Price stopping a few pips short of an obvious level

[Fairman]

Price stopping a few pips short of an obvious level

Everyone has the same round number and the same prior high drawn on their chart. So it shouldn't surprise anyone that price often turns two or three pips before it gets there.

The people who wanted to sell at the level don't all wait for the exact print. Some of them put their limit orders a little in front so they actually get filled. Once enough of them do that, the turn happens early and the level itself never trades.

For a scalper this creates a choice. You can keep your order on the level and accept that you'll miss some of these turns, or you can shade your entry in front of it like everyone else. Neither is wrong. They just produce different trade lists.

What I'd look at is whether the approach into the level was fast or slow. A fast push that stalls a few pips short usually has real selling behind it. A slow drift that stalls short is often just a pause before price tags the level properly.

If you shade entries, your stop logic has to change too. A stop that was placed beyond the level now sits further from your entry, so either the size comes down or the stop moves in.

One more thing on this. The early turn is far more common at levels that are widely watched, like the day's open, a big figure or yesterday's high. On a level that only you have drawn, there's no crowd to front run it, and price is more likely to trade right into it before reacting. So the question of shading an entry really only comes up at the obvious levels, and those are the ones to test first if you want to see how your pair behaves.

[Fairman]

The pair matters here. On something like EURUSD the early turn tends to happen a pip or two in front of a big figure. On GBPJPY or gold the same behaviour shows up several pips out because the typical tick size and noise are bigger. If you shade entries by a fixed number of pips across every pair, you'll be too close on the volatile ones and too far on the quiet ones. A fraction of the current M5 ATR is a cleaner way to set the offset, because it scales with whatever the market is doing that day. Recalculate it each session rather than once a week, since ATR on the same pair can halve between a busy Tuesday and a quiet Friday.

[Fairman]

Missing the trade by a pip is annoying, but it's a different problem from losing on a trade you took. Some traders react to a few misses by moving every order further from the level, and then they start getting filled on moves that were never going to turn. Write down the misses separately for a few weeks. If the level really does get front run most of the time on a certain pair and session, adjust. If it's only a handful of memorable cases, leave the order where it is and accept the occasional miss as a cost of the method. It's also fair to ask whether the missed trades would even have worked. Some of them reach the level a minute later and then reverse anyway.

[Fairman]

There's also the case where price stops short, pulls back, then comes back and blows straight through. That second approach is a warning sign for anyone still holding a short from the first turn. The sellers who were in front of the level already got filled, and the second push finds less resistance. If you shorted early and price returns without much hesitation, it makes sense to tighten up or step out rather than wait for the level to defend itself, because the orders that were going to defend it may already be used up. A partial exit is a reasonable middle ground here, keeping some of the position but banking part of it while the second approach plays out.

[Fairman]

Some platforms let you see a basic depth of market from the broker, and people sometimes try to use it to spot the orders sitting in front of a level. On retail FX that view is usually only the broker's own book or an aggregated feed, so it won't show the real picture. The chart behaviour is a better guide. Look at how many times price has turned early at that level in past sessions, and whether those turns came with a clear rejection candle or just a slow stall that faded. Volume on the turn can add something too. An early turn on a spike in tick volume reads differently from one on almost no activity, where the stall might just be a lack of interest.

[WhitG47]

Volume on the turn can add something too. An early turn on a spike in tick volume reads differently from one on almost no activity, where the stall might just be a lack of interest.

Fairman, my experience was a bit different on the volume side. I tried watching tick volume on the turns on EURUSD for a couple of months, and to be fair it didnt separate the real rejections from the stalls as cleanly as I expected... the spikes were mostly just news and the London open, so I ended up trusting the candle more than the volume.

Mind you, that was only one pair and one broker's feed, so fair play if it works better for you.

G.

[FXS London Scalper]

Front running round numbers is real, and it shows clearly on EURUSD in the London session. A practical adjustment: if you want to sell at a round number, place the limit two or three pips in front, the same way the crowd does, or wait for a rejection of that front zone. For stops the logic runs the other way. If others are selling a few pips short of the level, a stop placed just above the round number can sit in a quieter spot than one placed right at it.

[FXS Prop Desk]

WhitG47, that matches what many people find with tick volume in spot currencies. It is a count of price changes at one broker, so it mostly rises when everything rises: data, the open, the fix. If volume is going to separate a rejection from a stall, it probably has to be compared with the same time of day on previous sessions rather than with the bars just before. A spike at 08:00 London is normal; the same reading at 11:40 would say a lot more.

[WhitG47]

WhitG47, that matches what many people find with tick volume in spot currencies.

FXS Prop Desk, fair enough on comparing against the same time of day, but I'm not sure it would have rescued it for me. A lot of my turns were in the London session anyway, so I'd have been comparing one busy bar with other busy bars, and I suspect the difference would be small... Mind you, I never tried it that way, so I can't say it wouldnt work.

My bigger worry is that you'd need a fair few sessions of data before the baseline means anything, and I can't be bothered with that much spreadsheet work for one filter.

G.

[Fairman]

That matches a lot of what I've seen, to be fair. Tick volume on a retail feed is just a count of price updates from one broker, so it jumps whenever the market gets busy for any reason, news, the London open, a rollover. It says "something is happening", not "buyers are stepping in". Two months on EURUSD with nothing clean to show for it is a reasonable result, not a failure.

Where it's helped me a little is as a relative thing against the same time of day. Say the 9:00 to 9:05 candle on a normal morning averages around 450 ticks. If price stalls 3 pips short of yesterday's high on 1,100 ticks, that's worth noticing, since a lot of activity produced no progress. The same stall on 150 ticks is mostly nobody caring. But I wouldn't act on it alone, just use it to make me look at the candle more closely, which agrees with where you ended up.

On a 0.5 lot trade with a 6 pip stop, I'd rather skip a setup than size up because of a volume spike.

Did you try comparing the volume against the same hour on previous days, or only against the last few bars? And did the candle shape itself, like long upper wicks, hold up better for you?