Heikin Ashi, and what averaging removes
The answer
Heikin Ashi is a candlestick chart where each candle is built from averaged prices instead of real ones. It makes a trend far easier to see, and the cost is exact: the open and close it shows you are numbers nobody ever traded at, and every gap in the price history disappears.
Why this costs you money
A trader watches a Heikin Ashi chart during a strong move. The candles are a run of clean green bodies with no lower wicks. The chart is calm and the trend is obvious. He decides to add to his position at the current price, which the chart shows as 1,240.
He places the order and it fills at 1,268.
Nothing went wrong with the broker. The Heikin Ashi close of 1,240 is the average of the day's open, high, low and close. The actual last traded price was 1,268. The chart was never showing him a price. It was showing him an average, drawn in the shape of a price, on an axis labelled in rupees.
That is a 2.3% error on entry, and it is entirely avoidable.
The second cost is larger and slower. Heikin Ashi hides gaps completely. In the Indian market, an overnight gap is the one risk a trader genuinely cannot manage, because there is no trading between 15:30 and the next morning's pre-open. A chart that erases gaps is erasing exactly the risk you cannot do anything about, and after 6 months of looking at it, a trader stops believing that gaps happen.
Every smoothing method removes something. The mistake is not using them. The mistake is not knowing precisely what was removed.
How it works
The formula
Heikin Ashi means "average bar" or "average pace" in Japanese. Each candle uses 4 calculations.
| Value | Formula |
|---|---|
| HA Close | (Open + High + Low + Close) ÷ 4, from the real candle |
| HA Open | (previous HA Open + previous HA Close) ÷ 2 |
| HA High | the highest of: real High, HA Open, HA Close |
| HA Low | the lowest of: real Low, HA Open, HA Close |
Read those 4 lines carefully, because everything about the chart follows from them.
The HA Close is an average of 4 prices. It sits inside the day's range, never at the extreme. So the body is always narrower than the real body.
The HA Open depends on the previous HA candle, not on today's opening price. This is the important one. Today's HA Open is the midpoint of yesterday's HA candle. Today's real opening price does not appear in the calculation at all.
Because of that, today's HA Open always sits inside yesterday's HA body. Two consecutive HA candles always touch. A gap is arithmetically impossible on a Heikin Ashi chart, regardless of what the market did.
The high and the low do use the real high and low, so the extremes of the day survive, though they can be stretched by the averaged open and close.
Why it is recursive, and what that means for you
Each HA Open depends on the previous HA candle, which depends on the one before, and so on back to the first candle on the chart.
That first candle needs a starting value, usually the midpoint of the first real candle. So the entire chart depends on where it starts.
In practice the influence of the starting value decays quickly, within perhaps 10 to 20 candles. But it does mean 2 platforms can draw slightly different Heikin Ashi charts of the same stock if they load different amounts of history. If you ever see 2 Heikin Ashi charts disagree on a candle's colour, this is why.
How to read it
The reading is simple and this is its real strength.
A healthy trend shows consecutive same-coloured candles with reasonably long bodies and no wick on the opposing side. In an uptrend, no lower wicks. In a downtrend, no upper wicks. That absence is the signal: it means the averaged low never fell below the averaged open, which happens when a move is one-directional.
A weakening trend shows bodies getting shorter, and wicks appearing on both sides of the body.
A possible turn shows a small body with wicks above and below, similar to a doji on an ordinary chart, often followed by a colour change.
Notice what this is doing. Heikin Ashi is not adding an indicator. It is taking information you already had and presenting it so that the trend is the loudest thing on the screen. That is a legitimate and useful thing to do.
The lag, which is the price of the smoothing
Because HA Open carries yesterday forward, a Heikin Ashi chart changes colour after the real chart does. Typically 1 to 3 candles later, depending on how sharp the reversal is.
That is the whole trade-off in one sentence. You are exchanging the beginning of the move for the ability to sit through the middle of it.
For a trend follower that can be a good exchange. For anybody trading a reversal, it is a disaster.
The backtest trap
This is the most expensive misunderstanding about Heikin Ashi and it applies equally to Renko in the next article.
If you run a backtest that buys at the Heikin Ashi open and sells at the Heikin Ashi close, your results will look extraordinary. They will also be fiction, because those 2 prices did not exist. Nobody could have transacted at either of them.
Any Heikin Ashi strategy must be tested with entries and exits at real prices — typically the real open of the following session. Most published Heikin Ashi results do not do this, and the difference between the 2 versions is usually the difference between an excellent system and a losing one.
What it tells you, and what it does not
Heikin Ashi tells you about persistence. A run of 12 green candles with no lower wicks is a genuine statement that the closes have been consistently above the running average of recent activity.
It does not tell you a price. Not one number on the chart is a price you can act on. Read your entries from the ordinary candlestick chart, always.
It does not tell you about gaps, because it cannot represent them.
It does not tell you about a single dramatic session. A day with a violent reversal, which on an ordinary chart is a very visible candle, becomes an ordinary looking HA candle with 2 wicks. The drama is averaged away, and sometimes the drama was the information.
The decision rule
Use Heikin Ashi to decide whether to stay. Use ordinary candles to decide where to act.
Keep both open. Heikin Ashi answers "is this trend still intact", which is a question about the last 10 sessions. The ordinary chart answers "what price am I buying at and where is my stop", which is a question about today.
If you are ever about to place an order using a number read off a Heikin Ashi chart, stop. That number is an average.
Try this now
This takes about 5 minutes and it shows you precisely what the averaging removed from a chart you already know.
- Open a daily candlestick chart of a stock you follow. Set to 6 months.
- Find the 3 largest gaps — days where the opening price was clearly away from the previous close. Write down those 3 dates.
- Switch the chart type to Heikin Ashi, same stock, same 6 months. Go to your 3 dates. The gaps are gone. Not reduced. Gone. The candles touch.
- Now switch back to candlesticks and find a day with a large range. Note the real closing price for that day, from the price axis or the data readout.
- Switch to Heikin Ashi and read the HA close for the same day. Write down both numbers and the percentage difference between them.
- Last step. On the Heikin Ashi chart, count how many days in the 6 months were green on Heikin Ashi but red on the ordinary chart, or the reverse. You can do this by flipping between the 2 chart types on the same screen region.
What you should see. Three things, and every one of them is the point of the article.
The gaps vanished completely. On a volatile stock this can remove the single largest price event in 6 months from the chart.
The HA close differs from the real close, usually by a fraction of a percent on a quiet day and by 1% to 3% on a volatile one. That difference is what you would have paid if you had traded from the wrong chart.
There are more colour disagreements than you expected, often 15% to 25% of sessions. Each of those is a day where the market closed down and your chart is green, or the reverse. That is not an error. It is the smoothing doing its job, and you should know how often it happens before you rely on it.
Three real cases
1. Adani Enterprises, late January and early February 2023 (India) — the period where averaged prices were furthest from real ones On 24 January 2023 Hindenburg Research published a report on the Adani group. Over the following days Adani Enterprises fell very sharply, with large gaps between sessions and enormous intraday ranges. A follow-on public offer was fully subscribed on 31 January and withdrawn on 1 February 2023, with the money returned to investors.
Draw that period as Heikin Ashi. You get a clean run of red candles that reads as an orderly downtrend. Draw it as ordinary candles and you get gaps, reversals and sessions where the range was a large fraction of the price. The Heikin Ashi version is easier to read and it describes a market that did not exist. During those sessions the difference between an HA close and a real transactable price was very large, and a trader working from HA numbers would have been wrong about their entry by a great deal.
2. March 2020, the COVID crash and the turn (India and the United States) — the lag, measured Both the Nifty 50 and the S&P 500 fell heavily through February and March 2020 and turned upward in the second half of March 2020. Do this on your own chart rather than taking my word for it. Load either index for January to May 2020 as ordinary candles, note the date of the first strong up candle after the low. Then switch to Heikin Ashi and note the date of the first green HA candle.
You will typically find the Heikin Ashi turn arrives 2 or 3 sessions later . In an ordinary market 3 sessions is a small cost. In March 2020 the index moved a very large percentage in 3 sessions. This is the clearest available demonstration that the lag is not a constant cost; it is largest exactly when the market moves fastest.
3. Dan Valcu, 2004 (United States and Europe) — how it reached the West Heikin Ashi was largely unknown to Western traders until Dan Valcu published an article describing the technique in Technical Analysis of Stocks & Commodities . Almost every Western description of Heikin Ashi since then traces back to that article and the formula it set out.
This matters for a practical reason. Unlike momentum or trend following, Heikin Ashi has almost no independent academic literature testing it. It is a display method that was adopted because it looks clear, not because a body of evidence supported it. That is not a reason to avoid it. It is a reason to test it yourself rather than assume somebody else already did.
The question that resolves it
A novice asks: is Heikin Ashi better than candlesticks?
An expert asks: what did the average delete, and do I need that thing for the decision I am about to make?
The averaging deleted the exact open, the exact close, and every gap. If your decision is "should I still be holding this", none of those 3 matter and Heikin Ashi is the better chart. If your decision is "what price do I pay and where do I put my stop", all 3 matter and Heikin Ashi is unusable.
What would make this wrong
If Heikin Ashi produced no improvement in a trader's ability to stay in trends, the whole case for it would collapse, because staying in trends is its only claim. That is testable and you can test it: run the same trend rule on ordinary closes and on HA closes, exit at real prices in both cases, and compare how long the average winning position was held.
The honest limits are 3.
First, Heikin Ashi may simply be a slower moving average with a nice display. Its smoothing is not obviously better than a short exponential moving average of the close, and I am not aware of a study showing that it is.
Second, "easier to read" has a cost that is hard to measure. A chart that makes trends look clean also makes weak trends look clean, and a trader who is comfortable stays in positions longer, including the wrong ones.
Third, the whole method depends on the trend continuing. In a sideways market Heikin Ashi produces alternating short candles with wicks on both sides, which is correct information and also useless information. It will not tell you that you are in a range until you have already taken several losing signals.
In India
Gaps matter more here, so erasing them costs more. Between 15:30 and the next pre-open there is no trading in an Indian single stock. An overnight gap is therefore a price move that happened while you had no ability to act. It is the purest form of unmanageable risk in the Indian market, and Heikin Ashi removes all visual evidence that it occurred.
Price bands compound the problem. A stock locked at its lower band produces a real candle with a flat bottom. Run that through the Heikin Ashi formula and the flat bottom is averaged into an ordinary-looking body. Two separate mechanisms have now hidden the fact that the price was not free to trade.
Expiry sessions. Weekly and monthly index option expiries concentrate an enormous share of Indian derivatives volume. Heikin Ashi on an intraday Nifty chart during an expiry session is a smoothed picture of settlement mechanics, and smoothing it does not make it more readable. It makes it look like a trend.
Practical setting. On most Indian platforms and on TradingView, Heikin Ashi is a chart type rather than an indicator, which means your drawings, alerts and any strategy attached to the chart may use HA values rather than real ones. Check this. An alert set at a Heikin Ashi level will fire at a price that does not exist.
In the United States
The extended-hours setting interacts with Heikin Ashi in a way that surprises people. With extended hours off, an earnings release produces a gap. Heikin Ashi removes it. With extended hours on, the same move appears inside the candles and Heikin Ashi smooths it across several bars. Either way, the single most informative event of the quarter is softened into an ordinary looking sequence.
US futures make Heikin Ashi more usable. Index futures trade nearly around the clock, so gaps are small and rare, and the information Heikin Ashi deletes is much less valuable. This is a real reason the technique is popular among US futures traders and less suited to Indian single stocks. The method did not change. The instrument did.
Volume is unaffected. Heikin Ashi transforms price only. The volume bars under an HA chart are the real volume figures, which means volume is your one remaining link to what actually happened. Use it. A large volume bar under an ordinary looking HA candle is telling you that something happened that the price averaging concealed.
Where they differ, and what that tells you
Heikin Ashi deletes gaps in both markets. What a gap is differs, and that changes what the deletion costs you.
In the United States, a gap between 2 regular-hours candles usually has real transacted prices inside it, in the pre-market session. Deleting it hides a record of trading that happened somewhere else. It is a loss, but the information exists and you can go and look at it by switching on extended hours.
In India, a single-stock gap is genuinely empty. Nobody traded inside it. Deleting it hides the one thing you could not have participated in.
What that tells you is which market Heikin Ashi is better suited to. On a US index future, which trades almost continuously, Heikin Ashi removes very little that matters. On an Indian single stock, especially one outside the derivatives segment with a tight price band, it removes 2 of the most important features of the instrument: the overnight gap and the band lock.
The rule that follows is simple. The more continuously an instrument trades, the safer it is to smooth it. The more an instrument jumps, the more the jumps were the information, and the less you should be looking at an average of them.
Carry this
- HA Open and HA Close are averages. Nobody traded at them.
- Heikin Ashi cannot draw a gap. Ever.
- It turns 1 to 3 candles after the real chart does, and the lag is worst in fast markets.
- Read the trend from Heikin Ashi. Read the price from candlesticks.