi bar’s body, an average of averages. On SPY it sat above the real high or below the real low on 55.93% of days since 1996, and on those days the heikin-ashi high or low was that calculated open.

Can you see gaps on a heikin-ashi chart?

No. Each heikin-ashi open is drawn from inside the previous bar’s body, so the chart cannot show a gap. SPY had 728 real gap days from 1996 to 2 Oct 2026; the heikin-ashi version showed none.

Is heikin-ashi a lagging indicator?

Yes, by construction: each open averages the previous bar. On SPY since 1996 the color turned up a median of one session after the down run’s lowest low, with the close by then a median 1.75% above it.

Is heikin-ashi good for day trading?

It can show the direction of the last few bars on a 5-minute chart, but entries and stops still need real prices: on SPY 5-minute bars the heikin-ashi open sat outside the real range on 46.70% of bars.

Related

Candlesticks explains the real open, high, low and close and how a candle gets its color. Heikin-Ashi sets out the four averaging formulas behind every bar measured here. And price action is the practice of reading the real bars that heikin-ashi smooths away.

This page is educational, not financial advice. Test every idea on your own charts before risking money.