Introduction

The candlestick trade I’m about to describe made me enough money to pay for three months of living expenses. That’s not bad for an hour’s work! In a moment, I’ll outline the trading setup so you can tailor it to your liking.

On May 16, 2007, I went shopping for a stock to buy and found one in the diversified chemicals aisle of the market. What caught my eye first was a consolidation pattern called a descending triangle. Figure I.1 shows the chart pattern in May. Descending triangles have a flat bottom and a downward-sloping top. They break out downward 64% of the time, but upward breakouts can post spectacular results. I was looking for an upward breakout.

Figure I.1 The combination of a descending triangle, Big W, and morning doji star set up a profitable trade.

The next pattern I noticed was the Big W. A tall left side (C to A) leads to a reversal pattern, such as the Eve & Eve double bottom AB. Eve bottoms are wide, rounded turns, unlike Adam bottoms, which are narrow, often pointed—a single spike or two wide. Eve can have spikes like that shown at B, but the spikes are shorter and more numerous than what you see on Adam. The Eve & Eve double bottom is one of the more powerful and successful chart patterns. The theory behind a Big W is that the right side will mimic the left and price will climb after that.

The pattern that sealed the deal was the morning doji star candlestick. That candle pattern ended the day that price closed above the top of the descending triangle. My research said that the morning doji star is a highly reliable candle formation. Combined with additional analysis I did on the company, both fundamental and technical, the stock was a buy only if it gapped open higher. Why? Because the next day the company was holding a conference call before the market opened to discuss earnings. A higher open would mean the market liked its story.

The news reports the night before the meeting said that net profit was 53 cents versus 61 cents during the year-ago quarter even as revenue climbed by 30%. Just 1% of the revenue gain was from higher internal sales, though. Most was from acquisitions or currency translation. Analyst estimates ranged from $0.52 to $0.58, so earnings came in at the low end. All of this sounded bearish to me, but the technicals were shouting, “Buy!”

My candle research says that opening gap confirmation from a morning doji star results in the best performance. That means trading with the trend as soon as possible.

The next day, I watched the stock open and price took off. In the first minute, it shot from the prior close of 33.46 to 34. That left a tall white candle on the chart. Again, my research and knowledge of candles said that the body of a tall candle is often a support zone. So, I placed a buy order halfway down the body, at 33.78.

Sure enough, price turned down and nailed my buy order, filling most of it before moving up again. I had trouble fitting through the door because the smile on my face was so wide. Five minutes later, the remainder of the order filled.

By day’s end, the stock had recovered and closed higher by 1.16. The next day a brokerage firm upgraded the stock and price moved higher still, this time up another 1.93. The following day price coasted upward 33 cents (D).

Exit time. Why? This is one of those situations where you get a feeling that it’s time to leave, so I started my analysis. The height of the candle lines was diminishing, suggesting a trend change. The Commodity Channel Index (CCI, with default settings of 20 bars for the lookback and 5 for the DCCI line—dual CCI, a smoothing of the CCI), an indicator, was rounding over and looked as if the next day would produce a sell signal. In other words, upward price momentum was slowing. I didn’t want to hang around and give back my profits.

The falling window from C to E I thought added to overhead resistance, but my research said that happens only 25% of the time. The candle pattern at D also resembled a shooting star, with a tall upper shadow and small body after an upward price trend. This one wasn’t perfect, because the body was too tall in relation to the height of the upper shadow. If you have no idea what an upper shadow and a body are, don’t worry about it. I’ll explain them later.

The day after D, I vowed that if the stock opened lower, I would sell. I thought of selling it all at the open but the futures market suggested a higher open for equities, so I decided to wait and see.

Volume was thin in the stock but it opened lower, just as I expected. I timed the exit as best as I could and got out just before price plummeted. By the day’s close, however, the stock had gained it all back and then some. The candle that printed on the chart was a hanging man. Despite their reputation as a reversal, they act as a continuation pattern 59% of the time. That suggested more upside. But it doesn’t matter because I don’t own the stock anymore.

If you want to replicate this trading setup, look for:

• Good industry relative strength. If the stocks in the industry are doing well, then the chances improve that this stock will do well, too.
• Better than expected earnings. Only price can tell you how much the market likes the results, so watch the stock after the announcement of earnings. If price gaps upward, buy immediately or wait for a retracement and then buy.
• A reason to buy the stock from a technical or fundamental perspective. Even if price does not explode higher at the open, it should do well in the coming weeks based on your analysis.

I used an upward breakout from a descending triangle with a morning doji star inside a Big W pattern. The combination worked well but will be almost impossible to duplicate.

What Are Candlesticks?

Let’s talk about candlesticks, starting at the beginning so everyone comes up to speed at the same time. I’ll be brief because most people know what candle charts are. Figure I.2 shows two examples of candlesticks. The line (a single price bar) on the left is a white candle. This one shows the relative positions of the open, high, low, and close. Notice that the closing price is higher than the opening price. When that occurs, the body is white. On the right, the candle is black because price closed below the open. The upper shadow is hair growing from the top of the candle, and the lower shadow is a single leg dangling from the bottom of the candle. It may help to think of shadows as wicks.

Figure I.2 White and black candlestick lines.

Candles don’t need either an upper or a lower shadow. They don’t need a body, either (such as when the open and close are the same). The key concept to remember is that a black candle shows a close below the open and a white candle shows a close above the open. A black candle does not show price closing lower than the previous day, nor does a white candle show a higher close than the day before.

With this candle definition, you can have a stream of white candles in a declining price trend, and black candles forming a rising price trend. I’ve seen both situations, too.

That’s all there is to candle configuration. Multiple candle lines along with variations in shadow and body length make up the many candle patterns.

The Data

I wrote a computer program to recognize all of the candlestick patterns in this book. With nearly five million candle lines to explore for each of over 100 candle patterns, doing it manually was not an option.

I created what I call the standard database, a collection of data that doesn’t change in size. It contains the 500 stocks of the S&P 500 index for 10 years covering both bull and bear markets. From the standard database I derive the frequency rank and prorate the number of times a particular candlestick pattern is found to determine whether it appears more often in bull or bear markets. Details of the database are in the “Glossary and Methodology” section at the back of the book.

If I find too few candles in the standard database, then I use up to three more databases, one containing archived stocks I no longer follow, one that is a five-year/500-stock database, and one that is the current database I use for trading. Together they comprise almost five million candle lines (price bars). I removed any duplicated candles between databases, and all four contain split-adjusted, clean data.

The Price Trend

Many candles have a defined price trend that leads to the start of the candle pattern. For example, a hanging man appears in an upward price trend, and a hammer appears in a downtrend.

How do you determine the trend? I use a 10-day exponential moving average as a starting point and season it with special rules to allow price trends of a day or two to override the result. The method is a bit complicated but it works well. It is, however, not perfect, but the large sample size I use helps compensate.

Candle Performance

How do you measure candle performance? Since candlestick patterns often lead to short-term moves, I used the closest minor high or low (swing high or swing low), depending on the breakout direction, to gauge performance. The straight-line move often isn’t a long one, but it serves as a good proxy of what you can expect. The statistics in this book should be used to compare results from candle to candle, not as benchmarks of how well you will do trading...