Japanese candlesticks

What a Japanese candlestick is

Following on from What is a chart, a candlestick is one of three ways of presenting the same price. The open and the close define the body; the high and the low define the wicks, also called shadows. A long wick at a key zone tells you that price was rejected during that period; the same wick in the middle of a range often means nothing operationally. On the Warsaw Stock Exchange one D1 candle is a whole session, with a possible gap the next day; on an H1 CFD it is one hour of broker quotes carrying a spread.

Learn the behaviour at a level — rejection, acceptance, a strong close — instead of memorising forty Japanese names. The name helps you in your journal; the edge comes from invalidation and position size.

How to read a candle in context

A pin bar, an engulfing candle or a “strong close” works as a description of behaviour at a level, not as a stand-alone signal. The order is: higher timeframe bias → support/resistance zone → the candle as a trigger or a filter. An M5 candle against the daily trend is usually noise, not a “trend reversal”.

Measure instead of naming: the body-to-range ratio

Pattern names are debatable — two people will name the same candle differently. The ratio of the body to the whole range of the candle is a number, and there is nothing to argue about:

body  = |close − open|
range = high − low
ratio = body / range        (0 = doji, 1 = candle with no wicks)

Calculate this for the last 100 candles of your instrument on a single timeframe and sort the results into three buckets:

Write down the distribution you get. That is your own baseline: if 70% of candles on your instrument have a ratio below 0.25, then a “strong close” is a rarity and you certainly do not have to hunt for one every day.

When a candle says “do not trade”

The same shape in a different place is a different event. This is the most common source of losses on patterns.

Three hard “do not trade” rules worth writing into your plan:

1. An unclosed candle is not a signal. A decision based on a candle that is still forming can be undone in the last second of the interval.
2. A candle in the middle of a range is not a signal, however textbook it looks — no level means no place to put an invalidation.
3. A candle in the first minutes after the session opens (or in the first minutes after a data release) has a distorted range. Wait until the interval has closed.

Five-session protocol — fill it in yourself

This is the most important part of the lesson. Without filling it in, you do not know whether your “pretty candles” carry any edge at all.

After five rows, calculate two numbers: the average result in R for candles at a level and for candles without a level. If the second number is not clearly worse, then in your execution the level is not yet doing the job of a filter — and that is the correct conclusion to draw from the exercise.

Educational example (USD)

Assume a demo account of $10,000. You plan a long at a zone on an index CFD with 1% risk = $100. A stop beyond the low of the rejection candle corresponds, for teaching purposes, to 40 points. You calculate the position size so that being stopped out costs $100 — not because “the pin bar looks pretty”. The numbers are illustrative; on a live account always check the point value and the spread at your broker.

Three candle behaviours worth remembering

Instead of learning dozens of names, it is enough to recognise three basic behaviours at a level:

Any of these behaviours without the context of a level and a higher timeframe bias is just an interesting shape on the screen — it only becomes meaningful together with the zone in which it appears.

Multi-timeframe — hygiene

Bias from D1 or H4, the zone from H1/H4, the timing from a lower timeframe. Write down in your plan whether you require the trigger candle to close, or whether a wick is enough. Without that rule, every wick becomes an emotional discussion.

What to avoid with candles

• Entering on the shape alone, with no level and no invalidation plan.
• A tight stop in the middle of a wick — noise will eat the setup before any move appears.
• Overfitting: searching for a pattern backwards until history “fits”.
• Changing the timeframe mid-trade just so the pattern looks textbook.
• Treating one candle on the Warsaw Stock Exchange and one on a 24/5 CFD as identical risk, with no allowance for costs.

Common mistakes in candle interpretation

• Giving meaning to a single candle in the middle of a range, far from any significant zone.
• Entering on the shape of the pattern alone, without checking whether it agrees with the higher timeframe bias.
• Learning candle names in Polish and English like vocabulary for an exam, instead of practising how to recognise behaviour on a demo account.
• Ignoring the difference between a closed candle and a candle still forming — decisions based on an unclosed candle carry the extra risk that the picture will change.
• Measuring a pattern on far too small a sample — three successful pin bars are not an edge, they are three events.

Exercise

On a single instrument (EURUSD or WIG20), mark ten “pretty” candles from the past that sit away from any level, and ten that sit at a clear D1/H4 zone. For each one, calculate the body ratio and record whether over the next five candles on that timeframe price continued in the direction of the rejection — yes or no. Compare the percentages in a 20-row table, broken down by ratio bucket. The goal is a measurable difference between contexts, not an impression.

Measurable exercise

Pick one demo instrument. Write down: your capital, 1% risk in USD, the distance to invalidation, the resulting position size, and whether you end the session once you hit a −2R limit. One row in a spreadsheet beats three “gut feelings” from a chat.

Common questions

Does a pin bar always mean a reversal?

No. It is a rejection of price within a given interval. Without a level and an invalidation plan it is often noise — especially against the higher timeframe trend.

How many candlestick patterns do I need to know at the start?

It is enough to understand the body, the wicks and two or three behaviours: rejection, strong close and engulfing. The rest of the names will not replace a level and risk management.

Why calculate the body-to-range ratio when I can see it by eye?

Because 'seeing it' is different for two people looking at the same chart. The ratio gives you a number you can write in your journal and compare across sessions and instruments — and after 100 observations you have your own baseline for what counts as a strong close.

Do candles behave differently on the Warsaw Stock Exchange than on forex?

The OHLC logic is the same. What differs is the session, the gaps and the volume — on session-based equities, gaps after the close are normal; on 24/5 forex they are rarer.