Advertisement IC Markets

Candlestick patterns: what statistics say

Discuss 1-minute to 15-minute price action setups, fading intraday momentum, key support/resistance zones, and proven short-term trading methodologies.
Post Reply
Shadow Trader
Posts: 5594
Joined: Wed Sep 09, 2026 5:19 pm

Candlestick patterns: what statistics say

Post by Shadow Trader »

Hammers, engulfing candles and dojis fill trading books. The claims are bold: this pattern signals a reversal with high probability.

Large-scale studies on currency data are less enthusiastic. Many patterns show forecasting power close to random once costs are included. A few, in specific contexts, produce a small edge. Context seems to matter far more than the shape: a pin bar at a major support level in a range behaves differently from a pin bar in the middle of a trend.

Another issue is subjectivity. Two traders looking at the same candle can disagree on whether it qualifies, which makes testing hard.

The pragmatic view treats candlesticks as a compact description of what price did during an interval, showing rejection or acceptance of a level. They are a language, not a prediction engine.

Rules that define a pattern in numerical terms, such as wick at least twice the body, make testing possible.

Has any candlestick pattern been tested here with objective rules, and did it hold up?
Recommended broker for automated trading & scalping IC Markets
FXS London Scalper
Posts: 978
Joined: Sun Oct 18, 2026 12:25 pm

Re: Candlestick patterns: what statistics say

Post by FXS London Scalper »

Shadow Trader wrote: The pragmatic view treats candlesticks as a compact description of what price did during an interval, showing rejection or acceptance of a level.
The "once costs are included" part is what kills most of them on M1 and M5. On the London open specifically, an engulfing candle at 08:05 on a 3 pip range is a very different object from one at 11:30. If candle patterns have any value, it is as a location-dependent trigger at a level marked beforehand, not as a stand-alone signal.
FXS Official Team | London Session Desk
Amanhasnoname
Posts: 2060
Joined: Fri Oct 30, 2026 9:45 am

Re: Candlestick patterns: what statistics say

Post by Amanhasnoname »

I spent a few weekends trying to answer this for myself on GBPUSD five minute charts, and the honest result was humbling. I coded a pin bar as a wick at least two and a half times the body, closing in the top or bottom third of its range, and let it fire everywhere. Across roughly eighteen months of data it was close to a coin flip once I took off my normal spread and commission.

The picture changed when I only counted pin bars that touched a level I had marked before the session started, usually the prior day's high or low or the Asian range edges. The sample shrank to a small fraction of the original, but the follow through improved enough to cover costs with a little left over. That tells me the level was doing the work and the candle was just confirmation that someone defended it.

Engulfing patterns never held up for me in any version I tried. Dojis I gave up on quickly because defining them objectively was a mess.
Post Reply