Chapter 4

Candlesticks --- One Candle, One Story

The first time Rohan watched a candlestick form in real time, he was on the trading floor of nowhere in particular --- just his desk in Andheri, laptop open, Kite loaded, Nifty 50 fifteen-minute chart on one half of his screen and his work code on the other.

It was 9:17am on a Tuesday in November. The first candle of the day was forming. He watched the price tick up, tick down, tick up again. The candle body was growing, red for a moment, then green, then red again. At 9:30am the candle closed. The body was small. Both the upper and lower shadows were long. The price had moved significantly in both directions during those fifteen minutes, but the close was almost exactly where the open had been.

He did not know what this meant. He knew it was called a doji. He had read the word in three different places. He did not know what to do with it.

That week, on the recommendation of a colleague who traded occasionally, he bought a book. It was a well-known book on Japanese candlestick charting. It was 330 pages long and contained forty-seven distinct candlestick patterns, each with a name, a diagram, and a description of what it predicted.

He read it over four evenings. He highlighted sections. He made notes in the margin. He created a reference table in his notebook with the pattern names and their meanings.

By the end of the week he was more confused than before.

The problem was not that the patterns were complicated. The problem was that there were forty-seven of them, each with conditions and exceptions and caveats about market context and confirmation requirements. Morning Star, Evening Star, Abandoned Baby, Deliberation, Advance Block, Stalled Pattern, Three White Soldiers, Three Black Crows. He had learned, for example, that a Bullish Engulfing was a bullish reversal signal, but only when it appeared at the bottom of a downtrend, and only when the second candle completely engulfed the first, and only when confirmed by the next session. He had also learned that a Dark Cloud Cover was a bearish reversal signal with different but similarly contingent conditions.

He could identify patterns in past charts with some confidence. In real time, with money on the line, he could identify nothing. The candle was still forming. It might be an Engulfing. It might become a Doji. It might close and become neither.

He threw the book at his couch. Not hard. But he threw it.

He asked Meera if she had read any candlestick books.

"I tried," she said. "I got to twenty-three patterns and gave up. I use three."

"Only three?"

"What's the difference between a small body and a doji?" she asked. "I've never been able to see it in real time."

He did not know the answer precisely. He wrote the question in his notebook.

He asked KM Sir on Sunday.

KM Sir looked at the question in Rohan's notebook. He picked up his own notebook, read yesterday's line, and set it down.

*A candle tells you who won a battle. It does not tell you who will win the war.*

"How many patterns are in that book?" KM Sir asked.

"Forty-seven."

"How many do you need?"

Rohan thought about it. "Three?" he guessed, thinking of what Meera had said.

"Three is correct. And the difference between a small body and a doji is this: in a doji, the open and close are at or very near the same price. The battle ended in a draw. In a small body, one side won --- but only by a small margin. The distinction matters less in isolation and more in context."

The three patterns KM Sir considered worth learning for a beginner were these.

The first was the Doji. A candle where the open and close were at essentially the same price, producing a cross or plus-sign shape. What it communicated was this: buyers and sellers fought to a draw during this period. Neither side won. When a doji appears after a strong move --- up or down --- it suggests the momentum of that move may be slowing. Not reversing. Slowing. The doji alone proves nothing. What comes after the doji is what matters.

The second was the Engulfing candle. Two candles together: the second candle's body completely contains the first candle's body. A Bullish Engulfing at the bottom of a downtrend --- a small red candle followed by a larger green candle that covers it entirely --- communicates that sellers dominated the first period, but buyers came in with enough force in the second period to overwhelm them entirely. The reversal in force is visible in the candles.

The third was the Hammer. A candle with a very long lower shadow and a small body at the top. The price dropped significantly during the period, then recovered almost completely before the close. What this communicates: sellers tried to push the price down. Buyers came in forcefully at the lower price and pushed it back up. The long shadow is the evidence of the sellers' attempt. The close near the open is the evidence of the buyers' response.

"These three tell you enough to start reading price in context," KM Sir said. "The other forty-four exist. Some of them are useful. But you will not be able to use them effectively until you understand these three so thoroughly that you no longer need to think about them. That takes time. Forty-seven patterns at once is a way to know many things poorly. Three patterns over time is a way to know three things well."

The question Meera had asked --- the difference between a small body and a doji --- turned out to have a practical answer that was more important than the technical one.

In real time, the distinction often did not matter as much as the context. A doji after fifteen consecutive bullish candles communicated differently than a doji in the middle of a sideways range. A small body after a large engulfing candle communicated differently than a small body in isolation. The specific shape of the candle was one piece of information. The price action around it was the rest.

What the book with forty-seven patterns had taught Rohan, without meaning to, was to look at each candle in isolation. What KM Sir was teaching him was to look at each candle as part of a conversation --- one statement in a longer exchange between buyers and sellers.

A candle that said "buyers won today" meant something different if the previous ten candles had shown sellers winning steadily. It might mean the first sign of a reversal. It might mean a one-day pause in a downtrend. The candle alone could not tell you which.

He closed the browser tab with the forty-seven patterns and did not open it again.

He kept the tab with the three patterns. He printed the page and pinned it above his monitor at home, where it stayed until he moved out of the Malad flat three years later, by which point he no longer needed it.

He spent the following week looking at daily charts with only the three patterns as his filter. Every morning he would look at the Nifty 50 daily chart and find any of the three patterns that had appeared in the last ten candles. He would note what the price had done afterward.

The pattern that appeared most often was the Doji. The ones that appeared after strong directional moves --- after the price had moved cleanly in one direction for several days --- were often followed by a slowdown or a reversal. The ones that appeared in the middle of choppy, sideways price action were followed by nothing in particular.

Context, again. Always context.

He wrote in his notebook: A candle is a sentence. The chart is the paragraph. Read the paragraph, not just the sentence.

He did not know if KM Sir would agree with this. He asked at the next Sunday meeting.

KM Sir read it. He nodded once. He said nothing else.

Rohan would learn, over the coming months, that this was KM Sir's highest form of approval.

The practical problem with candlestick patterns was not learning them. It was learning to ignore the wrong ones.

Rohan had, in the third week of December, watched three doji candles form on three different stocks in a single afternoon. He had identified all three correctly. He had not known what to do with any of them.

A doji candle communicating indecision was only useful information if indecision at that specific moment, in that specific context, was significant. A doji after a strong upward trend of twelve sessions communicated one thing: the buyers who had been dominant for twelve sessions were, at this moment, uncertain. This was meaningful. A doji in the middle of a two-week sideways range communicated nothing that the range had not already communicated. It was indecision within indecision.

He brought this problem to KM Sir, who answered it with a question rather than an answer.

"Before you look at the candle," KM Sir said, "what should you already know?"

Rohan thought about the framework he had been building. "The trend. The direction the market has been moving."

"Yes. And?"

He thought further. "Whether price is at a significant level."

"Yes. So if you know the trend and you know the level, what does the candlestick pattern add?"

He understood. "Confirmation. The candlestick pattern tells you what happened at the level, within the context of the trend."

"Exactly. A hammer at a support level in an uptrend is three things agreeing simultaneously: the trend is up, the level has historical significance, and buyers appeared forcefully at that level. That is a meaningful combination. A hammer in the middle of a range without a significant level is a hammer in a vacuum. It adds nothing."

The three-pattern framework gave Rohan a specific way to look at charts that he had not had before.

For each chart, at each significant level, he asked three sequential questions. First: what was the trend when price arrived at this level? Second: had the price behaved specifically at this level before? Third: what did the candle at the level look like?

If he could not answer the first two questions clearly, the third question was irrelevant. The candle pattern was the last piece of evidence, not the first.

He spent a week going through Nifty 50 daily charts from the previous three months and marking every instance where a doji, engulfing, or hammer had appeared at or near a significant support or resistance level. He found eleven instances. He noted what the trend had been at each, whether the level was significant, and what price had done in the following five sessions.

Seven of the eleven had resulted in a move in the direction suggested by the pattern. Four had not.

He brought this to KM Sir.

"Seven out of eleven," KM Sir said. "Sixty-four percent." He looked at the four that had not worked. "What did these four have in common?"

Rohan looked at them. Two had appeared at levels that had been broken and then retested. The pattern had appeared at the right level but the level had not held. One had been at a level that had genuinely held, but the trend had been weak rather than clear. One had been a borderline case --- the pattern was close to a doji but the close was slightly below the open, making it technically bearish.

"The level did not hold in two cases. The trend was unclear in one. The pattern was ambiguous in one," Rohan said.

"Yes. So the pattern did not fail. The conditions for the pattern were not fully met." This distinction --- between a pattern failing and the conditions for the pattern not being present --- was one that he would carry forward.

The reader exercise for this chapter was observational rather than active.

For two weeks, Rohan opened the Nifty 50 and Bank Nifty daily charts every evening and noted any of the three patterns that had appeared in the last three sessions. He did not trade on any of them. He just noted them, and then the following evening, he noted what price had done afterward.

By the end of two weeks he had logged fourteen observations. Ten had moved in the direction consistent with the pattern. Four had not.

He looked at the four that had not worked and asked the same question KM Sir had asked: what had been different? Three of the four had appeared without a significant level nearby. One had appeared with the trend against the pattern direction.

The pattern that worked without all the conditions was lucky. The pattern that worked with all the conditions was systematic.

He closed the browser tab with the forty-seven patterns for the last time. He kept the page with the three patterns and the questions he had learned to ask before looking at them. The page stayed in his notebook until the binding wore through and he transferred it to the next notebook, then the next.