forex-scalping.com

Shrinking candles into a level and what they say about the push

Watch the last few candles before price reaches a level you care about. Not their direction, their size.

If each candle in the push is a bit smaller than the one before, the move is losing energy as it approaches. Buyers are still lifting price, but with less force each bar. That kind of approach often ends in a stall or a rejection at the level, because whoever was driving the move has mostly finished.

The opposite picture is candles getting bigger into the level. That's acceleration. Price arriving at resistance with expanding ranges is much more likely to break through, or at least to overshoot before anything else happens.

You can judge it by eye, or measure the high to low range of the last three or four bars and see if they're shrinking. Bodies matter too. A push where the bodies get smaller while the wicks stay the same size is a market that keeps trying and keeps failing to close far from the open.

For a scalper, this gives a cheap read before the level even gets touched. Shrinking candles into resistance favour a fade or at least no long. Expanding candles favour waiting for the break and the retest instead of fighting it.

It won't work every time. Nothing does. But it's visible on any chart without adding a single indicator.

Session context changes the reading. Candles naturally shrink as a session winds down, for example into the London lunch or late in New York, so shrinking bars at those times might just be the clock rather than exhaustion. The signal means more when it happens during an active window like the London open or the overlap, when you'd normally expect ranges to hold up. Comparing the bar sizes to the average for that time of day, rather than to the bars from an hour earlier, gives a fairer judgment of whether the push is really tiring.

Volume, even tick volume, can back it up. If the candles are shrinking and the tick volume is dropping too, fewer participants are involved in the push. If the candles are shrinking but tick volume is steady or rising, something else is going on, maybe absorption, where someone is selling into the buying and stopping price from moving far each bar. Both can lead to a rejection, but absorption tends to give a sharper turn. Tick volume on FX isn't real traded volume, but it's a reasonable proxy for activity on most brokers. Absorption usually shows as a cluster of small bars with wicks on the same side.

It helps to define the pattern tightly if you want to test it. Something like three consecutive candles each with a range at least a fifth smaller than the one before, ending within a few pips of a marked level. Then look back over a few weeks of charts and see what happened next. A vague version, such as candles looked smaller, can't really be tested, and you'll end up remembering only the times it worked. A strict definition keeps you honest about how often it actually leads to a rejection on your pairs. Thirty or forty examples per pair is enough for a first read.

Entry timing is where people go wrong with this. Seeing shrinking candles and selling straight away, before price has even reached the level, is jumping the gun. The level is still what makes the trade make sense. A cleaner approach is to wait for the touch and then a candle that closes back away from the level, and use the shrinking approach as the reason to be ready rather than the trigger itself. That way the stop has a logical place just beyond the level, instead of somewhere in the middle of the push. The approach tells you to get ready, the level tells you where.

A strict definition is the only way to test it. It is also worth recording the volume or tick count on those shrinking candles. If the ranges shrink and the tick count drops too, the push is fading. If the ranges shrink but tick count stays high, buyers and sellers are both active at the level and it can break either way. Many platforms show tick volume, and adding it as one extra column turns a vague pattern into something with two clear conditions. Logging that column alongside the ranges for a few weeks should show whether it actually separates the rejections from the breaks.