Nearly every week, someone sends TradingSchools.Org a request, asking about various topics related to technical analysis tools used to analyze stocks, futures, Forex or options. Nearly all modern trading platforms contain a plethora of basic trading tools. Some of the more common technical analysis tools are moving averages, stochastics, volume analysis, Japanese candlesticks, Heiken Ashi, MACD, divergence, Bollinger Bands, to name a few.
It seems like every few years, we get an entirely new set of tools that capture the imagination of the investing public. Crafty vendors and trading educators quickly pick up on these consumer trends and create their own “secret sauce” trading indicator that supposedly predicts future price direction. I like to call these sorts of characters “Indicator Hustlers”, as they put a little shine and spin on increasingly complex mathematical formulas. Slap a clever-sounding name on it and sell it to the next round of suckers. For instance, during the 1980’s, at the dawn of the personal computer, a few clever marketers created very rough looking programs that charted moving averages of stocks and commodities prices, and sold the idea as a “method of prediction”. A new industry was born.
As personal computers became cheaper and cheaper, the natural evolution was for trading platforms to offer more advanced feature sets. At some point, we dived off the deep end…into the esoteric. Some of those early esoteric hipsters include Wells Wilder, who claimed that future price direction could be universally predicted by charting the stars in the sky, or the gravitational pull of the moon. Yet others claimed to be able to predict future price direction through bible passages. Further down the rabbit hole we continually venture.
The truth is that even the most esoteric hipster, like a broken clock, will be correct at least twice-a-day. And being correct only twice a day is enough to sell the masses of desperate consumers the ‘next great thing’ in trading technology. Which brings me to Heiken Ashi Candlesticks.
What in the heck is a Heiken Ashi Candlestick? Quite simply, it is just a Japanese Candlestick with a tiny little trend component built into the equation. I will not be writing about Japanese Candlestick charting techniques. Have already spent a good deal of time writing about various hustlers in this particular niche. Hustler A, and Hustler B. Candlesticks are just another way to apply meaning to price charts.
Heiken Ashi Candlesticks are the next level of evolution of the original Japanese Candlestick. When you look at the Japenese Candlestick, you can see trend a little more clearly than a standard bar chart. A Heiken Ashi adds another subtle layer or calculation into the mix. Let’s take a look at the calculation for a Heiken Ashi bar, there are four calculations to form a bar.
Open of a Heiken Ashi Bar is calculated by adding the Open of the previous bar, plus the close of the previous bar, divided by 2. The exact formula is:
- Heiken Ashi Open = (Open of previous bar + Close of previous bar) / 2
Close of a Heiken Ashi Bar is calculated by adding the Open, plus the High, plus the Low, plus the Close, divided by 4. The exact formula is:
- Heiken Ashi Close = (Open + High + Low + Close) / 4
High of a Heiken Ashi Bar is calculated as the highest value of the High, Open, or Close. This might sound confusing. How could the Open or Close be higher than the High? Remember, the Heiken Ashi uses an average of the previous bar. So there are instances where the Open or Close can be higher than the High of the current bar. If still confused, these bars are easily seen as solid bars with no ‘tail’ or ‘candle wick’.
Low of a Heiken Ashi Bar is calculated as the lowest value of the Low, Open, or Close. Once again, if confused, these bars will be easily seen as solid bars with no tail.
Next, let us take a look at a Heiken Ashi Candlestick vs a standard Japanese Candlestick. The following screenshot is a 60-minute chart of the Emini SP500, with standard Japanese Candlestick on the top, and a Heiken Ashi Candlestick on the bottom.
As you can see, in this particular instance the Heiken Ashi did a better job of signaling a continuing uptrend for 60-minute Emini SP500 market. So why am I including this screenshot of a Heiken Ashi vs a regular Japanese Candlestick?
Day Trading Room Example
As many readers are aware, TradingSchools.Org records a massive amount of trading sessions within live day trading rooms being offered by various trading vendors. Recently, we spent a week inside of a popular trading room where the head moderator was using a 60 minute Heiken Ashi as his ‘anchor chart’ for day trading the Emini SP500 futures contract. The concept of an anchor chart is pretty simple. You would only trade in the direction of the dominant trend as per the anchor chart. If the ‘anchor chart’ is green and signalling higher prices, then you would only take trades in the direction of (hopefully) higher prices.
In this particular example, the trading vendor did OK by avoiding the possible trend change heading downwards, that was flashing on the RED as per the standard Japanese candlestick chart. Many traders might have been duped into taking a short trade, only to quickly discover that the signal was false. The market continued to grind higher.
After the market closed for the day. The trading vendor gave a ‘post-market’ wrap up and continued to talk about the amazing wonders of using a Heiken Ashi as a trend filter. In fact, he stated that for the prior 5 years of “being a full-time professional day trader” that specialized in the Emini SP 500, the Heiken Ashi 60 minute candlestick contains a robust and definitive trading edge. That it was as reliable and predictable as the sun setting in the west and rising in the east.” This guy was a true believer.
Next, I called him on the telephone and had a very pleasant conversation. We talked about Heiken Ashi and the use of the 60-minute bar as an anchor bar to determine the primary trend. I asked him if he had ever programmed the Heiken Ashi 60-minute bar and tested it on the Emini SP500? He explained that through many years of trial and error as well as visual observation, he simply ‘knew’ it was a valid concept. The thousands of hours of trading screen time gave him an intrinsic feel of the market that expressed itself through the Heiken Ashi. There is nothing like personal experience, right?
And so next, I offered to program his Heiken Ashi strategy in exact detail. And then present the results in a scientific manner. He thought this was a great idea. He was eager to prove how wonderful his method works for his paying subscribers.
Heiken Ashi Trading Strategy Number One
Rule A: If the 60 minute, Emini SP500 is flashing a Heiken Ashi green bar. Which is signaling an uptrend, then buy the close of the green bar. Simply exit 60 minutes later.
Rule B: If the 60 minute, Emini SP500 is flashing a Heiken Ashi red bar. Which is signaling a downtrend, then sell short the close of the red bar. Simply exit 60 minutes later.
The system trades only the day session. No trades taken overnight. All trades exited at the close of the day session.
We deducted $4 per trade in commission costs. We just wanted to test the raw signal. If an edge was present, it would have emerged. We got the following results…and they were not good.
Profit Factor of 1.03 with an average trade of $6 profit with a $37k drawdown. Very ugly equity curve.
The vendor was pretty shocked to discover that his supposed ‘edge’ of using a 60-minute Heiken Ashi bar chart contained no statistical advantage. When I have these sorts of conversations with trading vendors, about their trading indicators, it is usually filled with long silences. And there is a lot of initial disbelief.
The truth of the matter is that many well-intentioned trading Guru’s really, truly believe that their kool-aid contains magical properties. When presented (confronted) with hard evidence, most are pretty sad and depressed. The usual excuse from the Guru is that they have a ‘feel’ for which signals to take, and which signals to not take. And so they cherry pick based upon hunch or some other form of anecdotal evidence that cannot be defined. This would be akin to a medical doctor deciding which patient should be vaccinated and which patient should not be vaccinated.
Trading system rule sets are a lot like vaccinations. Either they work on the entire population, or they do not. The larger the population in which to test the theory, the better in which to evaluate the results.
Markets are just groups of people taking action. And within these groups of people are incredibly complex ecosystems of motivation. No trading indicator has the ability to encapsulate the motivations of all participants. A trading indicator does not predict anything, it simply displays what has already happened. The beauty of today’s modern backtesting platforms is that we now have the ability to define micro events being revealed by trading indicators. Once we have these micro events isolated, we can then ask the tough question…what typically happens after a micro event, or what happens after a 60 minute Heiken Ashi bar is signaling higher prices.
As we can see from the prior test, it simply does not work. But should we just toss the results and forget about Heiken Ashi charts? Does this mean that Heiken Ashi is not a valid concept? Many readers are probably quick to stop reading and hastily decide that Heiken Ashi charts and Heiken Ashi trading systems are useless. But this is not the truth.
The truth is that every single trading indicator in existence is guaranteed to work sometime. The hard part of this game is figuring out when to use it. And when to put it on the shelf. The next section, I am going to do my best to walk the audience through a research scenario and provide a look into how I would incorporate Heiken Ashi into an actual trading strategy.
Developing a Heiken Ashi Trading Strategy
For me, when I develop a trading strategy, the most important thing is the sample size. I want a massive amount of data in which to work within and analyze. The larger the amount of data, the greater the chance that I might accidentally stumble onto a genuine and exploitable bias. A bias that other market participants are not seeing. Once I stumble onto a bias, I then analyze the bias and attempt to form a logical hypothesis at to why the market is repeating the same behavior.
Let’s first look at the data that I personally use. First things first, since I mostly trade during the daytime, I want to focus on just the day session data. And since I like to sleep at night, without worry or being anxious about the evening session, I like to omit evening session data. Yes, there is plenty of overnight/evening session financial data, but since I like to sleep and watch TV like every other regular Joe, I don’t want to be bothered with night time anxiety. So the following are the trading symbols that I typically test different trading ideas.
- Emini SP500
- Emini Russell 2000
- Emini Dow
- Emini Nasdaq
US Treasury Bond and Note Futures
- Ultra Bond
- 30-Year Treasury Bond
- 10-Year Treasury Note
- 5-Year Treasury Note
Softs and Meats Futures
- Feeder Cattle
- Soybean Oil
- Soybean Meal
Precious Metals Futures
- Australian Dollar
- British Pound
- Canadian Dollar
- Euro FX
- Japanese Yen
In addition to futures contracts that have enough volume to actually trade, I also include the following stocks and ETF’s:
- 75 various ETF’s or exchange traded funds
- Dow 30 stocks
- Nasdaq 100 stocks
- SP500 stocks
- Russell 3000 stocks
As you can see, its nearly 4000 symbols! In addition, I test on 5-minute bars, 15-minute bars, 30-minute bars, 60-minute bars, daily bars, and weekly bars. So when I run a test for a particular pattern, you can imagine the hundreds of thousands of potential combinations in which I am testing. It is quite a time suck.
Most readers contemplating this sort of heavy lift will assume that a super fast, super high powered PC is the best option. Wrong! On the contrary, the best option for this sort of grunt work are cheap $200 laptop computers. For me, I install my backtesting software and data on 5 cheap little laptops. When I want to run a long test, using thousands of symbols on a multitude of time frames, I will spread out the workload. Since 99.9% of the patterns and ideas that I test will most likely prove worthless, then why should I expend so many resources on expensive computers? I work the poor little laptop until the bearings in the hard drive eventually give out. And they all eventually give out.
Since 99.9% of the patterns and ideas that I test will most likely prove worthless, then why should I expend so many resources on expensive computers? I work the poor little laptop until the bearings in the hard drive eventually give out. And they all eventually give out.
The point that I am trying to make is that when you are developing a trading strategy, cast a wide net. The wider the better. Capture as much information into the net as possible. After you begin programming and testing different trading strategies, you will find that the most precious catch will be the fish that you did not expect to find.
Heiken Ashi: a workable trading bias. A possible trading strategy.
To show the power of casting a wide net, in hopes that something tasty turns up inside of the net. I took the original idea of the 60-minute Heiken Ashi that the trading room moderator was touting as an amazing anchor chart with ES500 futures contract, and simply tested the idea on my vast ocean of data. Hoping that something would wash into the net. When you cast as wide as net as me, the chance that something workable will magically appear is quite good. What I want to see is an exploitable bias. A clear direction of where prices are headed. I want to know that the dice are loaded. So let’s take a look at a pattern that you might find as valuable.
Rule 1. Copper Futures. Using a 60-minute Heiken Ashi chart. Imagine that there are only two hours remaining in the day session. The Heiken Ashi chart is pointing down, the candle is Red. Market participants are anticipating a down market, are assuming the market will close lower still. However, as the market moves toward the close, we begin to see measurable buying coming into the copper futures market. The price begins to rally. With only one hour remaining in the trading day, the final Heiken Ashi bar flips from Red to Green. A signal that the final 60 minutes of the trading session is going to hopefully witness additional buying pressure and higher prices.
Rule 2. With only 60 minutes remaining in the day trading session, the Heiken Ashi candle is now Green and pointing higher. We enter to buy Copper futures at the market, and hope that there is an angry ninja about to bust out of the Heiken Ashi candle. We want that copper market to move higher for the final hour of the day. We wager it will. We exit this trade at the end of the day, without exception. And our copper futures trade is based on the following evidence…
What is this equity curve telling us? This equity curve is telling us quite plainly that the final hour of the day, with Copper Futures, the market tends to drift higher. Our trigger is when the 60-minute Heiken Ashi candle flips from (Red) down to up (Green) up. Let me remind folks, there is nothing fantastical or magical about a 60-minute Heiken Ashi candlestick. The Heiken Ashi candlestick did not predict the future. It simply gave us a trading signal that Copper prices were likely heading higher. Truth be told, there are hundreds of indicators that do the exact same thing. But that is not important, what is important is that we have found a tendency for Copper prices to rally into the close of the day if the price of Copper had been drifting lower the previous 60 minutes.
Refining our Heiken Ashi Strategy
Is an average trade size of $69 good enough as a standalone trading strategy. For me, it is not. I like to see an average trade size, for Copper futures to be +$90. So we are not that far away. How can we improve this strategy to increase average trade size? The good news is that we have a sample size of 786 trades. So we can add a few qualifiers that decrease the number of trades, and increase the average profit per trade over $100 per trade.
Without giving away the store, and to inspire you to take this concept further, I will give you a few hints on how you can ‘supercharge’ this bias into a very robust trading strategy. The following a few hints:
- What happens to this bias when the price of copper (intraday) is highly correlated with the price of gold (intraday)?
- Are certain ETFs and stocks that are highly correlated with Copper prices a good candidate for this strategy?
- What happens to this strategy when GDP or Gross Domestic Product is expanding?
- What happens to this strategy when the COT of Commitment of Traders Report reveals that producers are accumulating copper?
Let your imagination lead you forward. The answers to these questions are in plain site.
Using Stops with this Heiken Ashi Strategy
I have a simple philosophy regarding trading stops. Use the widest stop loss as possible. Yes, you can turn a losing strategy into a winning strategy by employing different stop sizes. But if your strategy requires a tight stop or a small stop, then you are playing with a delicate trading strategy. And a delicate trading strategy, that requires a lot of input is a strategy doomed to failure. Avoid this problem by only employing strategies that work across a wide spectrum of stop amounts. Let’s take a look at our Heiken Ashi 60-minute Copper Futures strategy, using stop losses ranging from $100 to $1000. We will backtest using increments of $100.
As you can see, the size of the stop loss is pretty irrelevant. The bias remains apparent, irrespective of the amount of the stop loss. For me, I would choose a stop loss of $1000 per contract. I want to give the market enough time and room in order for the bias to reveal itself.
When to turn off the Heiken Ashi Strategy Off?
Every system developer is going to have their own ‘special sauce’. No two developers are going to give the same answer as to when a trading strategy should be retired. The truth is that there is NO CORRECT ANSWER. But, I will tell you my own procedure for turning a strategy off. I simply calculate a 100 periods simple moving average of the equity curve. Please notice on the above equity curve, which you will see as a jagged red line. And then notice that under the equity curve is a smooth blue line.
The smooth blue line would be 100 periods, simple moving average of the equity curve. Once the equity curve dips below the average, then it would be time to turn off this trading strategy.
Once again, there is much debate as to when to turn on or turn off a trading strategy. But the hard and ugly truth about trading strategies is that all trading strategies are eventually doomed to failure. All markets change over time. Nothing lasts forever. So you need to have something built into all of your trading strategies that put them into the graveyard. This might be disappointing for many to read, but its the hard truth.
The key to long-term systematic trading success
In my opinion, the key to surviving is to accept that the markets are always in a state of change. What is apparent today, will be gone tomorrow. So you have to be always be searching. Always researching. Never give up asking questions and backtesting different hypothesis. Always keep digging. The moment you stop researching is the moment you seal your fate. Failure will never stop chasing you.
Wow, this blog post is now over 3000 words. Sorry, I took so much of your time. Thanks for making it this far!
I will wrap this up by stating the obvious…whenever a trading vendor begins howling that we “just had a moving average crossover!” or “big time hedge funds are hitting the bid!”. Stay calm. And use the backtesting tools the are readily available. Test stuff! Once you figure out how to code your own strategies and test differing trading indicators, a whole new world is going to open up.
Thanks for reading. And if you need help testing your own ideas, you can always email me with questions. I can usually code any trading idea in less than a minute.