Failed auction, in my notes, means price probes beyond a clear London level and comes back through it with pace, leaving the probe looking unloved.
On GBPUSD M1 I want: a defined level from Asia or early London, a break that fails within a few minutes, and a return through the level with spread still sane. Entry is back on the side of the failed probe, stop beyond the extreme, time stop short.
What is not a failed auction: a random wick in a quiet book, or a news spike I am fading without a pre-marked level. I tag these in the journal. Prop size stays smaller than personal because failed-auction losers can cluster when the day is actually trending. Recognition is pattern plus cost filter, not a name I slap on every reversal. Failed auctions that happen inside a Tier-1 window are ignored even if the pattern is pretty. Cost first.
Practical note on GBPUSD: cable tends to fake these more often around the 08:00 and 09:30 UK data slots, so I either skip failed auctions in those minutes or demand an extra candle of confirmation before entry.
Recognizing failed auction on GBPUSD M1
-
PropScalpDesk
- Posts: 521
- Joined: Sat Sep 19, 2026 7:50 pm
-
Shadow Trader
- Posts: 5594
- Joined: Wed Sep 09, 2026 5:19 pm
Re: Recognizing failed auction on GBPUSD M1
PropScalpDesk, on GBPUSD M1 I want a defined London level, a probe that fails inside a few bars, and price coming back through with costs still tolerable. Entry sits on the failed side, stop past the extreme, and I keep the clock short.
Near the early UK data windows I add a filter. Cable prints a lot of fake probes in those minutes, so I either stand aside or wait for one extra M1 close back through before I click.
On a prop account I keep size below what I would use on personal money. Two failed probes in a row on the same level is my cue to leave the idea alone for an hour, because losers stack when the session is trending.
A quiet wick with no pre marked level never counts for me. If I name the pattern only after the fill, that ticket was already wrong.
I cut the time stop shorter on these than on continuation trades because the edge fades quickly. Do you keep one clock for both, or tighten it on the failed probe version?
Near the early UK data windows I add a filter. Cable prints a lot of fake probes in those minutes, so I either stand aside or wait for one extra M1 close back through before I click.
On a prop account I keep size below what I would use on personal money. Two failed probes in a row on the same level is my cue to leave the idea alone for an hour, because losers stack when the session is trending.
A quiet wick with no pre marked level never counts for me. If I name the pattern only after the fill, that ticket was already wrong.
I cut the time stop shorter on these than on continuation trades because the edge fades quickly. Do you keep one clock for both, or tighten it on the failed probe version?
-
Shadow Trader
- Posts: 5594
- Joined: Wed Sep 09, 2026 5:19 pm
Re: Recognizing failed auction on GBPUSD M1
PropScalpDesk, one thing I have started measuring is the probe itself, because depth and duration separate real failures from the start of a new range. On GBPUSD M1 my better ones poke 3 to 8 pips past the level and are back through it inside four bars.
Probes deeper than about 12 pips that return are a different animal. Often the market has simply moved the auction and is building a fresh range, so fading the return puts me in the middle of it with a wide stop.
My stop goes 2 pips beyond the probe extreme plus the spread, never at the extreme itself. Cable likes to retest the high or low of a failed probe once, and a stop sitting exactly on it gets collected far too often.
The edge case I now skip completely is a failed probe on both sides of the same range within the first hour. When the high and the low each reject once, the session is two sided and the second fade usually goes nowhere.
For targets I take half at the range midpoint and treat the far side as a bonus. Do you size the probe in pips, or relative to how far the pair has travelled that morning?
Probes deeper than about 12 pips that return are a different animal. Often the market has simply moved the auction and is building a fresh range, so fading the return puts me in the middle of it with a wide stop.
My stop goes 2 pips beyond the probe extreme plus the spread, never at the extreme itself. Cable likes to retest the high or low of a failed probe once, and a stop sitting exactly on it gets collected far too often.
The edge case I now skip completely is a failed probe on both sides of the same range within the first hour. When the high and the low each reject once, the session is two sided and the second fade usually goes nowhere.
For targets I take half at the range midpoint and treat the far side as a bonus. Do you size the probe in pips, or relative to how far the pair has travelled that morning?
-
Shadow Trader
- Posts: 5594
- Joined: Wed Sep 09, 2026 5:19 pm
Re: Recognizing failed auction on GBPUSD M1
The NZD has a quirk that most major pairs lack: a scheduled commodity event. The Global Dairy Trade auction, held roughly every two weeks, produces a price index for milk powder, butter and other products that New Zealand exports heavily.
The result can nudge NZD pairs, particularly when it surprises the consensus. The effect is usually modest compared with a central bank decision, but in a quiet Asian morning it can be the main driver.
Because the auction concludes in the early hours of the New Zealand day, European and American traders are mostly asleep, and liquidity is thin. Spreads on NZD pairs are wider than usual.
For scalpers, the event is more of a map than a strategy. Knowing it exists explains an otherwise inexplicable 20 pip move in NZDUSD at an odd hour.
Similar commodity-linked patterns exist elsewhere: Australian iron ore, Canadian oil inventory data, Norwegian oil prices.
Which commodity linkage has proved useful here, and which turned out to be noise?
The result can nudge NZD pairs, particularly when it surprises the consensus. The effect is usually modest compared with a central bank decision, but in a quiet Asian morning it can be the main driver.
Because the auction concludes in the early hours of the New Zealand day, European and American traders are mostly asleep, and liquidity is thin. Spreads on NZD pairs are wider than usual.
For scalpers, the event is more of a map than a strategy. Knowing it exists explains an otherwise inexplicable 20 pip move in NZDUSD at an odd hour.
Similar commodity-linked patterns exist elsewhere: Australian iron ore, Canadian oil inventory data, Norwegian oil prices.
Which commodity linkage has proved useful here, and which turned out to be noise?