Candlestick Charts: What They Show, and What They Cannot Tell You

A candlestick chart compresses four numbers per period into a single shape. It is an efficient way to represent price information, and reading one properly is a genuinely useful skill.

The large industry of pattern names built on top of it rests on considerably weaker foundations, and it is worth separating the two.

Anatomy of a candle

Each candle covers one time period — a minute, an hour, a day — and encodes four prices:

• Open — the first traded price of the period
• High — the highest reached
• Low — the lowest reached
• Close — the final price

The body spans open to close. The wicks extend to the high and low.

Colour shows direction: conventionally green or white when the close is above the open, red or black when below.

What the shape communicates

The proportions carry information about the balance within the period.

Shape	Structure	Reasonable reading
Long body, short wicks	Sustained directional movement	One side controlled the period
Short body, long wicks	Wide range, little net change	Contested, no resolution
Long lower wick, close near high	Sold off, then recovered	Buying appeared at lower levels
Long upper wick, close near low	Rallied, then faded	Selling appeared at higher levels
Almost no body (doji)	Open and close nearly equal	Balance, indecision

Note the framing. These describe what happened during the period. They do not state what happens next, and the distinction matters more than anything else in this article.

The pattern industry

Dozens of named patterns exist — hammers, engulfing candles, morning stars, three white soldiers. The literature around them typically presents each with a stated implication and a reliability rating.

The evidence is considerably weaker than that presentation implies.

Academic studies testing candlestick patterns systematically find limited predictive power once transaction costs are included. Results are frequently not robust across different markets, different periods or different parameter choices. Broader surveys of technical analysis find that early positive results in some markets have generally weakened over time, consistent with any genuine inefficiency being competed away once it is widely known.

That does not mean chart reading is worthless. It means the specific claim — that a named three-candle formation reliably predicts direction — is not well supported.

Why patterns appear to work

Three mechanisms explain the persistent belief.

Selection in hindsight. Scrolling back through a chart, you find the pattern where it preceded a move. The identical pattern where nothing followed is not salient, and there are far more of those.

Vague definitions. Many patterns lack precise criteria. How long must the wick be relative to the body? How close must the close be to the high? Without exact rules, identification becomes flexible enough to fit whatever happened.

Multiple comparisons. With dozens of patterns across multiple timeframes, something will always be present. Some of those will precede moves purely by chance.

The honest test: define a pattern with precise numerical criteria, apply it mechanically to historical data across several markets, include realistic costs, and measure. Almost nobody promoting patterns has done this, and where it has been done the results have generally been unimpressive.

What is more defensible

Some chart-derived observations have better support, largely because they reflect real market structure rather than shape recognition.

Trend. The tendency for price movement to persist somewhat is one of the more robustly documented effects across asset classes, and it appears in academic research as momentum rather than as a chart pattern.

Support and resistance around obvious levels. Prior highs and lows, round numbers, and prior session extremes attract order clustering — because many participants place orders there. This is a self-referential effect but a real one.

Volatility regime. Whether ranges are expanding or contracting is directly observable and informative for position sizing, regardless of direction.

Volume confirmation. A move on unusually high volume reflects more participation than the same move on thin trading.

Choosing a timeframe

Lower timeframes are attractive because they produce more signals. That is precisely the problem.

Over one minute, price movement in a liquid market is dominated by order flow mechanics rather than by information. There is very little to extract, and each signal acted upon pays a spread. The ratio of noise to signal rises as the timeframe falls, while costs rise with frequency.

Daily and four-hour charts filter more noise, generate fewer decisions, and are consistent with the finding that lower trading frequency correlates with better outcomes.

Using charts sensibly

Read them as a record, not a forecast. A chart tells you what has happened and where participants have transacted. It is descriptive.

Combine with context. A technical level means more when it coincides with something — a policy meeting, an earnings release, a data print.

Define anything you rely on numerically. If a signal cannot be stated precisely enough to be tested, it cannot be evaluated, only believed.

Test before trusting. Apply the rule to historical data with realistic costs. Most rules do not survive this.

Never let a chart determine size. Position sizing derives from risk limits, not from how convincing a formation looks.

The bottom line

Candlestick charts are an efficient and useful way to see what a market has done. Learning to read them properly — range, close position, the balance within a period — is worthwhile.

The elaborate taxonomy of named patterns is a different proposition, and the evidence supporting it is thin. Treat charts as a description of what happened, be sceptical of anything presented as a reliable signal, and remember that the profitable part of trading is risk management rather than pattern recognition.

This article is educational and is not financial advice. Trading carries a high risk of loss.