forex-scalping.com

Entering on the close of the five-minute candle

A popular rule: wait for the five-minute candle to close, then enter in the direction of the signal. The close confirms the pattern and avoids false starts.

The cost is lag. By the close, the move has already happened. The entry is a few pips worse than it might have been, and the stop must be wider to be placed beyond the candle's range.

There's also crowding. Many traders and bots act on the five-minute close. Spreads can tick wider and liquidity thin out for a second as everyone sends orders. Fills on those moments are not at their best.

A variation is to enter at the open of the next candle, or a few seconds after. That sidesteps the crowd by a moment, at the expense of more lag.

A comparison tells which works better: record results of entries at the close, at close plus three seconds, and at close plus ten. Any measurable difference in slippage and outcome shows the crowd effect.

The five-minute close is one of several clock-based moments, like the top of the hour, where behavior clusters.

First second of a new five-minute candle versus a few seconds later: how do fills compare?

Entering on the close of a five minute candle, rather than during it, means waiting for the candle to complete before acting on a signal. That avoids signals that appear mid candle and then disappear before the close, which is a common problem with indicators that update on every tick. The trade off is that the entry is later, since the move may have started before the candle closes. For strategies that rely on confirmation, entering at the close is usually more reliable. For strategies that rely on catching moves early, waiting can mean missing part of the move or entering at a worse price.

I tested this on my own fills for a month on EURUSD with a simple script that logged the price at the candle close and the price I actually received. Entries sent exactly at the close were filled on average about 0.4 pips worse than entries sent six or seven seconds later, and the spread on the first second of a new candle was visibly wider on most of the samples. It wasn't a huge difference, but on a six pip target it added up.

The cleanest version I found was to wait for the close, confirm the signal still held, and then send the order after a short random delay of three to eight seconds. I didn't want a fixed delay, since a fixed one just creates another crowd at a different moment. My fill quality improved without losing many trades.

One caution is that waiting costs you on fast moves. On news days I switch it off, because the extra seconds can mean the entry is two pips further away.