At 9:32 New York time, a five minute Nasdaq candle is pushing to 19,220, well above the prior high, and it looks like a clean breakout with 90 seconds left on the clock. By 9:35, the candle has sold all the way back to 19,150 and closed below the level it appeared to break. A trader who bought the breakout at 9:32 based on the candle in progress is now down on a trade that the finished candle says never actually happened. This is the most expensive lesson in candlestick reading, and almost everyone learns it the hard way at least once.
Every candlestick pattern discussed anywhere, engulfing, pin bar, doji, star, is defined by where the candle closes, not by where it travels during the bar. An unfinished candle is not a smaller version of the pattern. It is not the pattern at all yet, because the close is the only part of the bar that cannot be undone by the next few ticks.
Why intrabar prices lie
During the life of a candle, price can spike through a level, trigger the visual shape of a breakout or a pin bar, and then fully reverse before the bar ends, and none of that intrabar movement is preserved once the candle closes. Only the open, high, low and close survive. A trader watching the tick by tick price sees every fake signal the bar produces along the way; a trader waiting for the close only sees the one version of events the market actually settled on. Acting on the intrabar picture means reacting to information the chart itself is about to discard.
What waiting for the close actually costs
The honest tradeoff is entry price. Waiting for a 5 minute candle to close before acting on a breakout means giving up whatever distance price moves in the confirmation, sometimes nothing, sometimes 10 or 15 pips on EURUSD if the move is running hard. That cost is real and worth naming rather than pretending discipline is free. But compare it to the alternative: a EURUSD breakout entered mid candle that reverses before the close, stopped out for a loss, is a considerably worse outcome than a slightly worse entry price on a breakout that actually held.
When acting before the close is defensible
The close rule is strongest on the timeframe you are trading, but it bends sensibly on higher timeframes. Waiting for a full daily candle to close on gold before reacting to an intraday level break can mean giving up hours of a legitimate move. A reasonable compromise many traders use is checking the close of a lower timeframe, say H1, as a proxy confirmation for a level marked on the daily chart. That is not abandoning the principle, it is applying the same logic, wait for a finished bar, at a resolution that fits how fast you need to act.
Building the habit
The fix is mechanical, not psychological: build a rule that the entry trigger is a closed candle beyond the level, full stop, and refuse orders based on price still moving inside the bar. On a fast market like Nasdaq around the open, this means literally waiting out the clock on a 5 minute or even 1 minute candle before clicking anything, which feels agonizingly slow in the moment and is nearly always the correct trade.