Heikin-Ashi and Bar Candles' Highs and Lows Don't Sync. Why?

Hello Traders,

As I (hopefully) understand it, Heikin-Ashi tail and nose tips are defined as showing true high and low price levels.

To quote Dale Woods,
"High Price = highest price out of the current candle’s high, open, or close price
Low Price = lowest price out of the current candles’s low, open, or close price

The high and low prices are nothing special. The Heikin Ashi candle will just show the highest and lowest data point achieved while it was active" Your Ultimate Guide to Trading with Heikin Ashi Candles.

Same with bar candles: candle tips show high and low; yet - and here’s the issue - when I superimpose bar candles over Heikin-Ashi, there are often vast differences in the highs and lows as shown by the two types of candles.

I’m baffled. If you can help me understand, I’d truly appreciate it.

Thanks,
Norm

Norm, I’m not an expert!

It’s my understanding the actual heiken ashi candle, tip to tail, is based on some type of price range averaging.

Hi Norm,

You are correct in your observations. Here is an image depicting exactly what you are describing —


click the image to enlarge it



I have never seen an understandable explanation of how the Heiken-Ashi high and low are calculated.

The formulas given in Investopedia are worded in typical fashion, and I think the wording is nonsense. Here are the Heiken-Ashi formulas, according to Investopedia

Heiken-Ashi formulas

Without a lot more explanation, the formulas for High and Low are just silly. It you have three data points to choose from – high, open, and close – obviously, you will choose the “high” to be your “high”. Duh.

Likewise, with the formula for the Low.

The problem is that Heiken-Ashi candles incorporate metrics from the current (open) time period, and from the previous (closed) time period. The internet sources which purport to explain how the H-A candles are constructed do a decent job of nailing down the H-A Open and Close, but fail to adequately explain how the H-A High and Low are calculated.

I vaguely recall trying to sort this out in a previous post, a long time ago (in a galaxy far, far away), but I don’t recall how I resolved it. (When you get old, memory is the second thing that goes.)

I will try to find my previous post. If it contained any useful insight, I’ll pass it along to you.

In the meantime, you guys in Hawaii are the last folks (in the civilized world) to ring in the New Year.
Better late than never :rofl:

I hope you will continue to be healthy, wealthy and wise in the New Year.

1 Like

Hey Clint,

Great hearing from you! Believe it or not, I just discovered your post. I don’t recall receiving notification of it.

In any case, I’ve said it before and I’ll say it again: I truly appreciate the input you’ve put into my forex understanding over the years. On that note, thanks for affirming my observations with your Japanese & H-A candle comparison. No, I’m not nuts - at least not on this count!

If you remember - being the old guy that you are - to seek out and find your ancient resolution to the problem (akin to squaring a circle), as usual, I’d be grateful if you sent it my way. I think that getting a handle on it would truly facilitate accurate placement of entries and stops. However, don’t lose sleep over it, not that I thought you would, anyway :grinning:

As far as my progress with forex goes, it’s been a long haul learning & testing interspersed with my other responsibilities, such as, work, which really cuts into a guy’s time! I haven’t laid a dime on the line yet, but I think I’ve found a winner with some H-A approaches. Some testing when I’m able and, Lord willing - watch out Mr. Soros! In the mean time, perhaps it’s best if I place entries and stops with bar charts or traditional candles and then manage with H-A.

P.S. I’ve recently taken up recreational canoe paddling with 12-person double hulls. I feel like King Kong on steroids!

Hope all is well with you, and . . .

Blessings in our Savior,
Norm

Hello Norm,

I’m working on a detailed comparison of standard candlestick charts versus Heiken-Ashi charts, but it’s not yet ready for posting. (I gave up trying to find something I wrote on the subject long ago.)

In the meantime, here’s a quick reply to your last post.

In my view, H-A charts should not be used for determining entry or exit levels (prices). All of the prices in a H-A chart are artificial – as you know, if you have ever tried to make sense of the H-A formulas as they are typically presented. To me, it doesn’t make sense to use these artificial prices for something as critical as choosing where to enter or exit.

The ForexGuy article (which you linked to) shows a method for using a series of H-A OPEN “prices” for a sequential trailing stop (that is, a trailing stop that you adjust manually after the close of each H-A candle). I’m very suspicious of this method, because of the artificial nature of the prices being used. But, I haven’t tested the method, so I will say no more about it, at this time.

In my opinion, H-A charts are a marvelous tool for price pattern analysis. The biases built into the construction of H-A charts make trends, trend-reversals, and other types of market-structure jump right out at you, and grab your attention. This can be extremely useful in pointing you to potential trade set-ups, and for confirming analyses you have made on “real” price charts.

I tend to think of a H-A chart as a sentiment indicator – but, the word sentiment is commonly used in other contexts, and probably should not be introduced here.

So, to return to the part of your quote (above) which I rendered in bold type, I think you should definitely place entries and exits on standard price charts (candlestick charts, bar charts, or whatever blows your skirt up) – not only “in the meantime”, but all the time.

I’ll post some analysis of the mathematics of H-A chart construction, when I’m able to get to it.

In the meantime, the big end of the paddle goes in the water. Don’t drown yourself :rofl:

Hi again, Norm

I want to discuss that image I posted in my first reply to this thread, back on December 31.

Ever since I posted that image, something about it has bothered me. As I did some research on the Heiken-Ashi formulas, I discovered what’s wrong with that image.

The image purports to show how HA charts smooth out some of the “noise” in Japanese candlestick charts. So far, so good. Chart “noise” is a fact of life, and HA charts do smooth out some of it.

Bur, here’s where the problem arises. To illustrate that smoothing, the image specifically highlights major differences in the highs and lows in the HA chart versus the Japanese candlestick chart. At first glance, what the image is showing looks perfectly reasonable, and appears to be a good illustration of the differences you observed in your original post. But, the image is fake.

While there are differences in the highs and lows displayed on a HA chart, versus the highs and lows on the corresponding candlestick chart, the “differences” shown in that image from forextrainggroup .com are IMPOSSIBLE. The image has been doctored.

It’s clear from the formulas for Heiken-Ashi highs and lows that:

  • HA highs cannot be lower than the actual highs. Yet, the red and green ovals in the forextrainggroup image purport to show exactly that.

  • And, HA lows cannot be higher than the actual lows. Yet, the blue oval and the gold rectangle purport to show that impossible situation.

As soon as the HA formulas proved to me that the image was doctored, I regretted having posted that image. If I could edit my post, I would delete that image, and rewrite the post. But, the window of opportunity for editing that particular post has closed.


None of this is meant to imply that HA highs and lows are the same as actual highs and lows. In many cases, they are not. Rather, HA highs can actually be higher than actual highs, and HA lows can actually be lower than actual lows. And, in many cases, this is exactly what we see – as you have noted.


I’m still working on a post detailing the construction of Heiken-Ashi candles, along with some comments on what traders can and can’t do with Heiken-Ashi charts. Since all of that sort of veers away from the original topic of your thread, I might post that analysis in a new thread. The way it’s developing, it will be a rather lengthy analysis. Maybe I’ll post it in several parts, in a series of posts.

Hi frandlost,

Apparently that’s the case, even for the H-A highs and lows.

Summary: None of the H-A levels reflect true price levels - highs, lows, opens, closes. They’re all doctored to provide a smoothing effect on the trend.

Application: H-A is great to see uncluttered trend continuations and reversals, but it would be disastrous to set any entry or exit level on those candles.

I have not found a way to get H-A and traditional candles on the same chart so they can be seen distinctly one from the other due to candle overlap, but this is not the case with H-A Smoothed. One can distinguish one from the other very clearly, especially if different colors are applied. Also, per my observation, H-A Smoothed follows the traditional candle trend beautifully, but also clears out the “noise” beautifully, better than the standard H-A! I’ll therefore use H-A Smoothed like a pair of eyeglasses that clears up the fuzz; but I’ll set all my entries and exits on traditional Japanese candles, which show true prices levels for all opens, closes, highs and lows.

Thanks and take care,
Norm

Hi Clint,

I’d be stoked to read your H-A posts. If you do place your series in a different thread, please, at least place a reference in this thread to where your new thread is located, or notify me in my personal message folder.

BTW, as I responded to frandlost above,

Also, it’s amazing to me that in all the H-A studies that I’ve read where they show where to place entries and stops, they all show their entry and stop levels on the H-A candles! If they intended that entries and stops should be placed on the corresponding traditional Japanese candles or bar candles, they certainly didn’t say so, leaving the poor reader to figure it out for themselves, no doubt leading many astray in their trials and trades!

Thanks and take care,
Norm

Hi Norm,

My Heiken-Ashi study is finally launched, although not complete. Here’s a LINK to it, as you requested.

Thanks Clint,

I already gave your first post the once-over. I’ll be looking at it more closely. Looking forward to seeing what else you come up with. Your series may be a real eye-opener for many.

As far as application goes, thus far, Heiken-Ashi Smoothed - really a different animal - has fallen into place as being a trend-clarifier and one of several possibilities in my arsenal of places behind which to place trailing stops.

Take care,
Norm